Notch Filter-Assisted DC-Link Stabilization for Improved Current Ripple Suppression in SLL Converter-Based BLDC Motor Drives

The brushless dc motor (BLDCM) drive is suffering from current/torque ripples. The DC-DC converter based BLDCM drive topology is widely preferred to suppress ripples. The improved characteristics such as low mechanical oscillation and stiff-speed regulation of BLDCM further introduces undamped AC signal which results destabilization of DC-link power (DCLP) supply. The different breeds of filter techniques have been recommended to stabilize the DCLP supply. However, the recommended filters are composed on the basic of concept of low pass filter (LPF) or band pass filter (BPF). In addition, these filters requires a precise association with active atonement, current reference and nonlinear interaction. In this context, a notch-filter (NF) has applied to stabilize the DCLP supply. The NF is investigated for a super-lift-Luo (SLL) converter assist BLDCM drive. The SLL converter is employed to suppress commutation current ripple (CCR) whereas the NF is applied to draw stabilized DCLP supply. A hardware prototype is developed and results are given to confirm the effectiveness of the proposed stabilization strategy and design. Thus, the suggested compensation technique effectively stabilized the DCLP supply of SLL converter employed to CCR mitigation in the BLDCM drive. Hosted file Manuscript.docx available at https://authorea.com/users/920992/articles/1293142-notch-filter- assisted-dc-link-stabilization-for-improved-current-ripple-suppression-in-sll-converter- based-bldc-motor-drives 1 Notch Filter -Assisted DC -Link Stabilization for Improved Current Ripple Suppression in SLL Con verter -Based BLDC Motor Drives Dileep Kumar1*, Surya Deo Choudhary1, Md Tabrez1, Saket Kumar Singh1 1Department of Electrical and Electronics Engineering, Motihari College of Engineering, Motihari 1Email: [email protected], [email protected], [email protected], [email protected] Abstract- The brushless dc motor (BLDCM) drive is suffering from current/torque ripples. The DC - DC converter based BLDCM drive topology is widely preferred to suppress ripples. The improved characteristics such as low mechanical oscillation and stiff -speed regulation of BLDCM further introduces undamped AC signal which results destabilization of DC -link power (DCLP) supply. The different breeds of filter techniques have been recommended to stabilize the DCLP supply. However, the recommended filters are composed on the basic of concept of low pass filter (LPF) or band pass filter (BPF). In addition, these filters requires a precise association with active atonement, current

and nonlinear interaction. In this context, a notch-filter (NF) has applied to stabilize the DCLP supply. The NF is investigated for a super-lift-Luo (SLL) converter assist BLDCM drive. The SLL converter is employed to suppress commutation current ripple (CCR) whereas the NF is applied to draw stabilized DCLP supply . A hardware prototype is developed and results are given to confirm the effectiveness of the proposed stabilization strategy and design. Thus, the suggested compensation technique effectively stabilized the DCLP supply of SLL converter employed to CCR mitigation in the BLDCM drive.

Notch-filter, DC-link, SLL converter, BLDCM drive, current /torque ripples 1. INTRODUCTION The BLDCM provides great torque to weight rate , compact size, and excellent efficiency. The performance of these motors has notably enhanced due to substantial advancements in power electronics, magnetic properties, and speed control systems in recent years [1-5]. Consequently, the BLDCM is extensively utilized in domestic and industrial appliances, as well as modern vehicles. Nevertheless, the BLDC M has a drawback of expensive cost c ompared to the DC motor since it is essential to apply an inverter and controller to detach the brush unit of DC motor [6-8]. In an optimal BLDCM, there exists trapezoidal back-electromotive force (back-EMF) with a fixed dc supply system [9-10]. Consequently, the motor’s current generated by the input voltage rapidly ascends to a peak value (i.e. steady-state) as seen in Fig. 1, and the ripple free torque is generated. Even so, current properties of the practical BLDCM diverge from the ideal scenario. The current, affected by inductance and resistance, possesses a time constant and cannot instantaneously reach a steady state, depicted in Fig. 1. Fig 1. Ideal and actual current waveforms. Consequently, a current ripple is originated by the effect of inductance. The torque ripple is directly influenced by current ripple when the back -EMF exhibits a trapezoidal pattern. Therefore, the CCR, typically arising from inadequate phase current control, has consistently been a major barrier to th e excellent performance of BLDCM. Numerous studies are being performed to mitigate the CCR. The origins of current/torque ripple s in BLDC Ms and develops a framework for its reduction adopting a phase current forec asting approach [1]. Furthermore, by adjusting the duty ratio based on prior assessment, the predictive control prevents the CCR, hence minimum torque ripple is drawn. The CCR reduction technique for a BLDCM that results from motor's parameters with an artistic single - cycle mean algorithm [3-4]. A precise investigator is not needed as neither the back EMF nor precise rotor location details are required. No current sensor is required for measuring the phase current. The rapid stabilization with torque ripple mitigation is demonstrated by a bio -inspired analytical approach in a BLDCM drive [5]. A spider -based mechanism has designed to produce PWM signals for the inverter unit and the control output for applied capacitor. The technique for integrating the motor's current supply that guarantees low torque ripple for functioning of the drive [7]. The synthesis is conducted theoretically using the vector technique, and the resultant vector trajectory is subsequently articulated in motor phase coordinates. A control strategy is provided to mitigate motor current harmonics and its negative impacts whi le enhancing the motor's overall performance [9]. An inverter powered by a boost converter, utilizing Targeted Harmonic Deletion PWM and a passive LPF, is being used to diminish harmonic and enhance overall system performance, including torque ripples. An efficient peak torque for each ampere control approach is presented for BLDCM drives [10]. In this approach, the impact of core/iron losses is considered. In addition, the Lagrange's theorem as well as nonlinear control have used to suppress the CCR. Previous studies present a BLDCM drive without considering a DC-link voltage. Nevertheless, the suppression of current/torque ripple is not efficient. Therefore, variety of DC-DC converter assist BLDCM drive topologies have been suggested to encounter the CCR. The multilevel inverter combining a SEPIC converter based BLDCM drive has developed to alleviate the CCR [ 11]. This approach possess a complex control due to increased switch count. A novel DC-link control is introduced for the BLDCM drive [12]. This approach exhibits a rapid active torque output that resembles that of the direct torque control (DTC) mechanism. This technique does not necessitate the transfer of motor variables to the stationary or synchronous reference frame, which often incurs a significant computational overhead. A broad range of speed regulation for BLDCM drives by the modulation of the DC-link voltage of a dc-ac unit utilizing a high-gain SEPIC circuit-based one Switch Converter has applied [13]. The DC -DC converter integrated into the system provides a substantial voltage gain suitable for low to extremely high -speed control. The converter continuously receives input current from the supply. The power sources are linked to the double input dc -dc converter to supply energy to the BLDCM [14]. The suggested double in put interspersed SEPIC converter minimizes input and output ripples facilitating smooth operation of the BLDCM drive during varying load conditions. Due to the commutation performed by inverter unit, ripples are produced in the torque spectrum, thereby impacting the input source power. To address this issue, a bridge-type DC-DC converter inputted a BLDCM drive has described [15]. The technique employing a boost converter circuit to mitigate the CCR of a BLDCM characterized by rectangular flux distribution [16]. A Luo converter is utilized to boost the input voltage. The desired transient voltage waveform is achieved by adjusting the settings of the DC-DC converter according to the mathematical modeling. To precisely compute the line current, the phase resistance is considered, and the rates of increase and decrease of the phase currents are compared. Moreover, it is demonstrated that the line current will remain constant if the DC voltage is increased throughout the commutation interval (CI) . The literature introduces a novel way for lowering the CCR [17]. The suggested circuit employs a Cuk converter at the leading end, characterized by high gain. An additional switch is coupled alongside the DC-link to adjust the slew rates of incoming phase current during the inverter's phase transition. The DC-DC converters consist of power semiconductor switches. The switches are triggered at very high switching frequency. Due to duty cycle, the power switches are turned -on and turned-off at high rate. Therefore, the desired output of the converter is no longer stable. During the continuous turn on/off process, an AC portions have overlapped to the DC -link termed as switching frequency harmonics (SFH). The SFH consists of several frequency terms which are multiple of fundamental triggered frequency of converter. So that, variety of DC -link stabilization methods have been illustrated. The inherent sliding mode technique (SMT) has designed to impro ve the DC -link profile [18]. This scheme offers characteristic of Kalman filter. This method is quit complex and not effective for BLDCM drive. To address the issues of sluggish reaction time and rattling in conventional sliding mode control of BLDCMs, an integrated fast intermediate SMT approach is developed to enhance the speed and the system's capability [19]. A Luenberger observer has applied to alleviate the effects of external load disturbances. The work aims to extend the permis sible range of the DC - link voltage in buck converter operations while minimizing the effective harmonics [20]. This study enhances the inside loop by integrating unique subsystems, including an auto -tuner and active damping with a customized loop gain. Using the typical disruptive spectator layout process, no actual motor characteristics or load details are needed for the controller [21]. The first characteristic is the use to avoid explicit divergence of speed information. The second is that active-damping required for system behavior. Besides, several traditional compensating methods have been incorporated. These techniques are designed on the principle of BPF and LPF. Nevertheless, the LPF necessitates a strong correlation between dynamic coupling and effective damping. On the other side, it is known that the BPF is designed for a particular range of frequency. As frequency range increases the bandwidth also increase which causes additional noise and inferior accuracy . Therefore, the DC-link stabilization is not satisfactory and the SFHs are still impose with the DC-link. The SFH produces undesirable regulation of speed, unintended mechanical oscillation, and unpleasant auditory noise. It also impacts the precision and rotor alignment of the BLDCM. In this context, this paper proposes a NF-based DC-link Stabilization of SLL converter allied to suppress CCR in BLDCM d rives. The PI controller (PIC) is extensively applied; yet, steady -error is not removed whenever the PI C tracks AC signals. Additionally, this approach will boost the control complexity and lower the sys tem's dynamic characteristic [22]. Moreover, the classical proportional-resonant compensator (PRC) offers negligible ste ady-error while following the AC signals at resonant frequency. Nevertheless, its tracking capability fal ls with frequency variations [23-24]. In comparison to the PRC, the NF has superior blocking capabilities, as well as enhanced overall effectiveness. This study introduces a NF that might be tuned at a switching frequency (i.e. single/selective frequency). The robustness of the NF is thoroughly investigated and assessed with the transfer characteristic and the bode-plot approach. The analytical design and functionality of the suggested NF are illustrated. Finally, the NF is applied w ith SLL converter to stabilize the DC -link voltage. The SLL converter replicates a traditional boost converter [25-27]. It is noteworthy that the SLL converter provides a broad gain in voltage with positive polarity with respect to a n existing Luo converter [26-28]. Consequently, the SLL converter has been em ployed to achieve regulated DCLP supply, with the duty ratio adjusted in accordance with the back -EMF. The DCLP supply circuit controls the DC-link voltage prior and later the CI in order to minimize CCR. The findings of the suggested SLL converter circuit which utilizes a NF, are demonstrated using MATLAB®/Simulink. Besides, a testing prototype using a STM32F407VGT controller is designed to confirm the fruitfulness of recommended control approach. 2. CURRENT/TORQUE RIPPLE ANALYSIS OF BLDCM A typical equivalent inside view of BLDCM is depicted in Fig. 2. It is assumed that the each phase consist of a series connection of resistance, self-inductance and back-EMF. Fig. 2. Exact inside structure of BLDC motor. As per Fig. 2, the voltage equation between phase and neutral could be derived as: a b c a n a a a ab ac a b a c b n b b b ab bc b c a b c n c c c ac bc c di di diV -V = r i +l + m m +edt dt dt di di diV -V = r i +l + m m +edt dt dt di di diV -V = r i +l + m m +edt dt dt +  +   +  (1) where Vn represents the differential potential between neutral and ground. Va, Vb and Vc symbols used for pole voltage. ia, ib, and ic used for phase current indication. ra, rb and rc indicate the phase resistance whereas symbols ea, eb, and ec show the back -EMFs. la, lb, and lc mentioned for phase inductances whereas mab, mbc, and mac used for mutual inductances, respectively. In practice, a balance three-phase star connected BLDCM having identical phase winding is used for mathematical analysis. As results, la=lb=lc=l, mab=mbc=mac=m and ra=rb=rc=r. Eq. (1) might be rearranged as: a n a lm a a b n b lm b c c n c lm c c V -V r 0 0 i 0 0 i e dV -V = 0 r 0 i + 0 0 i + e dtV -V 0 0 r i 0 0 i e                                                   (2) where ℾlm=l-m The motor speed is in radian per second and symbolized by r . Thus, the rotor gross torque is calculated from Eq. (3). †rotor a a b b c c gross r e i +e i +e i=  (3) It is known that the electrical length of the flattop area of back -EMF is equal to 120o electrical. To produce a constant power, the smooth phase current must flow in the region of flattop area of back - EMF only. This phenomenon can be achieved by a three-phase inverter which operates in 120o mode. Ideally, incoming (active) phase current must gain s the steady value and outgoing (inactive) current must reaches the zero value instantly to achieve a synchronized commutation. However, active and inactive phase current do not tens to steady and zero value during the CI due to phase inductance. In this study, it is supposed that the phases ‘a’ and ‘c’ are active and phase ‘b’ is inactive. Further, an assumption is considered that the commutation occurs between phases ‘ a’ and ‘b’. Hence, ia and ib are is taken as outgoing and incoming currents, respectively. Besides, phase ‘c’ remains unchanged and current ic hold in steady-state. Further, the Eq. (3) might be expressed as follow: †rotor a a c c gross r e i +e i=  (4) The standard back-EMF and torque equation of BLDCM is situated in Eq. (5). rotor gross τm peak back-EMF em r †= κi E= κv   (5) where Ktm = torque constant, ipeak=ia=ib=ic and Kem= back-EMF constant. During the short duration of commutation within the flattop region of the back EMF, the certain characteristics become apparent. The emax is the peak value and following observations have made: Va= 0 = Vc; Vb = Vinput; ea = eb = emax and ec= -emax. Before the CI, the torque could be derived from Eq. (4). † rotor max max gross r 2e i=−  (6) During the CI, the phase currents are expressed as follows: input maxa lm input maxb lm input maxc lm (V + 2e )di =-dt 3 2(V - e )di =dt 3 (V - 4e )di =dt 3         (7) The time required for outgoing current to falls from peak to zero value is known as fall time (ζ f). The time span ζf for current ia can be derived from Eq. (7) as per below equation. max 0 lm lm max fa i input max input max 3 3 i= - di =(V +2e ) (V +2e )   (8) Similarly, the time duration for which the incoming current reaches from zero to maximum value is term as raise time (ζr). The time span ζr for current ib can be derived from Eq. (7). lm max r input max 3i= 2(V - e )  (9) At the CI, the current ic might be estimated as below: input maxcommutation c lm (V - 4e )i = t 3 (10) Finally, the gross current through phase ‘c’ is as per Eq. (11). input maxinstantaneous c max lm (V - 4e )i = i + t 3Γ (11) Eqs. (4) and (11) are used to draw the torque equation for the BLDCM during the CI. † input maxcommutation max gross max r lm (V - 4e )2e= i + t 3Γ    (12) Finally, the torque ripple offers by the motor is obtained from Eqs. (6) and (12). † † † input maxripple -rotor commutation gross gross gross lm (V - 4e )=t 3Γ=− (13) The following observations are drawn from Eqs. (7-13) • As Vinput>4emax, i.e. ζr> ζf and ic decreases. Consequently, torque likewise continues to increase and rotor speed is less than the actual speed as shown in Fig. 3(a). • As demonstrated in Fig. 3(b), as Vinput<4emax, i.e. ζ r < ζf and ic raises. Consequently, the torque continues to decrease and the rotor speed is greater than the actual speed. • Referring Fig. 3(c), as Vinput=4emax then ζr=ζf, and ic remain constant. Hence, the torque ripple is zero and the rotor rotates at the actual speed. Fig. 3. characteristic of phase current in the CI, (a) low speed (b) high speed, and (c) actual speed. Eq. (13) demonstrates that the input supply (Vinput) and peak back -EMF (emax) at the CI are the only parameters that affect the sl ope of the active and inactive phase current. Typically, Vinput is invariant while emax is a variable parameter and depends on rotor speed. The torque ripple in Eq. (13) is proportional to the (Vinput -4emax). The torque ripple can be made zero if Vinput is kept near to 4emax during the CI. In this paper, the utilization of SLL converter enables the achievement of a regulated DC-link voltage (Vinput=Vs =4emax) during the CI. This controlled voltage ensures efficient suppression of torque ripples. By satisfying the condition Vinput=4emax as indicated in Eqs. (8) and (9), the rising and falling current exhibit similar slew rates (ζr=ζf), resulting in zero torque ripples during CI. The elimination of torque ripples facilitates the realization of a rectangular phase current waveform, leading to the production of a smoother electromagnetic torque. 3. DESIGN AND WORKING OF THE SLL CONVERTER The switching performance of SLL converter is as mimic of the boost converter. The SLL converter possess the step-up output voltage as well as power. It offers wide transfer gain with the lesser count of the circuit parameters as compare to the boost converter. The traditional Luo converter lifts the positive voltage to negative load voltage [16]. Besides, the SLL converter boosts the positive voltage to the positive load voltage. The circuit of SLL converter is depicted in Fig. 4. Fig. 4. Basic circuitry diagram of the SLL converter. The converter comprises of primary and secondary side circuits. The primary circuit consists of a MOSFET switch (Q), an inductor (Lp), and diode ( Dp) and capacitor (Cp). The secondary side circuit consists of a diode ( Ds) and a shunt combination of capacitor ( Cs) and a resistive load ( Ro) long with shunt load. Following considerations are taken in account: (a) All components are ideal and lossless. (b) The inductor current shows a linear characteristic during charging and discharging. (c) The capacitor holds a constant voltage itself. (d) The parasitic effects are avoided. (a) (b) Fig. 5. The operational stages of the SLL converter while Lp is in the CCM: (a) the switch (Q) is closed (Mode-A) (b) the switch (Q) is closed (Mode-B). The inductor works in continuous conduction mode (CCM). In order to investigate the effectiveness of SLL converter, two separate types of functionality have carried out given as: Mode-A (0 ≤ t ≥ T on/αT): In this mode, the switch ( Q) is closed the and diode ( Dp) is acti vated as illustrated in Fig. 5(a). The current ( ILp) through inductor ( Lp) starts to grow lineally with primary voltage (Vp). On the other side, the voltage ( VCp) across the capacitor (Cp) begins to charge to voltage (Vp). Additionally, the diode (Ds) is negatively biased hence it is opened. Therefore, the capacitor (Cs) starts to discharge via load resistance (Ro). The ideal voltage and current characteristic of the SLL converter is mentioned in Fig. 6. Referring to Fig. 6, the voltage across the Lp is given by: Lp pVV = (14) By applying generalized voltage equation of the inductor we get, Lp Lp p dIVL dt= (15) The Eq. (15) could be reform as follows: 0 1 on mx mn T Lp p p I Ton Lp p pI0 dI V dt L 1dI = V dtL  =         (16) The desired peak-to-peak ripple current might be derived as below: p mx mn Lp 0p αVI - I = = ΔIfL (17) Where Imx, Imn, and ΔILp are the maximum and minimum current values, and peak -peak ripple current respectively. Ton, T, and f0 are the switch closing time, time span of one cycle, and switching frequency, respectively. Fig. 6. Ideal voltage and current through circuit components. Mode-B (Ton ≤ t ≥ T/(1-α)T): In this mode, the switch (Q) and diode (Dp) are opened as illustrated in Fig. 5(b). Inherently, the inductor (Lp) and capacitor (Cp) are reconnected in series via diode (Ds). The current ILp starts to decrease via capacitor (Cp), diode (Ds) and load. The capacitor (Cp) begins to charge negatively toward the Vp. Finally, the capacitor (Cs) is charged and its voltage is equal to Vs. By referring Fig. 6, the voltage across the inductor (Lp) is follow: 2Lp p sV V V=− (18) By using classical method of voltage equation for inductor, the current ripple could be estimated by solving Eq. (19) 2 2 mn mx on I T Lp p s pIT ps mn mx Lp 0p 1dI = ( V V )dtL (1- α)( V V )I - I = = - ΔIfL  −   −     (19) As per voltage second balance theory, the static voltage across the inductor is zero. Therefore, using Fig.6, the static voltage expression for inductor (Lp) is derived as follow: on p p s (T -T )1 V + (2V -V )= 0TT (20) The Eq. (20) could be expressed as below: p on s onV (2T -T )=V (T -T ) (21) The desired transfer gain of the SLL converter might be obtained from Eq. (21). s p V 2- α= = T( α)V 1- α (21) where α is the duty cycle/ratio and it is ratio of the Ton to T. A comprehensive performance assessment of numerous DC-DC converters is listed in Table 1. Table 1: A comprehensive performance assessment of numerous DC-DC converters The SLL converter and DCLP supply mechanisms have constructed with MOSFETs. The desired DCLP supply is obtained by controlling the duty ratio. The inverter unit is inputted accordance with operation of switches P1 and P2. During CI, switch P2 is engaged to select the voltage ( Vs) of SLL converter. The inverter unit powered b y the selected voltage ( Vs). The inverter is fed via voltage ( Vp) during prior and later the commutation with the help of switch P1. The voltage (Vs) of SLL converter is expressed as: (2 ) s max pV = 4e = V (1- ) −  (22) By employing Eq. (5), Eq. (22) may be reformulated as (2 ) )em r p4κ v = V (1- −  (23) The ‘α’ might be calculated using Eq. (23) to ensure Vs = 4emax. Sl. No Variety of Converters Number of Parameters Activity Circuit Involution Polarity Primary Ripple Secondary Ripple D S L C Tot. 1. Cuk [17] 1 1 2 2 6 CCM Medium (-ve) Medium High 2. SLL converter [25-28] 2 1 1 2 6 CCM Low (+ve) High Medium 3. Z-source [27] 3 1 3 2 9 DCM Medium (-ve) High High 4. SEPIC [11, 26] 1 1 2 2 6 CCM High (+ve) Medium High 5. Landsman [27] 4 2 5 4 15 CCM High (+ve) High High em r p em r p κ v - 0.5Vα= κ v +0.25V   (24) As per Eq. (24), the duty ratio is calculated to maintain Vs = 4emax during the CI. The time length of commutation might be computed from Eq. (12). The exact communication length (Tx) is usually chosen over tr to accommodate fluctuations in load/ speed. The picking of Tx need not necessitate a n exact computation. 4. FORMULATION AND STABILITY CHARACTERIZATION OF THE NF Fig. 7 illustrates the fundamental component of the NF. By minimizing both phase errors as well as magnitude, the compensator provides an effective solution for DC-link stabilization. The NF can be adjusted specifically for a selected frequency. The fundamental switching frequency element exerts a minimal impact on DCLP supply as compared with the higher -order switching frequency harmonics. To remove the higher-order SFH, the NF is adjusted to the switching frequency to ensure precise DCLP regulation. The input s ignal undergoes multiplication with cosine and sine functions during the demodulation stage, facilitating the shift of higher -order SFH components to the dc component and effectively doubling the switching frequency term. The integrator term of a PI controller performs dual functions: it eliminates the double switching frequency component from the signal and provides integral action. The PI controller is characterized as a LPF in nature, and its transfer function can be represented as Ypi_dc(s). Both the su ggested filter and the usual PI controller exhibit akin dynamic characteristics over their designated bandwidth. Fig. 7. Schematic circuit network of the NR. The time-domain response of the recommended filter could be expressed as follow:     o q pi_dc q o q pi_dc q x(t)= v (t)cos(2 f t) Y (t) cos(2 f t) + v (t)sin(2 f t) Y (t) sin(2 f t) = X + X          (25) Xα and Xβ are two parts drawn from Eq. ( 25). Additionally, both sub -functions, x1(t) and x2(t), are defined as below:     1 o q pi_dc 2 o q pi_dc x (t)= v (t)cos(2 πf t) ×Y (t) x (t)= v (t)sin(2 πf t) ×Y (t)   (26) By applying Laplace transform, Eq. (26) might be expressed as:   qq 1 o q pi_dc j2 f t -j2 f t 1 o pi_dc x (s)= L v (t)cos(2 f t) Y (t) e +ex (s)= v (t)( ) Y (t) 2        (27) Therefore, 1 o q o q pi_dcx (s)= 0.5{v (s+ j2 f )+v (s - j2 f )}Y (s) (28) Likewise, x2(s) could be derived from Eq. (26). 2 o q o q pi_dcx (s)= 0.5j{v (s+ j2 f )+v (s - j2 f )}Y (s) (29) Applying Laplace transform, a complete form of Xα is estimated from Eq. (25). 1q 1 q 1 q X = L[x (t)cos(2 f t)] = 0.5[x (s+ j2 f )+ x (s - j2 f )]     (30) Eq. (30) might be restructured as. pi_dc q o q q o q q pi_dc q o q q o q q X = 0.5[0.5Y (s+ j2 f ){v (s+ j2 f + j2 f ) +v (s+ j2 f - j2 f )} +0.5Y (s - j2 f ){v (s+ j2 f - j2 f ) +v (s - j2 f - j2 f )}]             (31) Hence, pi_dc q o q pi_dc q o o q X = 0.25[Y (s)(s+ j2 f ){v (s+4j f )+V(s)} +Y (s - j2 f ){v (s)+v (s - 4j f )}]      (32) Likewise, Xβ could be obtained like: pi_dc q o o q pi_dc q o o q X = 0.25[Y (s)(s - j2 f ){v (s)- v (s - j4 f )} +Y (s)(s+ j2 f ){v (s)+v (s+ j4 f )}]      (33) Finally, o pi_dc pi_dcX + X = 0.5v (s)[Y (s - j2 f)+Y (s+ j2 f)]  (34) It is noteworthy that the frequency response of Eqs. (25) and (34) remains identical regardless of Ypi_dc(s). This approach yields the double switching frequency element within the demodulation phase. However, in the modulation phase, the double switching frequency part is effectively eliminated, as seen in Eq. (34). Consequently, just the fundamental switching frequency component is retained, while the remaining switching frequency aspects are greatly eliminated, so stabilizing the DCLP. Ypi_dc(s) is utilized to nullify error and achieve zero error in the DC system. The transfer function of Ypi_dc(s) is as similar as a PI controller given below: i pi_dc pY = + s  (35) Eq. (25) clearly indicates that Ypi_dc(s) acts as LPF. Consequently, alteration in Eq. (34) yields a LPF to frequency changing alteration as mentioned in Eq. (36). 22 q pi_ac pi_dc s +(2 f )Y = Y ( ) 2s  (36) The Ypi_ac(s) represents an AC term and act as ideal frequency changing alteration. Further, Ypi_ac(s) can be written as: i pi_ac p 22 q 2sY = + s +(2 f )   (37) The implementation of a DC integrator with infinite gain, or an analogous AC lossless resonant circuit, is unfeasible due to the inclusion of undesirable standards in AC systems. Nevertheless, the optimal Ypi_dc(s) is frequently predicted by Eq. (38). i pi_dc pY = + s+   (38) Eq. (38) might be expressed as follow: i pi_ac p 22 q 2sY = + s +2 s+(2 f )   (39) where Ψ = minimum breakpoint frequency The bode -plot for Eqs. (37) and (39) is illustrated in Fig. 8 to study a frequency response characteristic. The Fig. 8 indicates that the NF is specially implemented to provide high gain at the switching frequency ( 10 kHz switching frequency is considered for analysis). Additionally, the NF serves as an effective solution for DCLP supply stabilization due to its large gain (i.e. wide rejection ability. Hence, this paper employs a NF to remove greater -level SFHs from the SL L converter. The greater-level SFHs superimpose with voltage and current significantly impact the system's functionality in contrast to lower -order SFHs. hence, the filter is calibrated to the smallest frequency (i.e. f0=fq=10kHz) applied to the SLL converter. This setting aims to minimize voltage overshoot, enhance robustness and performance. (a) (b) Fig. 8 . Frequency response of notch-filter for Ψ=0.3 rad/sec, f0=10 kHz , p =.8, i =.8 (a) Ideal (b) non -ideal performance. Furthermore, the NF is an appropriate choice for DCLP supply stabilization owing to its large gain, (i.e. indicating strong rejection abilities). This study implements a NF to remove higher -levels SFH from the SLL converter to stabilize the DCLP supply. The higher-levels SFH in the voltage and current significantly influence converter performance in contrast to lower -levels SFH. Consequently, the NF is calibrated to the operating frequency (i.e. 10 kHz is taken for this analysis) applied in the SLL converter to prevent voltage overshoot, enhance robustness. Likewise, regarding the suggested control strategy, an empirical investigation indicates that the NF commonly yields unexpected peaks (>0 dB) around the specified frequencies (i.e., 10 kHz). This might result in unstable controller performance due to the switching frequency (i.e., 10 kHz) under actual circumstances is not consistently uniform and may deviate somewhat from the intended value. 5. SIMULATION AND EXPERIMENTAL RESULTS The simulation has conducted to assess the effectiveness and feasibility of the sugg ested control approach. Fig. 9 illustrates a framework of the SLL converter assists BLDCM drive system that relies on a NF-filter. Further, an experimental prototype is constructed to confirm the simulation results. The parameters utilized in the suggested system are laid out in Table 2. Table 2: List of parameters Parameters Rating Vin 24 V ra = rb = rc 0.30Ω Torque 0.4 N-m la = lb = lc 0.18mH Rate current 5 A Speed 4500 rpm Frequency of SSL converter (f0) 10 kHz Fig. 10 depicts the experimental findings of the SLL converter when it was operated at a switching frequency of 10 kHz. The SLL converter provides continuous inductor currents (ILp), as shown in Fig. 10(a). In addition, the experimental waveforms of voltage and current through Lp is mentioned in Fig. 10(b). On the other hand, the Vs shows notable ripples owing to SFH as referred in Fig. 11(a). The SFH destabilized the DCLP supply of the SLL converter. Fig. 9. Layout of working model of the proposed system. This is an aspect that should be taken into consideration. When the SFH is present, the output voltage of the SLL converter is impacted negatively. Hence, the SLL converter is tested by using the NF. Fig. 11(b) depicts the voltage waveform that is produced by the SLL converter when it is equipped with the NF. It is clearly noted that the SFH is effectivel y alleviated and a stabilized DC -link is obtained. We have arrived at the view that the NF is an efficient su ppressor of the SFH. In Fig. 9 , switch P1 is activated prior to and after the commutation times yet the BLDCM is powered by the primary voltage (Vp=Vinput). Switch P2 is consciously operated to ensure that Vs equals to the 4emax at the commutation. (a) (b) Fig. 10. The experimental findings of the SLL converter (a) voltage and current through the switch (Q) and C p (b) voltage and current through the Ls. (a) (b) Fig. 11. Input and output voltage of the SLL converter (a) without using NF (b) along with NF technique. It is important to note that the BLDCM obtaining power from the output voltage (i.e. Vs=4emax) of the SLL converter. The simulated phase ‘ a’ current (ia) of the traditional BLDC M is given in Fig. 12 for the full load conditions. Additionally, under the similar situation, the phase ‘a’ current (ia) of the SLL converter without NF scheme assist BLDCM drive is illustrated in Fig. 13. Fig. 12. Simulation result of standered BLDCM current ia. Figs 12 and 13 indicate that significant current ripples persist during the commutation due to SFH. The DC-link of the SLL converter is greatly affected by the SFH. Consequently, the NF is employed to reduce higher-order SFH and to stabilize the DC-link. Within the commutation, the NF is coupled with point T1. After the CI, the NF is coupled to point T2 to reduce the SFH associated to the Vs. The suggested filter technique based SLL converter assist BLDCM drive is tested under identical situation and simulated current (ia) is illustrated in Fig. 14. Fig. 13. Simulation result of current ia for sll converter assist BLDCM drive without filter Fig. 14. Simulation result of current ia for SLL converter assist BLDCM drive with NF method. A prototype of the said drive utilizing the SLL converter along with NF has designed to confirm the simulation findings. The gate signals are generated by using STM32F407VGT microcontroller. The experimental results have recorded in the four channel DSO manufactured by Tektronix TBS1102B- EDU and RIGOL-DS1102E. Fig. 15 illustrates the experimental findings of the switching sequences and voltage profiles for the switches P1 and P2. Fig. 15. Switching sequences and voltage profiles of P1 and P2. Additionally, the experimental results of voltage across the inverter are illustrated in Fig. 16, taking into account the switching performance of P1 and P2. It is important to observe that inverter is inputted by output of the SLL converter which is similar to four times the back-EMF at the CI (i.e., Vs=4emax). In the non-CI, the voltage pattern across VSI unit is equal to the primary voltage (Vinput). Fig. 16. Experimental results of voltage across the inverter (i.e.V0=4emax). The hardware findings of motor currents for a conventional BLDCM under different loading circumstances are depicted in Fig. 17. The excessive current ripples occur in the current as a result of the motors' inductances. The excessive current ripples occur in the current as a result of the motors' inductances. Referring to Fig. 17, the deviation in current is covered by a red color rectangle. (a) (b) (c) Fig. 17. Experimental result of standered BLDCM current (a) light load (b) moderate load (c) full laod. (a) (b) (c) Fig. 18. Practical result of phase current for SLL converter assist BLDCM without filter(a) light load (b) moderate load (c) full laod. At the CI, the rate of rising and falling currents are different mainly due to motors’ inductance. It is therefore, necessary to have a n SLL converter in order to keep Vinput (i.e. Vs) equal to 4emax (i.e. Vs=4emax) during the phase transition process. Hereby, the practical results of phase current for the SLL converter assist BLDCM without applying filter is mentioned in Fig. 18. It is seen that the ripple in current has been alleviated effectively. It is important to note that the SFH of the SLL converter is primarily responsible for the presence of the finite quantity of CCR that persists throughout the commutation (indicted by a circle (‘O’)). Phenomenal currents are irrelevant to the ideal rectangular waveform that is required due to the CCR caused by destabilized DC-link voltage of SLL converter. In order to obtain a stabilized DC -link voltage the NF is applied to the SLL converter. Fig. 19 depicts the experimental current results for light, moderate and full load conditions of the BLDCM assisting by NF filter based SLL converter. It is important to observe that theoretical and practical results are quite similar to one another, which demonstrates the effectiveness of the suggested control approach. At the various loading conditions, a comparative result for percentage of the CCR is presented in Fig. 20. This comparative assessment has been subjected to standard BLDCM , destabilized DC-link and stabilized DC-link converter-based drive. In addition, a comprehensive comparison evaluation of th e designed topology of the BLDCM drive against various existing topologies and control approaches from a variety of perspectives is required in order show the benefits of the suggested control strategy. Table 3 illustrates a tabular representation of the comparative study. The observation is drawn that the proposed system offers the lowest possible CCRs. It is clearly noted that the suggested strategy possess just 6.2% of CCRs that is minimum over other mentioned control techniques. The proposed system offers least CCR that confirms the diverse functionality of the recommended filter approach. (a) (b) (c) Fig. 19. Practical findings of current for SLL converter assist BLDC motor with filter. (a) light load (b) moderate load (c) full laod Fig. 20. A comparative result for percentage of the CCR. Table 3: Comprehensive investigation of contemporary strategies for CCR minimization Control Approach Filter Scheme DC-DC Converter CCR (%) Adaptive sliding mode [17] No No 34.3 Sliding Mode approach [19] No No 32.2 Kalman filter approach [18] No No 33.1 anticipatory current control [21] No No 32.6 pulse-width modulation [7] No No 32.1 Current-Loop adaptation [20] No Buck converter 27.8 PWM control [17] No Cuk converter 18.7 Multi-level inverter [11] No SEPIC 16.4 Proposed approach Yes/NF SLL converter 6.2 6. CONCLUSION In this study, an investigation is conducted to analyze the impact of SFH overlaid to output of SLL converter. The current and torque ripples further eva luated during the commutation. This is done in addition to the previous point. The NF approach has been developed to reduce the greater -level SFH superimposed with output of SLL converter. In addition, the SLL converter has applied in order to get a regulated dc-link. The suggested filter is tested at switching frequency of SLL converter. Based on the findings, it can be concluded that the present ripple has effectively reduced. It is estimated that the CCR that is found in the traditional BLDC M drive is roughly 45.8%, 43.7%, and 43.5% for circumstances that are light, moderate and full load, respectively. Likewise, the amount of CCR for SLL converter-based BLDCM drive corresponds to 16.5%, 14%, and 13% when not subjected to NF filter. However, when compensation is applie d, the percentages reduce to 7.2%, 7%, and 6.2%, respectively. When it comes to reducing the amount of current ripple that occurs during the CI, the presented control approach for the SLL converter is more efficient. The suggested method can be applied to various home appliances as well as industrial applications where highly precise motor is required.

1. K. Xia, Y. Ye, J. Ni, Y. Wang and P. Xu, "Model Predictive Control Method of Torque Ripple Reduction for BLDC Motor," in IEEE Transactions on Magnetics, vol. 56, no. 1, pp. 1 -6, Jan. 2020, Art no. 7503106. 2. Naresh, K., Suresh, D., & Pattnaik, S. (2025). High‐Voltage Gain DC –DC Converter for Sensorless Speed Control of BLDC Motor. International Journal of Circuit Theory and Applications. 3. Jagtap, N., & Pattnaik, S. (2025). Dual‐Input High Gain Step‐Up Converter for BLDC Motor Applications in Renewable Energy. International Journal of Circuit Theory and Applications. 4. Chaudhari, Parag Suresh, and Sanjaykumar L. Patil. "Reduction in harmonics of BLDC motor drive using controlled LC filter." Electric Power Components and Systems 46, no. 14-15 (2018): 1686-1703. 5. Dutta, A., Das, S. S., & Panda, A. (2024). A novel UEBASC‐qZSI‐based BLDC motor drive with modified DTC for EV application. International Journal of Circuit Theory and Applications, 52(10), 5184-5204. 6. Cheshmeh Beigi, H. M. (2017). An adaptive nonlinear internal -model control for the speed control of homopolar salient -pole BLDC motor. International Journal of Electronics, 105(5), 848–865. 7. Nagarajan, L., Bhoopathi Mariappan, P. A. Gowrisankar, and J. Karthikeyan. "Analysis and Controlling of Uncertainty in BLDC Motor Using Optimal Hybrid Algorithm in State Space." Electric Power Components and Systems (2023): 1-16. 8. Gopan, K. V., & Shree, J. D. (2022). Analysis of bridge type and bridgeless cuk converter fed motor drives for improved power quality. International Journal of Electronics, 110(7), 1179–1212. 9. Genc, Naci, and Zhansaya Seidakhanovna Kalimbetova. "Cuckoo Optimization Algorithm Based Fuzzy Logic Speed Controller for BLDC Motor." Electric Power Components and Systems 52, no. 11 (2024): 2065-2077. 10. Kivanc, O. C., & Ustun, O. (2019). Dynamic control of electronic differential in the field weakening region. International Journal of Electronics, 106(10), 1583–1601. 11. A. Khazaee, H. A. Zarchi, G. A. Markadeh and H. Mosaddegh Hesar, "MTPA Strategy for Direct Torque Control of Brushless DC Motor Drive," in IEEE Transactions on Industrial Electronics , vol. 68, no. 8, pp. 6692-6700, Aug. 2021 12. V. Viswanathan and J. Seenithangom, "Commutation Torque Ripple Reduction in the BLDC Motor Using Modified SEPIC and Three -Level NPC Inverter," in IEEE Transactions on Power Electronics, vol. 33, no. 1, pp. 535-546, Jan. 2018, doi: 10.1109/TPEL.2017.2671400. 13. R. Heidari and J. -W. Ahn, "Torque Ripple Reduction of BLDC Motor with a Low -Cost Fast- Response Direct DC-Link Current Control," in IEEE Transactions on Industrial Electronics, vol. 71, no. 1, pp. 150-159, Jan. 2024. 14. U. K. Kalla, S. P. S. Rajawat, S. Singh and N. Bhati, "A High -Gain Single -Switch DC -DC Converter Based Speed Control of PMBLDCM Drive for EV Applications," 2021 IEEE 2nd International Conference on Smart Technologies for Power, Energy and Control (STPEC) , Bilaspur, Chhattisgarh, India, 2021, pp. 1-6 15. Joshi, D., Deb, D., Giri, A. K., & Colak, I. (2024). Efficient Input‐Output Feedback Linearizing Control of BLDC Drives. International Journal of Circuit Theory and Applications. 16. S. Arunkumar, N. Mohana Sundaram, R. Thottungal and A. Shreya, "A Bridge Type DC -DC Converter fed BLDC Motor Drive for Household Appliances," 2021 International Conference on Advancements in Electrical, Electronics, Communication, Computing and Automation (ICAECA), Coimbatore, India, 2021, pp. 1-4, 17. Wei Chen, Changliang Xia and Mei Xue, "A torque ripple suppression circuit for brushless DC motors based on power DC/DC converters," 2008 3rd IEEE Conference on Industrial Electronics and Applications, Singapore, 2008, pp. 1453-1457, 18. R. A. Raj, D. S. Nair, S. MP and S. George, "A Novel High -Voltage Gain Circuit Topology for Commutation Torque Ripple Reduction," in IEEE Transactions on Industry Applications, vol. 58, no. 5, pp. 6227-6236, Sept.-Oct. 2022. 19. K. Guo, M. Yang, X. Li, P. Shi and P. Wang, "Research on a New Adaptive Integral Sliding Mode Controller Based on a Small BLDC," in IEEE Access, vol. 10, pp. 73204-73213, 2022, 20. Zhu, Changlin, et al. "Global fast terminal sliding mode control strategy for permanent magnet synchronous motor based on load torque Luenberger observer." IEICE Electronics Express 18.19 (2021): 21. Lim, Sun, Seok-Kyoon Kim, and Yonghun Kim. "Active damping injection output voltage control with dynamic current cut -off frequency for DC/DC buck converters." Energies 14.20 (2021): 6848. 22. J. K. Park, J. -H. Lee, S. -K. Kim and C. K. Ahn, "Output -Feedback Speed-Tracking Control Without Current Feedback for BLDCMs Based on Active -Damping and Invariant Surface Approach," in IEEE Transactions on Circuits and Systems II: Express Briefs , vol. 68, no. 7, pp. 2528-2532, July 2021 23. Liang, Yang, Deliang Liang, Yibin Liu, Mingkang Zhang, Peixin Jia, and Guanhua Sun. "Quasi proportional‐resonant control for the grid current of hybrid distribution transformer." The Journal of Engineering 2019, no. 16 (2019): 1926-1929. 24. Ye, Tao, NingYi Dai, Chi -Seng Lam, Man -Chung Wong, and Josep M. Guerrero. "Analysis, design, and implementation of a quasi -proportional-resonant controller for a multifunctional capacitive-coupling grid-connected inverter." IEEE Transactions on Industry Applications 52, no. 5 (2016): 4269-4280. 25. Herman, Leopold, Igor Papic, and Bostjan Blazic. "A proportional-resonant current controller for selective harmonic compensation in a hybrid active power filter." IEEE transactions on power delivery 29, no. 5 (2014): 2055-2065. 26. Bist V, Singh B. A brushless DC motor drive with power factor correction using isolated -zeta converter. IEEE Trans Ind Inform. 2014; 10 (4):2064-2072. 27. Vosoughi N, Maysam A, Abbasi E, Sabahi M. A Zeta-based switched-capacitor DC-DC converter topology. Int J Circuit Theory Appl. 2019; 47(8):1302-1322. 28. Vijay KO, Sriramalakshmi P. Comparison between zeta converter and boost converter using sliding mode controller. Int J Eng Res Technol. 2016; 5(7):368–373. 29. Singh S, Singh B, Bhuvaneswari G, Bist V. Power factor corrected zeta converter based improved power quality switched mode power supply. IEEE Trans Ind Electron. 2015; 62(9):0278-0046.