ANN-Based MPPT Control of a Standalone PV System with Bi-Directional Battery Management and NPC Inverter for Dynamic 3-Phase AC Loads

ANN-Based MPPT Control of a Standalone PV System with Bi-Directional Battery Management and NPC Inverter for Dynamic 3-Phase AC Loads | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article ANN-Based MPPT Control of a Standalone PV System with Bi-Directional Battery Management and NPC Inverter for Dynamic 3-Phase AC Loads K . Vinoth Bresnav, Dr.S. Singaravelu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6831075/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract This paper presents an intelligent energy management system for a standalone photovoltaic (PV) system integrated with a lithium-ion battery and a three-level neutral-point clamped (NPC) inverter to supply a three-phase AC load. An Artificial Neural Network (ANN)-based Maximum Power Point Tracking (MPPT) controller is employed to enhance the efficiency and dynamic response of the PV system under varying irradiance conditions. The extracted power is regulated through a DC-DC converter, maintaining an optimal DC link voltage. To ensure reliable power delivery and load balancing, a bi-directional DC-DC converter is interfaced between the battery and the DC link. A Proportional-Integral (PI) controller governs the charging and discharging of the battery based on the power difference between the PV generation and the AC load demand. This mechanism allows for smooth energy transition, enabling battery charging during surplus PV generation and discharging during power deficits. The DC link supplies power to a three-level NPC inverter, which is responsible for converting DC power into high-quality AC output for three-phase loads. An inverter control strategy is implemented to maintain voltage stability and minimize total harmonic distortion (THD). The proposed system is simulated in MATLAB/Simulink, and its performance is analyzed under dynamic load and environmental conditions. Results demonstrate improved MPPT accuracy, effective battery management and stable AC output voltage, showcasing the suitability of the system for off-grid applications. Artificial Neural Network (ANN) MPPT Standalone PV System Bi-Directional Converter Battery Energy Storage NPC Inverter 3-Phase AC Load DC Link Energy Management PI Controller Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 1. INTRODUCTION Electrical energy is fundamental not only to the daily lives of individuals but also plays a pivotal role in the economic development of nations. In light of global concerns regarding environmental sustainability and energy security, renewable energy sources have emerged as a promising alternative to conventional fossil-based systems. However, the inherently intermittent and weather-dependent nature of sources such as solar and wind limits their standalone reliability [ 1 ].To address this challenge, Hybrid Renewable Energy Systems (HRES), which combine multiple energy sources, have gained significant attention.As the deployment of renewable energy systems increases, efficient control and sizing of components become critical for ensuring system stability and economic feasibility. Modern HRES designs also incorporate demand-side strategies such as demand response and even electric vehicles (EVs) as flexible storage units or controllable loads, further expanding their potential. In this context, intelligent energy management systems that include accurate MPPT, optimal battery utilization and high-quality power conversion are essential. The standalone PV systems are often integrated with energy storage components such as secondary batteries, forming PV/B systems [ 2 ]. These configurations not only provide backup during periods of low irradiance but also enhance the system’s ability to meet dynamic load demands. Such systems are particularly critical in remote and off-grid areas where grid extension is not viable. Furthermore, applications extend beyond terrestrial uses into space missions, where PV/B systems act as the sole energy source for satellites and spacecraft. In the search to mitigate climate change and reduce carbon emissions, renewable energy systems particularly PV systemshave garnered substantial global attention. Among the core technical challenges in PV system deployment is the extraction of maximum power under varying environmental and dynamic load conditions. MPPT algorithms play a critical role in enhancing energy harvest by dynamically adjusting the operating point of the PV array to match optimal conditions [ 3 ].Over the years, a wide array of MPPT techniques has been developed, ranging from conventional methods such as Incremental Conductance (INC) [ 5 ] and Perturb & Observe (P&O) [ 4 ], to more advanced hybrid and artificial intelligence (AI) [ 7 ] based controllers including Fuzzy Logic [ 6 ] and Artificial Neural Networks (ANN).The existing comparative studies indicate that while traditional methods like INC and P&O offer simplicity and reasonable performance, they often fall short in complex, rapidly changing environments or partial shading conditions. On the other hand, AI-based controllers demonstrate superior adaptability, accuracy and robustness [ 8 ].AI-based approaches such as Fuzzy-PSO, ANN, ANFIS, and their combinations exhibit competitive or higher accuracy, often exceeding 98% [ 3 ]. These findings underscore the growing potential of intelligent control systems in optimizing PV output and improving energy management in standalone or hybrid configurations. Standalone PV systems are increasingly vital in remote and off-grid applications, but face challenges related to fluctuating solar irradiance and battery reliability [ 11 ]. Conventional charging techniques and basic MPPT algorithms often underperform in dynamic environments [ 10 ]. Recent studies have proposed adaptive neural control strategies using structures like ADALINE, achieving high MPPT efficiencies [ 9 ]. Problem Statement : Standalone photovoltaic systems face critical challenges in maintaining power continuity and voltage stability, particularly under fluctuating environmental conditions and variable load profiles. While traditional MPPT methods provide a basic level of power optimization, they often struggle to adapt in real time to sudden irradiance changes or partial shading, resulting in suboptimal energy capture. Similarly, conventional battery management strategies are limited in dynamically adjusting charging and discharging operations, leading to inefficient energy utilization and potential battery degradation. Moreover, the integration of advanced inverter topologies such asNPC inverters is seldom optimized in a unified control framework. A comprehensive, intelligent solution is needed to simultaneously optimize MPPT performance, battery behavior and AC power quality for reliable standalone operation. Current literature tends to focus separately on MPPT optimization or battery management, lacking an integrated control framework that coordinates both functions effectively within standalone PV systems. While ANNbased MPPT controllers have shown superior tracking efficiency, their integration with real-time battery control through a bi-directional converter is understated, especially in the context of supplyingthree-phase AC loads. Furthermore, limited research has been conducted on harmonizing ANN-based MPPT, dynamic battery interface control and three-level NPC inverters into a single system which is particularly in terms of ensuring consistent AC power quality and stable DC-link voltage across load and generation scenarios [ 12 ]. Inspired by recent advancements in intelligent energy management, this study implements a robust ANNbased MPPT control strategy integrated with a bi-directional DC-DC converter for effective battery management and a NPC inverter for three-phase AC load supply in standalone PV systems. This dual-layer intelligence harnesses the predictive and adaptive strengths of ANN to enhance power tracking efficiency, optimize battery utilization, and ensure stable, high-quality power delivery under dynamic environmental conditions. Objectives of the Study To develop a predictive and adaptive ANN-based MPPT controller to improve energy harvesting efficiency in standalone PV systems under rapidly changing irradiance conditions. To design a PI-controlled bi-directional DC-DC converter for managing battery charging/discharging based on PV output and AC load demand, ensuring optimal battery usage. To integrate a three-level NPC inverter with the DC-link and assess its ability to maintain high-quality three-phase AC power delivery. To evaluate the combined system's performance through simulation, focusing on MPPT accuracy, battery SOC dynamics DC-link voltage regulation and AC voltage waveform quality. To validate the proposed system’s capability in maintaining stable, uninterrupted power supply under practical and dynamic operating scenarios, including variable solar input and load profiles. This work is organized as follows: Section 1 presents the introduction, problem statement, research gap and objective of the work; Section 2 describes the system description which includes overview of proposed system architecture, including PV system description, the ANN-based MPPT controller, designof bi-directional converter with Battery configurations and NPC inverter; Section 3 discusses the modeling and simulation of the integrated system in MATLAB/Simulink; Section 4 presents the results and performance analysis under various environmental and dynamic load conditions. Finally, Section 5 concludes the study with key findings, contributions and suggestions for future research. 2. System Description A) Overview of the proposed system architecture The proposed standalone PV system architecture integrates intelligent control strategies to optimize energy harvesting, storage management, and AC power delivery. At the core of the system is an ANNbasedMPPT controller, which dynamically adjusts to varying environmental conditions to maximize power extraction from the solar PV array. This power is regulated via a DC-DC boost converter and stabilized across a DC-link capacitor. A PI-controlled bi-directional DC-DC converter interfaces a battery energy storage unit, enabling smart charge/discharge operations based on PV power flow and load demand. To ensure high-quality AC output, a three-level NPC inverter is employed which offering reduced harmonic distortion and improved efficiency. The coordination of ANN-based MPPT, battery management and NPC inverter operation forms a unified control scheme aimed at delivering consistent, high-quality three-phase power during fluctuating solar irradiance and load conditions as shown in Fig. 1 . B) Description of PV system Parameters The inclusion of a 2000 kW PV system in this research plays a critical role in demonstrating the practicality and scalability of the proposed ANN-based MPPT and bi-directional battery management strategy for large-scale standalone applications.The following section provides a detailed architecture of the 2000KW PV system, which is particularly chosen for integrating this system with a high current network. In addition to the commonly used models, other configurations such as single-diode and two-diode models are also available. The two-diode model introduces a second diode connected in parallel with the first, effectively mimicking the behavior of a single-diode equivalent circuit. While the single-diode model is simpler and involves fewer parameters, the two-diode model provides a more detailed representation of PV cell behavior. Studies have demonstrated that both simulated and experimental I–V and P–V characteristics of solar panels yield closely matching results. In modeling, a PV cell is typically depicted as an electrical circuit comprising a diode in parallel with a current source that represents the photo-generated current. These individual cells are then grouped into PV modules, configured in both series and parallel to deliver the required power output. An ideal PV cell can be modeled as a DC source in parallel with a diode [ 10 ].To represent real-world losses and leakages, series and shunt resistances are incorporated into the ideal model, forming a more accurate practical equivalent. Figure 2 illustrates the structure of such a PV cell model. The electrical behavior of a practical photovoltaic cell can be expressed through a modified current-voltage relationship, incorporating the effects of internal resistances and non-idealities. The output current ( \(\:{I}_{pv})\) from the PV cell is determined using the following expression: I = I pv = I ph – I s \(\:[{e}^{\left(\frac{\left({V}_{pv}+{I}_{pv}.{R}_{S}\right)}{{N}_{s}{V}_{T}}\right)}-1]\) - \(\:\frac{({V}_{pv}+{I}_{pv}.{R}_{S})}{{R}_{sh}}\) (1) Where, I ph represents the photo-generated current, \(\:{I}_{S}\) is the reverse saturation current of the diode, \(\:{V}_{pv}\) ​ and \(\:{I}_{pv}\) ​ are the output voltage and current of the PV cell, \(\:{R}_{S}\) ​ and \(\:{R}_{sh}\) ​ are the series and shunt resistances, \(\:{N}_{S}\) ​ denotes the number of series-connected cells and \(\:{V}_{T}\) ​ is the thermal voltage which equals \(\:a.K.T/q\) . Where a is diode ideality factor, q denotes electron charge, K for Boltzmann constant and T is cell temperature. Table 1 presents all the parameters of the PV module used in this study, specified under Standard Test Conditions (STC), which correspond to an irradiance of 1000 W/m² and a temperature of 25°C. Table 1 Intrinsic electrical parameters associated with the PV panel being evaluated SL. No. PARAMETERS Range 1 Number of cells per module 128 2 Open circuit voltage (Voc) 85.3V 3 Short circuit current (Isc) 6.09A 4 Voltage at MPP (Vmpp) 72.9V 5 Current at MPP (Impp) 5.69A 6 Power at MPP (Pmpp) 414.801W 7 Parallel Strings 700 8 Series connected Modules per string 7 9 Module Voltage at MPP (Vmpp) 510.3V 10 Module Current at MPP (Impp) 3983A 11 Module Power at MPP (Pmpp) 2032KW 12 Diode Ideality factor 0.87223 Figures 3 and 4 illustrate the graphical relationships between voltage and current, and between power and voltage, respectively, under different irradiance and temperature conditions. C)Model of ANN MPPT method A2000KW photovoltaic system using ANN for Maximum Power Point Tracking is a modern and intelligent technique aimed at maximizing energy extraction under varying environmental conditions. In this approach, key PV system parameters such as solar PV voltage and current are used as inputs to the ANN, while the output is the duty cycle, which adjusts the operating point of the DC-DC Link to track the maximum power point efficiently. To train the ANN model, historical data comprising PV voltage, current and the corresponding optimal duty cycles is collected. This data enables the ANN to learn the nonlinear relationship between the input conditions and the appropriate duty cycle needed to achieve MPPT. The training process involves optimizing the network’s weights and biases to minimize the error between the predicted and actual duty cycles.The ANN-based MPPT controller in this work is designed and implemented using the SIMULINK platform, as shown in Fig. 5 . The difference between the predicted and actual duty cycle values is evaluated using a loss function, which serves as a measure of prediction error. This function helps quantify how far the model’s output deviates from the expected results. To minimize this error, the ANN adjusts its internal weights and biases based on the gradient of the loss function. In this work, the Levenberg-Marquardt optimization algorithm is used for this purpose.The Levenberg-Marquardt algorithm is specifically designed to solve nonlinear least squares problems and combines the advantages of both the gradient descent and Gauss-Newton methods. Named after Kenneth Levenberg and Donald Marquardt, who independently developed the approach, the algorithm works by iteratively updating the model parameters to reduce the sum of squared differences between the actual and predicted values. A properly trained ANN model typically produces an error histogram with values centered around zero, indicating high prediction accuracy, as illustrated in Fig. 6 . The division of data into training, validation, and testing sets used for developing the ANN model is shown in Fig. 7 . A regression value (R) close to 1 signifies a strong correlation between predicted and actual outputs, confirming the effectiveness of the trained model. The developed ANN model achieved a mean squared error of 1.6450e-10 and a minimum gradient of 9.99e-8, indicating its strong performance and suitability for the intended application. Following the training phase, the ANN is integrated into the closed-loop control system, where it continuously receives real-time inputs of PV current and voltage. Based on these inputs, the ANN predicts the optimal duty cycle required for efficient operation under the given conditions. The SIMULINK implementation of the trained ANN model for this study is depicted in Fig. 8 . D) Design of PI controller bi-directional converter with Battery configurations The bi-directional DC-DC converter facilitates both charging and discharging of the battery bank based on PV surplus and load demand. An interleaved bidirectional buck-boost topology [ 14 ]is used to support high power with reduced ripple and improved thermal distribution. The battery specifications are as shown in Table 2 . Table 2 Battery Specifications Battery Parameters Range Nominal Voltage( \(\:{Vbat}_{Nominal}\) ) 300V Capacity 6500ah Initial State of Charge ( \(\:{SOC}_{Initial}\) ) 25% Response time 30sec A buck-boost converter topology is ideal due to its ability to handle both power directions. The converter operates in buck mode to charge the battery when PV generation exceeds load demand and in boost mode to discharge when the load exceeds PV generation.The converter is governed by a PI (Proportional-Integral) controller, regulating the battery current to:Maintain SOC within operational limits, avoid overcharge/over-discharge and ensure smooth transitions between charging and discharging.The controller compares the reference current (based on power mismatch) with the actual battery current and adjusts the duty cycle accordingly. The battery power range \(\:{\mathbf{P}\text{b}\text{a}\text{t}}_{\text{m}\text{a}\text{x}}\) has been decided by the expression 2, $$\:{\mathbf{P}\text{b}\text{a}\text{t}}_{\text{m}\text{a}\text{x}}={\mathbf{V}\text{b}\text{a}\text{t}}_{\text{N}\text{o}\text{m}\text{i}\text{n}\text{a}\text{l}}\times\:{\mathbf{I}\text{b}\text{a}\text{t}}_{\text{m}\text{a}\text{x}}=975\text{K}\text{W}\:\text{f}\text{o}\text{r}\:\text{t}\text{h}\text{e}\:\text{s}\text{e}\text{l}\text{e}\text{c}\text{t}\text{e}\text{d}\:\text{b}\text{a}\text{t}\text{t}\text{e}\text{r}\text{y}$$ 2 For high power converters, ripple current should be < 20% of max current. The control objective is to maintain battery current smoothness (30s dynamic response), preserve SOC profile and prevent oscillations during load transients. The design of dc-dc bi-directional converter as follows, \(\:\text{L}=\frac{{\text{V}}_{\text{d}\text{c}}\times\:\text{D}}{{\text{f}}_{\text{s}}\times\:\varDelta\:\text{I}\text{b}\text{a}\text{t}\text{t}}\) = 47µH (3) \(\:\text{C}=\frac{{I}_{Bat}^{max}}{{\text{f}}_{\text{s}}\times\:\varDelta\:\text{V}\text{b}\text{a}\text{t}\text{t}}\) = 33mF (4) Where switching frequency ( \(\:{\text{f}}_{\text{s}})\) is 10KHZ, maximum battery current ( \(\:{I}_{Bat}^{max}\) ) is 3250A, allowable ripple current ( \(\:\varDelta\:\text{I}\text{b}\text{a}\text{t}\text{t}\) ) is 650A, dc link voltage is 600V.The transfer function of converter with PI controller for battery current control (current mode control) is given as, G(s) = \(\:\frac{Vbat}{L.s}\) (5) PI controller transfer function is given as, $$\:{G}_{c}\left(s\right)={K}_{p}+\frac{{K}_{i}}{s}$$ 6 Where \(\:{K}_{p}\) is 0.05 and \(\:{K}_{i}\) is 0.01 as derived form automatic tuning in Matlab platform for 30 seconds settling time. E) Design of Low-voltage high-current 3-level NPC inverter Multilevel inverters have emerged as a powerful solution for medium- to high-power applications, especially in renewable energy systems. Among them, the Three-LevelNPC inverter offers several advantages such as, reduced voltage stress on switching devices, improved harmonic performance and higher efficiency due to lower switching losses.In an NPC topology, each leg of the inverter includes four switches and two clamping diodes, enabling three distinct output voltage levels: +Vdc​/2, 0 and − Vdc​/2. This finer resolution results in a closer approximation to a sinusoidal waveform, thus reducing output Total Harmonic Distortion (THD) and filter requirements. The three-phase 1500 kW load is evenly distributed which is resulting in a phase power of 500 kW. Given an output line-to-line RMS voltage of 400 V, the phase voltage calculates to approximately 230.9 V. Assuming a power factor of 0.9, the RMS phase current is approximately 2402 A which is requiring the use of multiple high-current power switches in parallel within each inverter leg. To accommodate the 550 V DC-link voltage with adequate margin, each switch should have a voltage rating of at least 900 V and a current rating of at least 400 A.Therefore, six such devices per switch are implemented. The DC-link capacitor bank is designed to store 10 kJ of energy, requiring a total capacitance of around 66,000 µF which is implemented using low-ESR, high-ripple film capacitors. To limit output current ripple, the inverter uses an output filter inductor of approximately 31.8 µH which is capable of handling over 2500 A. For modulation and control, a hybrid approach combining Space Vector PWM and level-shifted PWM is employed to ensure efficient DC-link usage and neutral-point voltage balance, while an ANN-based current controller enhances dynamic response and nonlinear load compensation. The design-oriented specifications are as listed in Table 3 . Table 3 NPC Inverter Specifications Parameters Range Load Power 1500 kW Output Voltage (L-L RMS) 400 V (3-phase standard) Output Frequency 50 Hz DC-Link Voltage 550 V Topology 3-Level NPC Switching Frequency 5–10 kHz Voltage Regulation \(\:\pm\:\) 2% Response time 96% Output THD < 2% 3. Simulink Model of Integrated system The Fig. 9 illustrates the NPC Inverter Circuit, which features a three-level NPC topology powered by a 550 V DC-link. The inverter converts the regulated DC voltage into a three-phase AC output, ensuring high-quality voltage delivery to the load. A harmonic filter and a Y-grounded transformer are employed to minimize output harmonics and provide electric isolation. Each phase leg of the NPC inverter is configured to handle high current by paralleling multiple power switches, and appropriate filtering is added to maintain grid-compatible output. The Fig. 10 shows the Inverter Control Circuit, which employs an ANN-based MPPT Controller to ensure optimal power extraction from the PV array. The controller dynamically adjusts the reference voltage ( \(\:{\text{U}}_{\text{r}\text{e}\text{f}}\) ) for the inverter based on real-time irradiance and PV current/voltage inputs. A phase-locked loop (PLL) enables synchronization with the reference frame and a PI-regulated \(\:{\text{V}}_{\text{D}\text{C}}\) controller maintains DC-link voltage stability by generating the appropriate d-axis current reference. This is followed by a current regulator that translates these references into modulation signals for the inverter switches. The control system integrates both the MPPT and inverter regulation schemes to ensure seamless operation under dynamic solar and load conditions. The integrated system includes a solar PV array connected with a battery, which is interfaced to the AC load via a 3-level NPC inverter and a harmonic filter, ensuring sinusoidal current injection into the grid as shown in Fig. 11 . The control architecture further incorporates a Phase-Locked Loop (PLL) for synchronization, a voltage and current regulator and \(\:{\text{U}}_{\text{a}\text{b}\text{c}}\) reference generation for stable operation.The battery energy storage system (BESS) is interfaced through a bi-directional DC-DC converter controlled by a PI-based feedback control loop. This controller adjusts the battery's charge/discharge power in real time, considering the instantaneous PV current and load demand. The Battery Control Block, as highlighted in the second schematic, regulates the charging process using a logic that compares a fraction of PV current with the battery current and adjusts the usage signal accordingly to ensure effective energy balancing and SOC (state of charge) management as shown in Fig. 12 . The coordinated interaction of these subsystems enables efficient power flow regulation, voltage stability and system reliability across different load scenarios. The subsequent section presents detailed simulation results and performance analyses under various dynamic operating conditions which demonstrating the effectiveness of the proposed ANN-based MPPT control and energy management strategies. The integrated Simulink model effectively demonstrates the coordinated operation of the solar PV system, battery storage, NPC inverter and ANN-based MPPT control. This system ensures the optimal power management and system stability under three phase load and variable environmental conditions. 4. Results and Performance Evaluation under Dynamic Environmental and Load Conditions This section presents a comprehensive analysis of the PV/battery/NPC inverter system under dynamically varying environmental and load scenarios. The simulation is designed to reflect realistic operating conditions by incorporating sudden changes in load demand and assessing the system's response. The evaluation focuses on four key aspects: the accuracy of PV voltage and power tracking using the P&O MPPT algorithm, the battery's charging/discharging behavior governed by a PI controller, the quality of AC load voltage includingTHD and the overall system stability and energy flow management. These performance metrics are critical for ensuring reliable and efficient energy delivery in standalone or microgrid applications. i) PV power and voltage tracking accuracy The implemented ANN MPPT algorithm effectively tracks the Maximum Power Point of the PV array. The PV voltage oscillates tightly around the Vmpp (510.3 V), and power stabilizes near the rated 2032 kW under standard test conditions, with minimal deviations, indicating high tracking accuracy as observed from Fig. 13 . ii) Battery charging/discharging behavior 0–4.5 s : The PV output exceeds the load demand (1500 kW), and the battery charges with the excess 500 kW. 4.5 s onwards : When load increases to 2500 kW, the PV is insufficient and the battery discharges to support the remaining demand (500 kW). SOC profile confirms this behavior with a smooth charging tendency initially which is followed by controlled discharging after the load step as shown in Fig. 14 . iii) AC load voltage waveform and THD analysis The three-phase output voltage maintains a stable 400 V (rms) with proper 120° phase shift.FFT analysis reveals a THD (< 2%) which proves the NPC inverter’s high-quality output and filtering effectiveness.Waveform plots before and after the load step show no significant distortion, indicating strong voltage regulation as visualized in Fig. 15 . iv) Insights into system stability and energy flow The system remains stable under both static and dynamic conditions, with continuous power flow management as shown in Fig. 16 .Duty cycle modulation through the PI controller ensures smooth current transitions.Energy balance is maintained throughout the simulation, with the battery acting as an efficient buffer during load fluctuation.No voltage collapse or instability is observed, confirming the robust design and control logic as proved with Fig. 17 . Overall, the energy management strategy effectively coordinates power among PV, battery and load to maintain system reliability.The unique summary of this work has been elaborated using Table 4 , which captures key technical observations of PV voltage and power tracking, battery dynamics, AC load performance and system stability under variable load and environmental conditions. Table 4 Results and Performance Metrics under Dynamic Environmental and Load Conditions Metric Description Observation PV Voltage Tracking (Vmpp) MPPT voltage convergence accuracy ~ 510.3 V with minor ripple (< 1%) PV Power Output (Pmpp) Maximum power generation capability ~ 2032 kW sustained under STC Battery Operation Mode (0–4.5s) Charging phase (excess PV power utilized) Battery charges at ~ 500 kW Battery Operation Mode (4.5–14s) Discharging phase (PV power deficit) Battery discharges ~ 500 kW to support extra load SOC Behavior Response to charging/discharging transitions Smooth variation, with logical increase/decrease AC Load Voltage (RMS) Per-phase voltage magnitude maintained 400 V (± small ripple) Phase Shift in Load Voltage Verification of 3-phase sinusoidal quality 120° phase shift among phases THD Measured using FFT analysis < 2% Load Step Change at 4.5s Additional 1000 kW applied Total load rises from 1500 kW → 2500 kW System Stability Under dynamic PV/load conditions Stable operation, no overshoot or voltage dip Energy FlowManagement Balance between PV, battery, and load Dynamic and accurate with PI controller support Duty Cycle Response PI controller modulation of DC-DC converter Smooth variation without oscillations 5. Conclusion This study presents a robust and adaptive energy management system for a standalone PV/battery-powered three-phase load configuration which is validated through a 14-second MATLAB simulation incorporating realistic irradiance fluctuations and dynamic load variations. The system employed AI based MPPT tracking and a PI-controlled bidirectional DC-DC converter, ensuring reliable power flow continuity even during sudden demand surges and irradiance dips. Key findings include high PV power extraction efficiency (2032 kW), rapid and stable battery mode switching (charging/discharging) and THD levels below 2% by maintaining a steady 400 V AC output with accurate 120° phase shifts.The synchronized operation between the PV source, battery and NPC inverter under dynamic load transitions demonstrates the system’s practical viability and resilience. This work uniquely highlights the integration of AI control strategies with traditional components to form a scalable framework for renewable-powered standalone systems. Future research could explore Cybersecurity-aware control algorithms, especially for IoT-enabled PV-battery microgrids to safeguard communication and operational integrity. Declarations AUTHOR BIOGRAPHIES Mr. K. Vinoth Bresnav M.E., an Assistant Professor Department of Electrical Engineering, with over 13 years of academic and industrial experience. Currently pursuing his Ph.D. at Annamalai University, he specializes in Power Systems, Control Systems, and Electrical Machines. He has authored books and published research in reputed journals and conferences. His contributions include academic coordination, placement, and NBA documentation. He is also GATE qualified and actively engaged in FDPs, workshops, and NPTEL-certified courses. Dr. S. Singaravelu, Professor in the Department of Electrical Engineering, Annamalai University, Annamalai Nagar, Tamilnadu, holds a Ph.D. with expertise in Electrical Machines, Power Systems, Renewable Energy, HVDC, and Smart Grid technologies. With over three decades of academic and research experience, he has published extensively in reputed national and international journals. His research focuses on self-excited induction generators, hybrid renewable systems, intelligent controllers, and smart grid advancements. He is recognized for his contributions to sustainable power system solutions and innovative engineering methodologies. Author Contribution Mr.K. Vinoth Bresnav conceptualized the research idea, developed the system architecture, implemented the ANN-based MPPT algorithm, and performed system-level simulations. He also led the manuscript preparation and revision process.Dr.S. Singaravelu contributed to the design and control of the bi-directional battery management system and the NPC inverter. He assisted with result analysis, verification, and provided critical feedback during manuscript development.Both authors reviewed and approved the final version of the manuscript and agree to be accountable for the content and integrity of the work. Acknowledgement We gratefully acknowledge the support and facilities provided by the authorities of the Annamalai University, Annamalai Nagar, Tamilnadu, India to carry out this research. We would also like to thank our supporting staffs [Department of Electrical Engineering, Annamalai University] who gave insight and knowledge that considerably aided the research. Data Availability All data generated or analysed during this study are included in this published article and its supplementary files. Simulation files (MATLAB/Simulink) are available from the corresponding author on reasonable request. References Bhimaraju, A., Mahesh, A. Recent developments in PV/wind hybrid renewable energy systems: a review. Energy Syst (2024). https://doi.org/10.1007/s12667-024-00679-3. Jingyan Xie, Yun-Ze Li, Lizhu Yang, Yuehang Sun, Man Yuan, A review of the recent progress of stand-alone photovoltaic-battery hybrid energy systems in space and on the ground, Journal of Energy Storage, Volume 55, Part C, 2022, 105735, ISSN 2352-152X, https://doi.org/10.1016/j.est.2022.105735. Sarang, S.A., Raza, M.A., Panhwar, M. et al. Maximizing solar power generation through conventional and digital MPPT techniques: a comparative analysis. Sci Rep 14 , 8944 (2024). https://doi.org/10.1038/s41598-024-59776-z K. Keerthana, S. Singaravelu, "Enhancing the Robustness of P and O Algorithm-Based MPPT Control in Stand-Alone PV Systems through FineTuned PI Controller for Dynamic Load Variations," SSRG International Journal of Electronics and Communication Engineering , vol. 11, no. 6, pp. 9-19, 2024. Crossref, https://doi.org/10.14445/23488549/IJECE-V11I6P102 S. Astaomar and B. Erkal, "Improving the Performance of an Incremental Conductance MPPT Algorithm Using Harris-Hawks Optimization in Photovoltaic Systems," 2024 Global Energy Conference (GEC) , Batman, Turkiye, 2024, pp. 128-134, doi: 10.1109/GEC61857.2024.10881949. M. Farhat, O. Barambones, L. Sbita and J. M. G. de Durana, "A stable FLC-based MPPT technique for photovoltaic system," 2015 IEEE International Conference on Industrial Technology (ICIT) , Seville, Spain, 2015, pp. 890-895, doi: 10.1109/ICIT.2015.7125210. Farhat, M., Flah, A., &Sbita, L. (2014). Photovoltaic maximum power point tracking based on ANN control. International Review on Modelling and Simulations, 7(3), 474-480. Y. Gao, S. Wang, T. Dragicevic, P. Wheeler and P. Zanchetta, "Artificial Intelligence Techniques for Enhancing the Performance of Controllers in Power Converter-Based Systems—An Overview," in IEEE Open Journal of Industry Applications , vol. 4, pp. 366-375, 2023, doi: 10.1109/OJIA.2023.3338534. Y. Triki, A. Triki, A. Bechouche, D. O. Abdeslam and R. Porumb, "An Efficient Battery-Charging Algorithm with ANN based MPPT Method for Off-Grid PV Systems," 2022 IEEE 21st international Ccnference on Sciences and Techniques of Automatic Control and Computer Engineering (STA) , Sousse, Tunisia, 2022, pp. 106-111, doi: 10.1109/STA56120.2022.10019093. Katche, M. L., Makokha, A. B., Zachary, S. O., &Adaramola, M. S. (2023). A Comprehensive Review of Maximum Power Point Tracking (MPPT) Techniques Used in Solar PV Systems. Energies, 16(5), 2206. https://doi.org/10.3390/en16052206 Tamer Khatib, Ibrahim A. Ibrahim, Azah Mohamed,A review on sizing methodologies of photovoltaic array and storage battery in a standalone photovoltaic system,Energy Conversion and Management,Volume 120,2016, Pages 430-448, ISSN 0196-8904, https://doi.org/10.1016/j.enconman.2016.05.011. Alok Jain, Suman Bhullar,Design and performance analysis of solar PV-battery energy storage system integration with three-phase grid,Journal of Power Sources,Volume 640,2025,236486, ISSN 0378-7753, https://doi.org/10.1016/j.jpowsour.2025.236486 Kirubakaran, V., & Singaravelu, S. (2024). A generalized mathematical modeling and performance analysis of GaAs, m-Si and Ge single - junction solar PV cells. Multidisciplinary Science Journal, 6(9), 2024176. https://doi.org/10.31893/multiscience.2024176 J. Edler and N. Kondrath, "Bidirectional Interleaved Buck/Boost DC-DC Converter Design to Improve Power Density in High-Current Applications," 2019 IEEE 62nd International Midwest Symposium on Circuits and Systems (MWSCAS) , Dallas, TX, USA, 2019, pp. 403-406, doi: 10.1109/MWSCAS.2019.8885244. N. D. Bhat., D. B. Kanse., S. D. Patil., and S. D. Pawar., 2020, "DC/DC Buck Converter Using Fuzzy Logic Controller," 5th International Conference on Communication and Electronics Systems (ICCES), Coimbatore, India., pp. 182-187. Yavuz Bahadır KOCA, Yılmaz ASLAN, Ahmet YÖNETKEN3, Yüksel OĞUZ4, “Boost Converter Design and Analysis for Photovoltaic Systems”, Conference Paper · April 2019. H. Patel and V. Agarwal, “MATLAB-Based Modeling to Study the Effects of Partial Shading on PV Array Characteristics,” IEEE Transactions on Energy Conversion, vol. 23, no. 1, pp. 302–310, March 2008. DOI: 10.1109/TEC.2007.914308 S. Jain and V. Agarwal, “Comparison of the Performance of Maximum Power Point Tracking Techniques for a PV System,” Solar Energy Materials and Solar Cells, vol. 90, no. 5, pp. 672–685, 2006. DOI: 10.1016/j.solmat.2005.04.007 M. A. Elobaid, S. Z. Kassas, and M. F. El-Naggar, “Artificial Neural Network-Based Photovoltaic Maximum Power Point Tracking Techniques: A Survey,” Solar Energy, vol. 141, pp. 22–45, 2017. DOI: 10.1016/j.solener.2016.11.018 N. Femia, G. Petrone, G. Spagnuolo, and M. Vitelli, “Optimization of Perturb and Observe Maximum Power Point Tracking Method,” IEEE Transactions on Power Electronics, vol. 20, no. 4, pp. 963–973, July 2005. DOI: 10.1109/TPEL.2005.850975 T. Esram and P. L. Chapman, “Comparison of Photovoltaic Array Maximum Power Point Tracking Techniques,” IEEE Transactions on Energy Conversion, vol. 22, no. 2, pp. 439–449, June 2007. S. Mirbagheri, A. Dorri, and H. R. Najafi, “Modeling and Simulation of Photovoltaic Cells/Modules/Arrays with MATLAB/Simulink,” International Journal of Computer and Electrical Engineering, vol. 3, no. 5, 2011. S. Jain and V. Agarwal, “Comparison of the performance of maximum power point tracking techniques for a PV system,” Solar Energy Materials and Solar Cells, vol. 90, no. 5, pp. 672–685, May 2006. M. A. Elobaid, S. Z. Kassas, and M. F. El-Naggar, “Artificial neural network-based photovoltaic maximum power point tracking techniques: A survey,” Solar Energy, vol. 141, pp. 22–45, Jan. 2017. doi: 10.1016/j.solener.2016.11.018 N. Femia, G. Petrone, G. Spagnuolo, and M. Vitelli, “Optimization of perturb and observe maximum power point tracking method,” IEEE Trans. Power Electron., vol. 20, no. 4, pp. 963–973, Jul. 2005. T. Esram and P. L. Chapman, “Comparison of photovoltaic array maximum power point tracking techniques,” IEEE Trans. Energy Convers., vol. 22, no. 2, pp. 439–449, Jun. 2007. M. Nader, A. Dib, and S. Ouchen, “A fuzzy logic controller based MPPT technique for photovoltaic generation system,” Int. J. Smart Grid, vol. 5, no. 3, pp. 144–151, Sep. 2021. P. C., R. Geethamani, G. Radhakrishnan, S. Kishore Kumar, and C. Manoj, “Comparison of solar P&O and FLC-based MPPT controllers & analysis under dynamic conditions,” EAI Endorsed Transactions on Energy Web, vol. 10, no. 6, e3, 2023. C. R. Algarín, J. T. Giraldo, and O. R. Álvarez, “Data from a photovoltaic system using fuzzy logic and the P&O algorithm under sudden changes in solar irradiance,” Data in Brief, vol. 21, pp. 190–195, Aug. 2018. M. S. Ali, A. A. Elserougi, and A. M. Massoud, “A comprehensive comparison of different MPPT techniques for photovoltaic systems,” IEEE Int. Symp. Power Electron. Distrib. Gener. Syst., 2014. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6831075","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":497238954,"identity":"6f22cb4e-599e-4c95-9529-dfe097419603","order_by":0,"name":"K . Vinoth Bresnav","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA10lEQVRIiWNgGAWjYBACCQYeKIu9AUgYWJCihecASIsEKVokEiB8gkCy/eyxDx/+3JE3uPn86oYfBRIM/O3dCXi1SPPkJc+c2fbMcMPtnLKbPUCHSZw5uwGvFjmGHGNm3obDjDNn56Td4AFqMZDIJaCF/40x858/h+1nzjyTdvMPMVqkJYC2MLAdTuyXYD92myhbJGe8S2bsbTuc3M+Tw3ZbxkCCh6BfJM7nHmb48eewbRv78Wc33/yxkeNv78WvBQnwGIBJYpWDAPsDUlSPglEwCkbBCAIArLhG95q9zCMAAAAASUVORK5CYII=","orcid":"","institution":"Annamalai University","correspondingAuthor":true,"prefix":"","firstName":"K","middleName":". Vinoth","lastName":"Bresnav","suffix":""},{"id":497238955,"identity":"e27d24bf-976b-4ae5-bb5b-0442ea46a890","order_by":1,"name":"Dr.S. Singaravelu","email":"","orcid":"","institution":"Annamalai University","correspondingAuthor":false,"prefix":"Dr.","firstName":"S.","middleName":"","lastName":"Singaravelu","suffix":""}],"badges":[],"createdAt":"2025-06-05 16:53:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6831075/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6831075/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":88641544,"identity":"98063ef3-b39d-4341-8779-2c5b4b14cd84","added_by":"auto","created_at":"2025-08-08 16:08:47","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":101668,"visible":true,"origin":"","legend":"\u003cp\u003eSystem Architecture Diagram\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/9d330b22868ed6f12054e633.png"},{"id":88641553,"identity":"7f967503-2df6-45c2-9860-eb6714e55086","added_by":"auto","created_at":"2025-08-08 16:08:48","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":115984,"visible":true,"origin":"","legend":"\u003cp\u003eEquivalent circuit model of a solar cell [13].\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/4b12f37843832f2b48a7deeb.png"},{"id":88643599,"identity":"688ec6a9-3391-4f25-8d15-d013e03e30cf","added_by":"auto","created_at":"2025-08-08 16:16:49","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":92798,"visible":true,"origin":"","legend":"\u003cp\u003eGraphical representations of the V–I and P–V characteristics of the selected PV module under varying irradiance levels at a constant temperature of 25 °C.\u003c/p\u003e","description":"","filename":"image3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/f9b660f6eb28e91822cc4898.jpeg"},{"id":88641537,"identity":"fdc3a343-eb08-4d4a-aeb2-4e87fdfb09d2","added_by":"auto","created_at":"2025-08-08 16:08:47","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":75747,"visible":true,"origin":"","legend":"\u003cp\u003eGraphical illustrations of the V–I and P–Vcharacteristics of the considered PV module under varying temperature conditions at a constant irradiance of 1000 W/m².\u003c/p\u003e","description":"","filename":"image4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/1c3f74b68054d1a7a679a327.jpeg"},{"id":88641549,"identity":"10173007-1304-4130-8673-86235103eb9f","added_by":"auto","created_at":"2025-08-08 16:08:47","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":17723,"visible":true,"origin":"","legend":"\u003cp\u003eANN Structure\u003c/p\u003e","description":"","filename":"image5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/f223a96b403232dbc69c6bea.jpeg"},{"id":88641534,"identity":"c37f29aa-a581-4f77-b0a6-7b04e8c6330b","added_by":"auto","created_at":"2025-08-08 16:08:46","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":61945,"visible":true,"origin":"","legend":"\u003cp\u003eError Histogram\u003c/p\u003e","description":"","filename":"image6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/e0741560c724cf8d9cc0e05b.jpeg"},{"id":88643594,"identity":"66955a94-9d37-4548-bb44-8ee2f4088a85","added_by":"auto","created_at":"2025-08-08 16:16:48","extension":"jpeg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":76604,"visible":true,"origin":"","legend":"\u003cp\u003eTraining, validation and test data\u003c/p\u003e","description":"","filename":"image7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/c9477e6a9e17aae0ef4ecaaf.jpeg"},{"id":88641536,"identity":"5d180438-1291-404a-ab3a-bc96d74238f5","added_by":"auto","created_at":"2025-08-08 16:08:46","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":81668,"visible":true,"origin":"","legend":"\u003cp\u003eSIMULINK model of ANN\u003c/p\u003e","description":"","filename":"image8.png","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/8866eb43e561f7331cb30532.png"},{"id":88641576,"identity":"311c8c78-1893-4b8d-91ee-4f657fc7cf03","added_by":"auto","created_at":"2025-08-08 16:08:49","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":187784,"visible":true,"origin":"","legend":"\u003cp\u003eSimulation model of NPC inverter circuit\u003c/p\u003e","description":"","filename":"image9.png","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/c754e883e3b3a88aeec2fa51.png"},{"id":88641560,"identity":"54469101-061f-43cd-a849-1c57a4f03fe4","added_by":"auto","created_at":"2025-08-08 16:08:48","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":283592,"visible":true,"origin":"","legend":"\u003cp\u003eSimulation model of NPC inverter control circuit\u003c/p\u003e","description":"","filename":"image10.png","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/52d4802553d8008ab47e5633.png"},{"id":88643572,"identity":"e9962ff2-c04f-4716-ad63-827bd0afc48d","added_by":"auto","created_at":"2025-08-08 16:16:47","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":232105,"visible":true,"origin":"","legend":"\u003cp\u003eSimulation model of overall work\u003c/p\u003e","description":"","filename":"image11.png","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/b87552b817cd1bc18b29c4b4.png"},{"id":88641554,"identity":"4b3c5e07-d1a2-44c4-83be-aeb0e5e56c77","added_by":"auto","created_at":"2025-08-08 16:08:48","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":105882,"visible":true,"origin":"","legend":"\u003cp\u003eSimulation model of battery control circuit\u003c/p\u003e","description":"","filename":"image12.png","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/a5bca513283805e0982e6079.png"},{"id":88641529,"identity":"e7606311-7003-4883-abf4-599b4fb126ec","added_by":"auto","created_at":"2025-08-08 16:08:46","extension":"jpeg","order_by":13,"title":"Figure 13","display":"","copyAsset":false,"role":"figure","size":78193,"visible":true,"origin":"","legend":"\u003cp\u003ePerformance of PV system with ANN MPPT model\u003c/p\u003e","description":"","filename":"image13.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/ee29df7beed71301bd337a7c.jpeg"},{"id":88641545,"identity":"3e7d4c56-07ff-4722-ae41-d6f39412654a","added_by":"auto","created_at":"2025-08-08 16:08:47","extension":"jpeg","order_by":14,"title":"Figure 14","display":"","copyAsset":false,"role":"figure","size":66258,"visible":true,"origin":"","legend":"\u003cp\u003eBattery charging and discharging behavior for change in load\u003c/p\u003e","description":"","filename":"image14.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/e9d1c190cce6c1ea3227308c.jpeg"},{"id":88641566,"identity":"21050645-d14e-418f-8e90-9683bdf6a57c","added_by":"auto","created_at":"2025-08-08 16:08:49","extension":"jpeg","order_by":15,"title":"Figure 15","display":"","copyAsset":false,"role":"figure","size":231819,"visible":true,"origin":"","legend":"\u003cp\u003eThree-phase load voltage characteristics during system operation\u003c/p\u003e","description":"","filename":"image15.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/2f1d2db2edfe02fd19551b17.jpeg"},{"id":88641543,"identity":"51b09c85-9480-41d8-9b50-ec32a6736e55","added_by":"auto","created_at":"2025-08-08 16:08:47","extension":"jpeg","order_by":16,"title":"Figure 16","display":"","copyAsset":false,"role":"figure","size":193602,"visible":true,"origin":"","legend":"\u003cp\u003eCurrent profile of three-phase load under transient conditions\u003c/p\u003e","description":"","filename":"image16.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/3e0e80192d6db52722141bc4.jpeg"},{"id":88641574,"identity":"e9aa2ec2-d305-4353-9ce7-458a3cac4ec8","added_by":"auto","created_at":"2025-08-08 16:08:49","extension":"jpeg","order_by":17,"title":"Figure 17","display":"","copyAsset":false,"role":"figure","size":57553,"visible":true,"origin":"","legend":"\u003cp\u003eCurrent comparison of pv generation and load demand over time\u003c/p\u003e","description":"","filename":"image17.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/7f1993afb79bcb8d4779c2e6.jpeg"},{"id":100529826,"identity":"23740866-ec60-4d29-8628-15263eae74b7","added_by":"auto","created_at":"2026-01-19 01:08:48","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2901934,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6831075/v1/9bd26df4-7d53-4860-b26e-a63abec4112d.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"ANN-Based MPPT Control of a Standalone PV System with Bi-Directional Battery Management and NPC Inverter for Dynamic 3-Phase AC Loads","fulltext":[{"header":"1. INTRODUCTION","content":"\u003cp\u003eElectrical energy is fundamental not only to the daily lives of individuals but also plays a pivotal role in the economic development of nations. In light of global concerns regarding environmental sustainability and energy security, renewable energy sources have emerged as a promising alternative to conventional fossil-based systems. However, the inherently intermittent and weather-dependent nature of sources such as solar and wind limits their standalone reliability [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e].To address this challenge, Hybrid Renewable Energy Systems (HRES), which combine multiple energy sources, have gained significant attention.As the deployment of renewable energy systems increases, efficient control and sizing of components become critical for ensuring system stability and economic feasibility. Modern HRES designs also incorporate demand-side strategies such as demand response and even electric vehicles (EVs) as flexible storage units or controllable loads, further expanding their potential. In this context, intelligent energy management systems that include accurate MPPT, optimal battery utilization and high-quality power conversion are essential. The standalone PV systems are often integrated with energy storage components such as secondary batteries, forming PV/B systems [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. These configurations not only provide backup during periods of low irradiance but also enhance the system\u0026rsquo;s ability to meet dynamic load demands. Such systems are particularly critical in remote and off-grid areas where grid extension is not viable. Furthermore, applications extend beyond terrestrial uses into space missions, where PV/B systems act as the sole energy source for satellites and spacecraft.\u003c/p\u003e\u003cp\u003eIn the search to mitigate climate change and reduce carbon emissions, renewable energy systems particularly PV systemshave garnered substantial global attention. Among the core technical challenges in PV system deployment is the extraction of maximum power under varying environmental and dynamic load conditions. MPPT algorithms play a critical role in enhancing energy harvest by dynamically adjusting the operating point of the PV array to match optimal conditions [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].Over the years, a wide array of MPPT techniques has been developed, ranging from conventional methods such as Incremental Conductance (INC) [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e] and Perturb \u0026amp; Observe (P\u0026amp;O) [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e], to more advanced hybrid and artificial intelligence (AI) [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e] based controllers including Fuzzy Logic [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e] and Artificial Neural Networks (ANN).The existing comparative studies indicate that while traditional methods like INC and P\u0026amp;O offer simplicity and reasonable performance, they often fall short in complex, rapidly changing environments or partial shading conditions. On the other hand, AI-based controllers demonstrate superior adaptability, accuracy and robustness [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e].AI-based approaches such as Fuzzy-PSO, ANN, ANFIS, and their combinations exhibit competitive or higher accuracy, often exceeding 98% [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. These findings underscore the growing potential of intelligent control systems in optimizing PV output and improving energy management in standalone or hybrid configurations.\u003c/p\u003e\u003cp\u003eStandalone PV systems are increasingly vital in remote and off-grid applications, but face challenges related to fluctuating solar irradiance and battery reliability [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Conventional charging techniques and basic MPPT algorithms often underperform in dynamic environments [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Recent studies have proposed adaptive neural control strategies using structures like ADALINE, achieving high MPPT efficiencies [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e].\u003c/p\u003e\u003cp\u003e\u003cb\u003eProblem Statement\u003c/b\u003e:\u003c/p\u003e\u003cp\u003eStandalone photovoltaic systems face critical challenges in maintaining power continuity and voltage stability, particularly under fluctuating environmental conditions and variable load profiles. While traditional MPPT methods provide a basic level of power optimization, they often struggle to adapt in real time to sudden irradiance changes or partial shading, resulting in suboptimal energy capture. Similarly, conventional battery management strategies are limited in dynamically adjusting charging and discharging operations, leading to inefficient energy utilization and potential battery degradation. Moreover, the integration of advanced inverter topologies such asNPC inverters is seldom optimized in a unified control framework. A comprehensive, intelligent solution is needed to simultaneously optimize MPPT performance, battery behavior and AC power quality for reliable standalone operation.\u003c/p\u003e\u003cp\u003eCurrent literature tends to focus separately on MPPT optimization or battery management, lacking an integrated control framework that coordinates both functions effectively within standalone PV systems. While ANNbased MPPT controllers have shown superior tracking efficiency, their integration with real-time battery control through a bi-directional converter is understated, especially in the context of supplyingthree-phase AC loads. Furthermore, limited research has been conducted on harmonizing ANN-based MPPT, dynamic battery interface control and three-level NPC inverters into a single system which is particularly in terms of ensuring consistent AC power quality and stable DC-link voltage across load and generation scenarios [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eInspired by recent advancements in intelligent energy management, this study implements a robust ANNbased MPPT control strategy integrated with a bi-directional DC-DC converter for effective battery management and a NPC inverter for three-phase AC load supply in standalone PV systems. This dual-layer intelligence harnesses the predictive and adaptive strengths of ANN to enhance power tracking efficiency, optimize battery utilization, and ensure stable, high-quality power delivery under dynamic environmental conditions.\u003c/p\u003e\u003cp\u003e\u003cb\u003eObjectives of the Study\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eTo develop a predictive and adaptive ANN-based MPPT controller to improve energy harvesting efficiency in standalone PV systems under rapidly changing irradiance conditions.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eTo design a PI-controlled bi-directional DC-DC converter for managing battery charging/discharging based on PV output and AC load demand, ensuring optimal battery usage.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eTo integrate a three-level NPC inverter with the DC-link and assess its ability to maintain high-quality three-phase AC power delivery.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eTo evaluate the combined system's performance through simulation, focusing on MPPT accuracy, battery SOC dynamics DC-link voltage regulation and AC voltage waveform quality.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eTo validate the proposed system\u0026rsquo;s capability in maintaining stable, uninterrupted power supply under practical and dynamic operating scenarios, including variable solar input and load profiles.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eThis work is organized as follows: Section 1 presents the introduction, problem statement, research gap and objective of the work; Section 2 describes the system description which includes overview of proposed system architecture, including PV system description, the ANN-based MPPT controller, designof bi-directional converter with Battery configurations and NPC inverter; Section 3 discusses the modeling and simulation of the integrated system in MATLAB/Simulink; Section 4 presents the results and performance analysis under various environmental and dynamic load conditions. Finally, Section 5 concludes the study with key findings, contributions and suggestions for future research.\u003c/p\u003e"},{"header":"2. System Description","content":"\u003cp\u003e\u003cb\u003eA) Overview of the proposed system architecture\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe proposed standalone PV system architecture integrates intelligent control strategies to optimize energy harvesting, storage management, and AC power delivery. At the core of the system is an ANNbasedMPPT controller, which dynamically adjusts to varying environmental conditions to maximize power extraction from the solar PV array. This power is regulated via a DC-DC boost converter and stabilized across a DC-link capacitor. A PI-controlled bi-directional DC-DC converter interfaces a battery energy storage unit, enabling smart charge/discharge operations based on PV power flow and load demand. To ensure high-quality AC output, a three-level NPC inverter is employed which offering reduced harmonic distortion and improved efficiency. The coordination of ANN-based MPPT, battery management and NPC inverter operation forms a unified control scheme aimed at delivering consistent, high-quality three-phase power during fluctuating solar irradiance and load conditions as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cb\u003eB) Description of PV system Parameters\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe inclusion of a 2000 kW PV system in this research plays a critical role in demonstrating the practicality and scalability of the proposed ANN-based MPPT and bi-directional battery management strategy for large-scale standalone applications.The following section provides a detailed architecture of the 2000KW PV system, which is particularly chosen for integrating this system with a high current network.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eIn addition to the commonly used models, other configurations such as single-diode and two-diode models are also available. The two-diode model introduces a second diode connected in parallel with the first, effectively mimicking the behavior of a single-diode equivalent circuit. While the single-diode model is simpler and involves fewer parameters, the two-diode model provides a more detailed representation of PV cell behavior. Studies have demonstrated that both simulated and experimental I\u0026ndash;V and P\u0026ndash;V characteristics of solar panels yield closely matching results. In modeling, a PV cell is typically depicted as an electrical circuit comprising a diode in parallel with a current source that represents the photo-generated current. These individual cells are then grouped into PV modules, configured in both series and parallel to deliver the required power output. An ideal PV cell can be modeled as a DC source in parallel with a diode [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e].To represent real-world losses and leakages, series and shunt resistances are incorporated into the ideal model, forming a more accurate practical equivalent. Figure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e illustrates the structure of such a PV cell model.\u003c/p\u003e\u003cp\u003eThe electrical behavior of a practical photovoltaic cell can be expressed through a modified current-voltage relationship, incorporating the effects of internal resistances and non-idealities. The output current (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{I}_{pv})\\)\u003c/span\u003e\u003c/span\u003efrom the PV cell is determined using the following expression:\u003c/p\u003e\u003cp\u003eI\u0026thinsp;=\u0026thinsp;I\u003csub\u003epv\u003c/sub\u003e = I\u003csub\u003eph\u003c/sub\u003e \u0026ndash; I\u003csub\u003es\u003c/sub\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:[{e}^{\\left(\\frac{\\left({V}_{pv}+{I}_{pv}.{R}_{S}\\right)}{{N}_{s}{V}_{T}}\\right)}-1]\\)\u003c/span\u003e\u003c/span\u003e-\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\frac{({V}_{pv}+{I}_{pv}.{R}_{S})}{{R}_{sh}}\\)\u003c/span\u003e\u003c/span\u003e (1)\u003c/p\u003e\u003cp\u003eWhere, I\u003csub\u003eph\u003c/sub\u003e represents the photo-generated current,\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{I}_{S}\\)\u003c/span\u003e\u003c/span\u003eis the reverse saturation current of the diode,\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{V}_{pv}\\)\u003c/span\u003e\u003c/span\u003e​ and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{I}_{pv}\\)\u003c/span\u003e\u003c/span\u003e​ are the output voltage and current of the PV cell,\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}_{S}\\)\u003c/span\u003e\u003c/span\u003e​ and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}_{sh}\\)\u003c/span\u003e\u003c/span\u003e​ are the series and shunt resistances,\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{S}\\)\u003c/span\u003e\u003c/span\u003e​ denotes the number of series-connected cells and\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{V}_{T}\\)\u003c/span\u003e\u003c/span\u003e​ is the thermal voltage which equals \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:a.K.T/q\\)\u003c/span\u003e\u003c/span\u003e. Where a is diode ideality factor, q denotes electron charge, K for Boltzmann constant and T is cell temperature.\u003c/p\u003e\u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents all the parameters of the PV module used in this study, specified under Standard Test Conditions (STC), which correspond to an irradiance of 1000 W/m\u0026sup2; and a temperature of 25\u0026deg;C.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eIntrinsic electrical parameters associated with the PV panel being evaluated\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSL. No.\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003ePARAMETERS\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRange\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNumber of cells per module\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e128\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eOpen circuit voltage (Voc)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e85.3V\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eShort circuit current (Isc)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e6.09A\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eVoltage at MPP (Vmpp)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e72.9V\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCurrent at MPP (Impp)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5.69A\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ePower at MPP (Pmpp)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e414.801W\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eParallel Strings\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e700\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSeries connected Modules per string\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eModule Voltage at MPP (Vmpp)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e510.3V\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eModule Current at MPP (Impp)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3983A\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eModule Power at MPP (Pmpp)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2032KW\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDiode Ideality factor\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.87223\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eFigures \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e illustrate the graphical relationships between voltage and current, and between power and voltage, respectively, under different irradiance and temperature conditions.\u003c/p\u003e\u003cp\u003e\u003cb\u003eC)Model of ANN MPPT method\u003c/b\u003e\u003c/p\u003e\u003cp\u003eA2000KW photovoltaic system using ANN for Maximum Power Point Tracking is a modern and intelligent technique aimed at maximizing energy extraction under varying environmental conditions. In this approach, key PV system parameters such as solar PV voltage and current are used as inputs to the ANN, while the output is the duty cycle, which adjusts the operating point of the DC-DC Link to track the maximum power point efficiently.\u003c/p\u003e\u003cp\u003eTo train the ANN model, historical data comprising PV voltage, current and the corresponding optimal duty cycles is collected. This data enables the ANN to learn the nonlinear relationship between the input conditions and the appropriate duty cycle needed to achieve MPPT. The training process involves optimizing the network\u0026rsquo;s weights and biases to minimize the error between the predicted and actual duty cycles.The ANN-based MPPT controller in this work is designed and implemented using the SIMULINK platform, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe difference between the predicted and actual duty cycle values is evaluated using a loss function, which serves as a measure of prediction error. This function helps quantify how far the model\u0026rsquo;s output deviates from the expected results. To minimize this error, the ANN adjusts its internal weights and biases based on the gradient of the loss function. In this work, the Levenberg-Marquardt optimization algorithm is used for this purpose.The Levenberg-Marquardt algorithm is specifically designed to solve nonlinear least squares problems and combines the advantages of both the gradient descent and Gauss-Newton methods. Named after Kenneth Levenberg and Donald Marquardt, who independently developed the approach, the algorithm works by iteratively updating the model parameters to reduce the sum of squared differences between the actual and predicted values.\u003c/p\u003e\u003cp\u003eA properly trained ANN model typically produces an error histogram with values centered around zero, indicating high prediction accuracy, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e. The division of data into training, validation, and testing sets used for developing the ANN model is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e. A regression value (R) close to 1 signifies a strong correlation between predicted and actual outputs, confirming the effectiveness of the trained model.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe developed ANN model achieved a mean squared error of 1.6450e-10 and a minimum gradient of 9.99e-8, indicating its strong performance and suitability for the intended application. Following the training phase, the ANN is integrated into the closed-loop control system, where it continuously receives real-time inputs of PV current and voltage. Based on these inputs, the ANN predicts the optimal duty cycle required for efficient operation under the given conditions. The SIMULINK implementation of the trained ANN model for this study is depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cb\u003eD) Design of PI controller bi-directional converter with Battery configurations\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe bi-directional DC-DC converter facilitates both charging and discharging of the battery bank based on PV surplus and load demand. An interleaved bidirectional buck-boost topology [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]is used to support high power with reduced ripple and improved thermal distribution. The battery specifications are as shown in Table \u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eBattery Specifications\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"2\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBattery Parameters\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRange\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNominal Voltage(\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{Vbat}_{Nominal}\\)\u003c/span\u003e\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e300V\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCapacity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6500ah\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInitial State of Charge (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{SOC}_{Initial}\\)\u003c/span\u003e\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e25%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eResponse time\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e30sec\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003cp\u003eA buck-boost converter topology is ideal due to its ability to handle both power directions. The converter operates in buck mode to charge the battery when PV generation exceeds load demand and in boost mode to discharge when the load exceeds PV generation.The converter is governed by a PI (Proportional-Integral) controller, regulating the battery current to:Maintain SOC within operational limits, avoid overcharge/over-discharge and ensure smooth transitions between charging and discharging.The controller compares the reference current (based on power mismatch) with the actual battery current and adjusts the duty cycle accordingly. The battery power range \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\mathbf{P}\\text{b}\\text{a}\\text{t}}_{\\text{m}\\text{a}\\text{x}}\\)\u003c/span\u003e\u003c/span\u003ehas been decided by the expression 2,\u003c/p\u003e\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:{\\mathbf{P}\\text{b}\\text{a}\\text{t}}_{\\text{m}\\text{a}\\text{x}}={\\mathbf{V}\\text{b}\\text{a}\\text{t}}_{\\text{N}\\text{o}\\text{m}\\text{i}\\text{n}\\text{a}\\text{l}}\\times\\:{\\mathbf{I}\\text{b}\\text{a}\\text{t}}_{\\text{m}\\text{a}\\text{x}}=975\\text{K}\\text{W}\\:\\text{f}\\text{o}\\text{r}\\:\\text{t}\\text{h}\\text{e}\\:\\text{s}\\text{e}\\text{l}\\text{e}\\text{c}\\text{t}\\text{e}\\text{d}\\:\\text{b}\\text{a}\\text{t}\\text{t}\\text{e}\\text{r}\\text{y}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eFor high power converters, ripple current should be \u0026lt;\u0026thinsp;20% of max current. The control objective is to maintain battery current smoothness (30s dynamic response), preserve SOC profile and prevent oscillations during load transients. The design of dc-dc bi-directional converter as follows,\u003c/p\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\text{L}=\\frac{{\\text{V}}_{\\text{d}\\text{c}}\\times\\:\\text{D}}{{\\text{f}}_{\\text{s}}\\times\\:\\varDelta\\:\\text{I}\\text{b}\\text{a}\\text{t}\\text{t}}\\)\u003c/span\u003e\u003c/span\u003e = 47\u0026micro;H (3)\u003c/p\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\text{C}=\\frac{{I}_{Bat}^{max}}{{\\text{f}}_{\\text{s}}\\times\\:\\varDelta\\:\\text{V}\\text{b}\\text{a}\\text{t}\\text{t}}\\)\u003c/span\u003e\u003c/span\u003e = 33mF (4)\u003c/p\u003e\u003cp\u003eWhere switching frequency (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{f}}_{\\text{s}})\\)\u003c/span\u003e\u003c/span\u003e is 10KHZ, maximum battery current (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{I}_{Bat}^{max}\\)\u003c/span\u003e\u003c/span\u003e) is 3250A, allowable ripple current (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varDelta\\:\\text{I}\\text{b}\\text{a}\\text{t}\\text{t}\\)\u003c/span\u003e\u003c/span\u003e) is 650A, dc link voltage is 600V.The transfer function of converter with PI controller for battery current control (current mode control) is given as,\u003c/p\u003e\u003cp\u003eG(s) = \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\frac{Vbat}{L.s}\\)\u003c/span\u003e\u003c/span\u003e (5)\u003c/p\u003e\u003cp\u003ePI controller transfer function is given as,\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:{G}_{c}\\left(s\\right)={K}_{p}+\\frac{{K}_{i}}{s}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e6\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eWhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{K}_{p}\\)\u003c/span\u003e\u003c/span\u003e is 0.05 and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{K}_{i}\\)\u003c/span\u003e\u003c/span\u003e is 0.01 as derived form automatic tuning in Matlab platform for 30 seconds settling time.\u003c/p\u003e\u003cp\u003e\u003cb\u003eE) Design of Low-voltage high-current 3-level NPC inverter\u003c/b\u003e\u003c/p\u003e\u003cp\u003eMultilevel inverters have emerged as a powerful solution for medium- to high-power applications, especially in renewable energy systems. Among them, the Three-LevelNPC inverter offers several advantages such as, reduced voltage stress on switching devices, improved harmonic performance and higher efficiency due to lower switching losses.In an NPC topology, each leg of the inverter includes four switches and two clamping diodes, enabling three distinct output voltage levels: +Vdc​/2, 0 and \u0026minus;\u0026thinsp;Vdc​/2. This finer resolution results in a closer approximation to a sinusoidal waveform, thus reducing output Total Harmonic Distortion (THD) and filter requirements.\u003c/p\u003e\u003cp\u003eThe three-phase 1500 kW load is evenly distributed which is resulting in a phase power of 500 kW. Given an output line-to-line RMS voltage of 400 V, the phase voltage calculates to approximately 230.9 V. Assuming a power factor of 0.9, the RMS phase current is approximately 2402 A which is requiring the use of multiple high-current power switches in parallel within each inverter leg. To accommodate the 550 V DC-link voltage with adequate margin, each switch should have a voltage rating of at least 900 V and a current rating of at least 400 A.Therefore, six such devices per switch are implemented. The DC-link capacitor bank is designed to store 10 kJ of energy, requiring a total capacitance of around 66,000 \u0026micro;F which is implemented using low-ESR, high-ripple film capacitors. To limit output current ripple, the inverter uses an output filter inductor of approximately 31.8 \u0026micro;H which is capable of handling over 2500 A. For modulation and control, a hybrid approach combining Space Vector PWM and level-shifted PWM is employed to ensure efficient DC-link usage and neutral-point voltage balance, while an ANN-based current controller enhances dynamic response and nonlinear load compensation. The design-oriented specifications are as listed in Table \u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eNPC Inverter Specifications\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"2\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eParameters\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRange\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLoad Power\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1500 kW\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOutput Voltage (L-L RMS)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e400 V (3-phase standard)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOutput Frequency\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e50 Hz\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDC-Link Voltage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e550 V\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTopology\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3-Level NPC\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSwitching Frequency\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5\u0026ndash;10 kHz\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVoltage Regulation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e2%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eResponse time\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;1 ms\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInverter Efficiency\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;96%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOutput THD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;2%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"3. Simulink Model of Integrated system","content":"\u003cp\u003eThe Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e illustrates the NPC Inverter Circuit, which features a three-level NPC topology powered by a 550 V DC-link. The inverter converts the regulated DC voltage into a three-phase AC output, ensuring high-quality voltage delivery to the load. A harmonic filter and a Y-grounded transformer are employed to minimize output harmonics and provide electric isolation. Each phase leg of the NPC inverter is configured to handle high current by paralleling multiple power switches, and appropriate filtering is added to maintain grid-compatible output.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e shows the Inverter Control Circuit, which employs an ANN-based MPPT Controller to ensure optimal power extraction from the PV array. The controller dynamically adjusts the reference voltage (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{U}}_{\\text{r}\\text{e}\\text{f}}\\)\u003c/span\u003e\u003c/span\u003e) for the inverter based on real-time irradiance and PV current/voltage inputs. A phase-locked loop (PLL) enables synchronization with the reference frame and a PI-regulated \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{V}}_{\\text{D}\\text{C}}\\)\u003c/span\u003e\u003c/span\u003e controller maintains DC-link voltage stability by generating the appropriate d-axis current reference. This is followed by a current regulator that translates these references into modulation signals for the inverter switches. The control system integrates both the MPPT and inverter regulation schemes to ensure seamless operation under dynamic solar and load conditions.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe integrated system includes a solar PV array connected with a battery, which is interfaced to the AC load via a 3-level NPC inverter and a harmonic filter, ensuring sinusoidal current injection into the grid as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003e. The control architecture further incorporates a Phase-Locked Loop (PLL) for synchronization, a voltage and current regulator and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{U}}_{\\text{a}\\text{b}\\text{c}}\\)\u003c/span\u003e\u003c/span\u003e reference generation for stable operation.The battery energy storage system (BESS) is interfaced through a bi-directional DC-DC converter controlled by a PI-based feedback control loop. This controller adjusts the battery's charge/discharge power in real time, considering the instantaneous PV current and load demand. The Battery Control Block, as highlighted in the second schematic, regulates the charging process using a logic that compares a fraction of PV current with the battery current and adjusts the usage signal accordingly to ensure effective energy balancing and SOC (state of charge) management as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003e.\u003c/p\u003e\u003cp\u003eThe coordinated interaction of these subsystems enables efficient power flow regulation, voltage stability and system reliability across different load scenarios. The subsequent section presents detailed simulation results and performance analyses under various dynamic operating conditions which demonstrating the effectiveness of the proposed ANN-based MPPT control and energy management strategies.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe integrated Simulink model effectively demonstrates the coordinated operation of the solar PV system, battery storage, NPC inverter and ANN-based MPPT control. This system ensures the optimal power management and system stability under three phase load and variable environmental conditions.\u003c/p\u003e"},{"header":"4. Results and Performance Evaluation under Dynamic Environmental and Load Conditions","content":"\u003cp\u003eThis section presents a comprehensive analysis of the PV/battery/NPC inverter system under dynamically varying environmental and load scenarios. The simulation is designed to reflect realistic operating conditions by incorporating sudden changes in load demand and assessing the system's response. The evaluation focuses on four key aspects: the accuracy of PV voltage and power tracking using the P\u0026amp;O MPPT algorithm, the battery's charging/discharging behavior governed by a PI controller, the quality of AC load voltage includingTHD and the overall system stability and energy flow management. These performance metrics are critical for ensuring reliable and efficient energy delivery in standalone or microgrid applications.\u003c/p\u003e\u003cp\u003e\u003cb\u003ei) PV power and voltage tracking accuracy\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe implemented ANN MPPT algorithm effectively tracks the Maximum Power Point of the PV array. The PV voltage oscillates tightly around the Vmpp (510.3 V), and power stabilizes near the rated 2032 kW under standard test conditions, with minimal deviations, indicating high tracking accuracy as observed from Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e13\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cb\u003eii) Battery charging/discharging behavior\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003e0\u0026ndash;4.5 s\u003c/b\u003e: The PV output exceeds the load demand (1500 kW), and the battery charges with the excess 500 kW.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003e4.5 s onwards\u003c/b\u003e: When load increases to 2500 kW, the PV is insufficient and the battery discharges to support the remaining demand (500 kW).\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eSOC profile confirms this behavior with a smooth charging tendency initially which is followed by controlled discharging after the load step as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig14\" class=\"InternalRef\"\u003e14\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cb\u003eiii) AC load voltage waveform and THD analysis\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe three-phase output voltage maintains a stable 400 V (rms) with proper 120\u0026deg; phase shift.FFT analysis reveals a THD (\u0026lt;\u0026thinsp;2%) which proves the NPC inverter\u0026rsquo;s high-quality output and filtering effectiveness.Waveform plots before and after the load step show no significant distortion, indicating strong voltage regulation as visualized in Fig.\u0026nbsp;\u003cspan refid=\"Fig15\" class=\"InternalRef\"\u003e15\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cb\u003eiv) Insights into system stability and energy flow\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe system remains stable under both static and dynamic conditions, with continuous power flow management as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig16\" class=\"InternalRef\"\u003e16\u003c/span\u003e.Duty cycle modulation through the PI controller ensures smooth current transitions.Energy balance is maintained throughout the simulation, with the battery acting as an efficient buffer during load fluctuation.No voltage collapse or instability is observed, confirming the robust design and control logic as proved with Fig.\u0026nbsp;\u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e17\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eOverall, the energy management strategy effectively coordinates power among PV, battery and load to maintain system reliability.The unique summary of this work has been elaborated using Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, which captures key technical observations of PV voltage and power tracking, battery dynamics, AC load performance and system stability under variable load and environmental conditions.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eResults and Performance Metrics under Dynamic Environmental and Load Conditions\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMetric\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDescription\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eObservation\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePV Voltage Tracking (Vmpp)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMPPT voltage convergence accuracy\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e~\u0026thinsp;510.3 V with minor ripple (\u0026lt;\u0026thinsp;1%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePV Power Output (Pmpp)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMaximum power generation capability\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e~\u0026thinsp;2032 kW sustained under STC\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBattery Operation Mode (0\u0026ndash;4.5s)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCharging phase (excess PV power utilized)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eBattery charges at ~\u0026thinsp;500 kW\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBattery Operation Mode (4.5\u0026ndash;14s)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDischarging phase (PV power deficit)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eBattery discharges\u0026thinsp;~\u0026thinsp;500 kW to support extra load\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSOC Behavior\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eResponse to charging/discharging transitions\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSmooth variation, with logical increase/decrease\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAC Load Voltage (RMS)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ePer-phase voltage magnitude maintained\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e400 V (\u0026plusmn;\u0026thinsp;small ripple)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePhase Shift in Load Voltage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eVerification of 3-phase sinusoidal quality\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e120\u0026deg; phase shift among phases\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTHD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMeasured using FFT analysis\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;2%\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLoad Step Change at 4.5s\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAdditional 1000 kW applied\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eTotal load rises from 1500 kW \u0026rarr; 2500 kW\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSystem Stability\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eUnder dynamic PV/load conditions\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eStable operation, no overshoot or voltage dip\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEnergy FlowManagement\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBalance between PV, battery, and load\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eDynamic and accurate with PI controller support\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDuty Cycle Response\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ePI controller modulation of DC-DC converter\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSmooth variation without oscillations\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eThis study presents a robust and adaptive energy management system for a standalone PV/battery-powered three-phase load configuration which is validated through a 14-second MATLAB simulation incorporating realistic irradiance fluctuations and dynamic load variations. The system employed AI based MPPT tracking and a PI-controlled bidirectional DC-DC converter, ensuring reliable power flow continuity even during sudden demand surges and irradiance dips. Key findings include high PV power extraction efficiency (2032 kW), rapid and stable battery mode switching (charging/discharging) and THD levels below 2% by maintaining a steady 400 V AC output with accurate 120\u0026deg; phase shifts.The synchronized operation between the PV source, battery and NPC inverter under dynamic load transitions demonstrates the system\u0026rsquo;s practical viability and resilience.\u003c/p\u003e\u003cp\u003eThis work uniquely highlights the integration of AI control strategies with traditional components to form a scalable framework for renewable-powered standalone systems. Future research could explore Cybersecurity-aware control algorithms, especially for IoT-enabled PV-battery microgrids to safeguard communication and operational integrity.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAUTHOR BIOGRAPHIES\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMr. K. Vinoth Bresnav M.E.,\u0026nbsp;an Assistant Professor Department of Electrical Engineering, with over 13 years of academic and industrial experience. Currently pursuing his Ph.D. at Annamalai University, he specializes in Power Systems, Control Systems, and Electrical Machines. He has authored books and published research in reputed journals and conferences. His contributions include academic coordination, placement, and NBA documentation. He is also GATE qualified and actively engaged in FDPs, workshops, and NPTEL-certified courses.\u003c/p\u003e\n\u003cp\u003eDr. S. Singaravelu, Professor in the Department of Electrical Engineering, Annamalai University, Annamalai Nagar, Tamilnadu, \u0026nbsp;holds a Ph.D. with expertise in Electrical Machines, Power Systems, Renewable Energy, HVDC, and Smart Grid technologies. With over three decades of academic and research experience, he has published extensively in reputed national and international journals. His research focuses on self-excited induction generators, hybrid renewable systems, intelligent controllers, and smart grid advancements. He is recognized for his contributions to sustainable power system solutions and innovative engineering methodologies.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eMr.K. Vinoth Bresnav conceptualized the research idea, developed the system architecture, implemented the ANN-based MPPT algorithm, and performed system-level simulations. He also led the manuscript preparation and revision process.Dr.S. Singaravelu contributed to the design and control of the bi-directional battery management system and the NPC inverter. He assisted with result analysis, verification, and provided critical feedback during manuscript development.Both authors reviewed and approved the final version of the manuscript and agree to be accountable for the content and integrity of the work.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe gratefully acknowledge the support and facilities provided by the authorities of the Annamalai University, Annamalai Nagar, Tamilnadu, India to carry out this research. We would also like to thank our supporting staffs [Department of Electrical Engineering, Annamalai University] who gave insight and knowledge that considerably aided the research.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eAll data generated or analysed during this study are included in this published article and its supplementary files. Simulation files (MATLAB/Simulink) are available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBhimaraju, A., Mahesh, A. Recent developments in PV/wind hybrid renewable energy systems: a review. \u003cem\u003eEnergy Syst\u003c/em\u003e (2024). https://doi.org/10.1007/s12667-024-00679-3.\u003c/li\u003e\n\u003cli\u003eJingyan Xie, Yun-Ze Li, Lizhu Yang, Yuehang Sun, Man Yuan, A review of the recent progress of stand-alone photovoltaic-battery hybrid energy systems in space and on the ground, Journal of Energy Storage, Volume 55, Part C, 2022, 105735, ISSN 2352-152X, https://doi.org/10.1016/j.est.2022.105735.\u003c/li\u003e\n\u003cli\u003eSarang, S.A., Raza, M.A., Panhwar, M. \u003cem\u003eet al.\u003c/em\u003e Maximizing solar power generation through conventional and digital MPPT techniques: a comparative analysis. \u003cem\u003eSci Rep\u003c/em\u003e\u003cstrong\u003e14\u003c/strong\u003e, 8944 (2024). https://doi.org/10.1038/s41598-024-59776-z\u003c/li\u003e\n\u003cli\u003eK. Keerthana, S. Singaravelu, \u0026quot;Enhancing the Robustness of P and O Algorithm-Based MPPT Control in Stand-Alone PV Systems through FineTuned PI Controller for Dynamic Load Variations,\u0026quot; \u003cem\u003eSSRG International Journal of Electronics and Communication Engineering\u003c/em\u003e, vol. 11, no. 6, pp. 9-19, 2024. \u003cem\u003eCrossref,\u003c/em\u003ehttps://doi.org/10.14445/23488549/IJECE-V11I6P102\u003c/li\u003e\n\u003cli\u003eS. Astaomar and B. Erkal, \u0026quot;Improving the Performance of an Incremental Conductance MPPT Algorithm Using Harris-Hawks Optimization in Photovoltaic Systems,\u0026quot; \u003cem\u003e2024 Global Energy Conference (GEC)\u003c/em\u003e, Batman, Turkiye, 2024, pp. 128-134, doi: 10.1109/GEC61857.2024.10881949.\u003c/li\u003e\n\u003cli\u003eM. Farhat, O. Barambones, L. Sbita and J. M. G. de Durana, \u0026quot;A stable FLC-based MPPT technique for photovoltaic system,\u0026quot; \u003cem\u003e2015 IEEE International Conference on Industrial Technology (ICIT)\u003c/em\u003e, Seville, Spain, 2015, pp. 890-895, doi: 10.1109/ICIT.2015.7125210.\u003c/li\u003e\n\u003cli\u003eFarhat, M., Flah, A., \u0026amp;Sbita, L. (2014). Photovoltaic maximum power point tracking based on ANN control. International Review on Modelling and Simulations, 7(3), 474-480.\u003c/li\u003e\n\u003cli\u003eY. Gao, S. Wang, T. Dragicevic, P. Wheeler and P. Zanchetta, \u0026quot;Artificial Intelligence Techniques for Enhancing the Performance of Controllers in Power Converter-Based Systems\u0026mdash;An Overview,\u0026quot; in \u003cem\u003eIEEE Open Journal of Industry Applications\u003c/em\u003e, vol. 4, pp. 366-375, 2023, doi: 10.1109/OJIA.2023.3338534.\u003c/li\u003e\n\u003cli\u003eY. Triki, A. Triki, A. Bechouche, D. O. Abdeslam and R. Porumb, \u0026quot;An Efficient Battery-Charging Algorithm with ANN based MPPT Method for Off-Grid PV Systems,\u0026quot; \u003cem\u003e2022 IEEE 21st international Ccnference on Sciences and Techniques of Automatic Control and Computer Engineering (STA)\u003c/em\u003e, Sousse, Tunisia, 2022, pp. 106-111, doi: 10.1109/STA56120.2022.10019093.\u003c/li\u003e\n\u003cli\u003eKatche, M. L., Makokha, A. B., Zachary, S. O., \u0026amp;Adaramola, M. S. (2023). A Comprehensive Review of Maximum Power Point Tracking (MPPT) Techniques Used in Solar PV Systems. Energies, 16(5), 2206. https://doi.org/10.3390/en16052206\u003c/li\u003e\n\u003cli\u003eTamer Khatib, Ibrahim A. Ibrahim, Azah Mohamed,A review on sizing methodologies of photovoltaic array and storage battery in a standalone photovoltaic system,Energy Conversion and Management,Volume 120,2016, Pages 430-448, ISSN 0196-8904, https://doi.org/10.1016/j.enconman.2016.05.011.\u003c/li\u003e\n\u003cli\u003eAlok Jain, Suman Bhullar,Design and performance analysis of solar PV-battery energy storage system integration with three-phase grid,Journal of Power Sources,Volume 640,2025,236486, ISSN 0378-7753, https://doi.org/10.1016/j.jpowsour.2025.236486\u003c/li\u003e\n\u003cli\u003eKirubakaran, V., \u0026amp; Singaravelu, S. (2024). A generalized mathematical modeling and performance analysis of GaAs, m-Si and Ge single - junction solar PV cells. Multidisciplinary Science Journal, 6(9), 2024176. https://doi.org/10.31893/multiscience.2024176\u003c/li\u003e\n\u003cli\u003eJ. Edler and N. Kondrath, \u0026quot;Bidirectional Interleaved Buck/Boost DC-DC Converter Design to Improve Power Density in High-Current Applications,\u0026quot; \u003cem\u003e2019 IEEE 62nd International Midwest Symposium on Circuits and Systems (MWSCAS)\u003c/em\u003e, Dallas, TX, USA, 2019, pp. 403-406, doi: 10.1109/MWSCAS.2019.8885244.\u003c/li\u003e\n\u003cli\u003eN. D. Bhat., D. B. Kanse., S. D. Patil., and S. D. Pawar., 2020, \u0026quot;DC/DC Buck Converter Using Fuzzy Logic Controller,\u0026quot; 5th International Conference on Communication and Electronics Systems (ICCES), Coimbatore, India., pp. 182-187.\u003c/li\u003e\n\u003cli\u003eYavuz Bahadır KOCA, Yılmaz ASLAN, Ahmet Y\u0026Ouml;NETKEN3, Y\u0026uuml;ksel OĞUZ4, \u0026ldquo;Boost Converter Design and Analysis for Photovoltaic Systems\u0026rdquo;, Conference Paper \u0026middot; April 2019.\u003c/li\u003e\n\u003cli\u003eH. Patel and V. Agarwal, \u0026ldquo;MATLAB-Based Modeling to Study the Effects of Partial Shading on PV Array Characteristics,\u0026rdquo; IEEE Transactions on Energy Conversion, vol. 23, no. 1, pp. 302\u0026ndash;310, March 2008. DOI: 10.1109/TEC.2007.914308\u003c/li\u003e\n\u003cli\u003eS. Jain and V. Agarwal, \u0026ldquo;Comparison of the Performance of Maximum Power Point Tracking Techniques for a PV System,\u0026rdquo; Solar Energy Materials and Solar Cells, vol. 90, no. 5, pp. 672\u0026ndash;685, 2006. DOI: 10.1016/j.solmat.2005.04.007\u003c/li\u003e\n\u003cli\u003eM. A. Elobaid, S. Z. Kassas, and M. F. El-Naggar, \u0026ldquo;Artificial Neural Network-Based Photovoltaic Maximum Power Point Tracking Techniques: A Survey,\u0026rdquo; Solar Energy, vol. 141, pp. 22\u0026ndash;45, 2017. DOI: 10.1016/j.solener.2016.11.018\u003c/li\u003e\n\u003cli\u003eN. Femia, G. Petrone, G. Spagnuolo, and M. Vitelli, \u0026ldquo;Optimization of Perturb and Observe Maximum Power Point Tracking Method,\u0026rdquo; IEEE Transactions on Power Electronics, vol. 20, no. 4, pp. 963\u0026ndash;973, July 2005. DOI: 10.1109/TPEL.2005.850975\u003c/li\u003e\n\u003cli\u003eT. Esram and P. L. Chapman, \u0026ldquo;Comparison of Photovoltaic Array Maximum Power Point Tracking Techniques,\u0026rdquo; IEEE Transactions on Energy Conversion, vol. 22, no. 2, pp. 439\u0026ndash;449, June 2007.\u003c/li\u003e\n\u003cli\u003eS. Mirbagheri, A. Dorri, and H. R. Najafi, \u0026ldquo;Modeling and Simulation of Photovoltaic Cells/Modules/Arrays with MATLAB/Simulink,\u0026rdquo; International Journal of Computer and Electrical Engineering, vol. 3, no. 5, 2011.\u003c/li\u003e\n\u003cli\u003eS. Jain and V. Agarwal, \u0026ldquo;Comparison of the performance of maximum power point tracking techniques for a PV system,\u0026rdquo; Solar Energy Materials and Solar Cells, vol. 90, no. 5, pp. 672\u0026ndash;685, May 2006.\u003c/li\u003e\n\u003cli\u003eM. A. Elobaid, S. Z. Kassas, and M. F. El-Naggar, \u0026ldquo;Artificial neural network-based photovoltaic maximum power point tracking techniques: A survey,\u0026rdquo; Solar Energy, vol. 141, pp. 22\u0026ndash;45, Jan. 2017. doi: 10.1016/j.solener.2016.11.018\u003c/li\u003e\n\u003cli\u003eN. Femia, G. Petrone, G. Spagnuolo, and M. Vitelli, \u0026ldquo;Optimization of perturb and observe maximum power point tracking method,\u0026rdquo; IEEE Trans. Power Electron., vol. 20, no. 4, pp. 963\u0026ndash;973, Jul. 2005.\u003c/li\u003e\n\u003cli\u003eT. Esram and P. L. Chapman, \u0026ldquo;Comparison of photovoltaic array maximum power point tracking techniques,\u0026rdquo; IEEE Trans. Energy Convers., vol. 22, no. 2, pp. 439\u0026ndash;449, Jun. 2007.\u003c/li\u003e\n\u003cli\u003eM. Nader, A. Dib, and S. Ouchen, \u0026ldquo;A fuzzy logic controller based MPPT technique for photovoltaic generation system,\u0026rdquo; Int. J. Smart Grid, vol. 5, no. 3, pp. 144\u0026ndash;151, Sep. 2021.\u003c/li\u003e\n\u003cli\u003eP. C., R. Geethamani, G. Radhakrishnan, S. Kishore Kumar, and C. Manoj, \u0026ldquo;Comparison of solar P\u0026amp;O and FLC-based MPPT controllers \u0026amp; analysis under dynamic conditions,\u0026rdquo; EAI Endorsed Transactions on Energy Web, vol. 10, no. 6, e3, 2023.\u003c/li\u003e\n\u003cli\u003eC. R. Algar\u0026iacute;n, J. T. Giraldo, and O. R. \u0026Aacute;lvarez, \u0026ldquo;Data from a photovoltaic system using fuzzy logic and the P\u0026amp;O algorithm under sudden changes in solar irradiance,\u0026rdquo; Data in Brief, vol. 21, pp. 190\u0026ndash;195, Aug. 2018.\u003c/li\u003e\n\u003cli\u003eM. S. Ali, A. A. Elserougi, and A. M. Massoud, \u0026ldquo;A comprehensive comparison of different MPPT techniques for photovoltaic systems,\u0026rdquo; IEEE Int. Symp. Power Electron. Distrib. Gener. Syst., 2014.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Artificial Neural Network (ANN), MPPT, Standalone PV System, Bi-Directional Converter, Battery Energy Storage, NPC Inverter, 3-Phase AC Load, DC Link, Energy Management, PI Controller","lastPublishedDoi":"10.21203/rs.3.rs-6831075/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6831075/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis paper presents an intelligent energy management system for a standalone photovoltaic (PV) system integrated with a lithium-ion battery and a three-level neutral-point clamped (NPC) inverter to supply a three-phase AC load. An Artificial Neural Network (ANN)-based Maximum Power Point Tracking (MPPT) controller is employed to enhance the efficiency and dynamic response of the PV system under varying irradiance conditions. The extracted power is regulated through a DC-DC converter, maintaining an optimal DC link voltage.\u003c/p\u003e\u003cp\u003eTo ensure reliable power delivery and load balancing, a bi-directional DC-DC converter is interfaced between the battery and the DC link. A Proportional-Integral (PI) controller governs the charging and discharging of the battery based on the power difference between the PV generation and the AC load demand. This mechanism allows for smooth energy transition, enabling battery charging during surplus PV generation and discharging during power deficits.\u003c/p\u003e\u003cp\u003eThe DC link supplies power to a three-level NPC inverter, which is responsible for converting DC power into high-quality AC output for three-phase loads. An inverter control strategy is implemented to maintain voltage stability and minimize total harmonic distortion (THD). The proposed system is simulated in MATLAB/Simulink, and its performance is analyzed under dynamic load and environmental conditions. Results demonstrate improved MPPT accuracy, effective battery management and stable AC output voltage, showcasing the suitability of the system for off-grid applications.\u003c/p\u003e","manuscriptTitle":"ANN-Based MPPT Control of a Standalone PV System with Bi-Directional Battery Management and NPC Inverter for Dynamic 3-Phase AC Loads","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-08 16:08:38","doi":"10.21203/rs.3.rs-6831075/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"daa7d892-2cbc-4586-a3d9-e9f53b1ab202","owner":[],"postedDate":"August 8th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-01-19T01:08:13+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-08 16:08:38","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6831075","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6831075","identity":"rs-6831075","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}