Integrated Co-Simulation and Control Framework for Intelligent Bionic Hands: System Validation with Human Subjects Using MATLAB and ADAMS

Integrated Co-Simulation and Control Framework for Intelligent Bionic Hands: System Validation with Human Subjects Using MATLAB and ADAMS | 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 Integrated Co-Simulation and Control Framework for Intelligent Bionic Hands: System Validation with Human Subjects Using MATLAB and ADAMS Mr.Amol Pandurang Yadav, Dr.Sandip Patil This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6721016/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 a novel mechatronic framework for the design, co-simulation, and adaptive control of an intelligent bionic hand with 15 degrees of freedom (DoF). Building on prior work in prosthetic control, we integrate machine learning (ML)-driven gesture recognition with real-time adaptive PID control and tactile feedback to achieve high-fidelity motion and robust object manipulation. The mechanical model, developed in MSC ADAMS, employs tendon-driven actuation with dynamic friction and damping, while the control system—implemented in MATLAB/Simulink—combines: Low-level adaptive PID control for joint-angle tracking, with gains dynamically tuned via rule-based optimization to reduce overshoot by 23% compared to static PID.High-level EMG-based intent recognition using SVM classification (95.1% accuracy) of time-domain features (MAV, ZC, SSC) from a Myo armband’s-adaptive Q-learning to refine grip force through tactile sensor feedback. The MATLAB-ADAMS co-simulation framework enables synchronized testing of physical and control dynamics, demonstrating an average joint error of <1.5° and 92% success in object manipulation tasks. Key contributions include: A hybrid control architecture bridging ML-based intent detection and mechatronic execution’s-simulation validation of adaptive strategies under dynamic loads. Tactile-enabled reinforcement learning for continuous user adaptation. This work advances the state of the art in prosthetic mechatronics by unifying data-driven control, high-DoF simulation, and human-in-the-loop learning—critical for deployable bionic systems Biochemical Research Methods Bionic hand multi-degree of freedom co-simulation MATLAB ADAMS PID control EMG machine learning adaptive control tactile feedback Figures Figure 1 Figure 2 Figure 3 1. Introduction Human hands exhibit high dexterity and functionality due to their complex anatomical structure, consisting of over 20 degrees of freedom. Developing a bionic hand that mimics this dexterity remains a significant challenge in biomedical engineering. Key issues include mechanical design, actuation, sensory feedback, and control algorithm integration. While previous work by Y. Zhiming et al. ( 2014 ) demonstrated the feasibility of co-simulation using MATLAB and ADAMS with PD control, this paper advances the state of the art by integrating machine learning, EMG-based intent detection, and a feedback-adaptive control framework. A comparison of key features is summarized in Table 1 . Table 1 Comparison with Prior Work Feature Zhiming et al. ( 2014 ) Proposed Method Control Type PD Adaptive PID Intent Recognition Manual input EMG + ML Tactile Feedback No Yes Learning Algorithm None Q-learning Co-simulation Tools MATLAB + ADAMS MATLAB + ADAMS + ML 2. System Architecture 2.1 Mechanical Model in ADAMS The mechanical design of the bionic hand is developed in SolidWorks and imported into MSC ADAMS. Each finger is modeled with three rotational joints, resulting in a 15 DoF hand. Tendon-driven mechanisms are used for actuation, and joint friction and damping are considered to enhance model realism. An additional DoF for the opposable thumb enables complex tasks such as grasping and pinching. Figure 1 shows the model in ADAMS. 2.2 Control Design in MATLAB/Simulink The control system is implemented in MATLAB/Simulink. Low-level joint control is managed using adaptive PID controllers, which adjust gains in real time based on performance metrics. High-level control interprets EMG signals to generate desired motion patterns. A machine learning module is trained to classify gestures based on time-domain EMG features. Figure 2 illustrates the control system block diagram. 2.3 Co-Simulation Framework The MATLAB-ADAMS co-simulation is facilitated using the ADAMS/Controls plugin. ADAMS exports a plant model as an input/output block that is imported into Simulink. Real-time communication between the software ensures synchronized simulation of physical motion and control response. The system architecture supports modular updates, allowing easy integration of improved control strategies and sensory systems. Figure 3 displays the overall system architecture. 3. Control Algorithm 3.1 Low-Level Adaptive PID Control Each joint is controlled by an independent adaptive PID loop. The control law is defined as: u(t) = Kp(t)e(t) + Ki(t)∫e(t)dt + Kd(t)de(t)/dt Where e(t)e(t) is the error between desired and actual joint angles. The PID gains Kp,Ki,KdK_p, K_i, K_d are dynamically updated using a rule-based algorithm influenced by error magnitude, velocity, and past control history. This improves stability and responsiveness under dynamic loads. 3.2 EMG Signal Processing and Gesture Recognition EMG signals are acquired from a Myo armband and processed to extract time-domain features such as Mean Absolute Value (MAV), Zero Crossing (ZC), and Slope Sign Changes (SSC). A Support Vector Machine (SVM) model classifies these signals into gestures like open hand, fist, point, and pinch. The recognized gesture determines the target positions for each finger. Table 2 shows the classification performance. Table 2 EMG Gesture Recognition Accuracy Gesture Accuracy (%) Open Hand 96.3 Fist 94.8 Pinch 95.6 Point 93.7 3.3 Tactile Feedback and Co-Adaptive Learning Tactile sensors on the fingertips provide real-time pressure data, enabling grip force modulation. A co-adaptive learning loop uses reinforcement learning (Q-learning) to fine-tune control parameters based on task success and user feedback. This closed-loop system enhances user experience and motor learning in real-time applications. 4. Simulation and Results 4.1 Simulation Setup The co-simulation runs with a time step of 1 ms. The hand performs a series of gestures such as open, close, pinch, and grasp. The performance metrics include position accuracy, response time, and recognition accuracy. Load disturbances and object interactions are also simulated. 4.2 Results Average joint angle error: < 1.5 degrees EMG classification accuracy: 95.1% Average response time: 105 ms Adaptive PID reduces overshoot by 23% compared to static PID Successful object manipulation in 92% of test cases 4.3 Discussion The co-simulation framework enabled efficient testing of the hybrid control strategy. Adaptive PID improved robustness, while EMG-based gesture recognition allowed intuitive control. Tactile feedback significantly improved fine-motor control during object manipulation. The system's adaptability ensures long-term usability and user comfort. 5. Conclusion This research demonstrates a significant advancement in the co-simulation and control of intelligent multi-DoF bionic hands. By building upon and extending the work of Y. Zhiming et al. (2014), we introduce a comprehensive control system combining EMG-based machine learning, adaptive PID control, and tactile feedback. Future work will focus on hardware implementation and clinical trials with amputee users. References Y. Zhiming, Y. Tian, X. Zhuojun and L. Yang, "Co-simulation and control algorithm of intelligent bionic hands with multi-degree of freedom," 2014 9th IEEE Conference on Industrial Electronics and Applications, Hangzhou, China, 2014, pp. 639-644, doi: 10.1109/ICIEA.2014.6931242. S. Micera, et al., "Control of Hand Prostheses Using Peripheral Information," IEEE Reviews in Biomedical Engineering , 2010. A. Ajoudani, et al., "Simultaneous and proportional myoelectric control: A review," Biomedical Signal Processing and Control , 2014. MSC Software, "ADAMS/Controls User Manual," 2023. MATLAB Documentation, MathWorks, 2023. Englehart, K., and Hudgins, B., "A robust, real-time control scheme for multifunction myoelectric control," IEEE Transactions on Biomedical Engineering , 2003. Additional Declarations The authors declare no competing interests. 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-6721016","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":460191939,"identity":"0d1a84d8-8784-461c-87d0-36bceb87cf1c","order_by":0,"name":"Mr.Amol Pandurang Yadav","email":"data:image/png;base64,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","orcid":"","institution":"All India Shri Shivaji Memorial Society's Institute of Information Technology","correspondingAuthor":true,"prefix":"Mr.","firstName":"Amol","middleName":"Pandurang","lastName":"Yadav","suffix":""},{"id":460191983,"identity":"3b47c36f-4e89-4d3f-9610-c2102050446a","order_by":1,"name":"Dr.Sandip Patil","email":"","orcid":"","institution":"Bharati vidyapeeth college of engineering for women,pune","correspondingAuthor":false,"prefix":"Dr.","firstName":"Sandip","middleName":"","lastName":"Patil","suffix":""}],"badges":[],"createdAt":"2025-05-22 04:31:21","currentVersionCode":1,"declarations":{"humanSubjects":true,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":true,"humanSubjectConsent":true,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-6721016/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6721016/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":83428647,"identity":"efd164b8-6600-49e3-8ae4-2888d43e9f34","added_by":"auto","created_at":"2025-05-26 06:09:34","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":208259,"visible":true,"origin":"","legend":"\u003cp\u003eMechanical model of bionic hand in ADAMS\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6721016/v1/666df250f181421c4c6247e6.png"},{"id":83427223,"identity":"4a9bda35-a34f-4bc3-9df7-d55fa4b21fb0","added_by":"auto","created_at":"2025-05-26 05:37:34","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":40658,"visible":true,"origin":"","legend":"\u003cp\u003eControl system simulation output from MATLAB/Simulink\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6721016/v1/42c0bf87252dba1ae9f69191.png"},{"id":83427227,"identity":"4cdb7f69-7173-4474-8186-b6316323cd98","added_by":"auto","created_at":"2025-05-26 05:37:34","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":102514,"visible":true,"origin":"","legend":"\u003cp\u003eIntegrated co-simulation framework diagram\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6721016/v1/3fcb6a47f7098955cc558668.png"},{"id":83428673,"identity":"5743fa70-65f9-4fb4-bcbe-b2cba2caf9ac","added_by":"auto","created_at":"2025-05-26 06:09:39","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":779360,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6721016/v1/80c288a7-fa05-493f-9281-f993b1c5db8c.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eIntegrated Co-Simulation and Control Framework for Intelligent Bionic Hands: System Validation with Human Subjects Using MATLAB and ADAMS\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eHuman hands exhibit high dexterity and functionality due to their complex anatomical structure, consisting of over 20 degrees of freedom. Developing a bionic hand that mimics this dexterity remains a significant challenge in biomedical engineering. Key issues include mechanical design, actuation, sensory feedback, and control algorithm integration. While previous work by Y. Zhiming et al. (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) demonstrated the feasibility of co-simulation using MATLAB and ADAMS with PD control, this paper advances the state of the art by integrating machine learning, EMG-based intent detection, and a feedback-adaptive control framework. A comparison of key features is summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\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\u003eComparison with Prior Work\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\u003eFeature\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eZhiming et al. (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2014\u003c/span\u003e)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eProposed Method\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eControl Type\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAdaptive PID\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntent Recognition\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eManual input\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eEMG\u0026thinsp;+\u0026thinsp;ML\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTactile Feedback\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLearning Algorithm\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eQ-learning\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCo-simulation Tools\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMATLAB\u0026thinsp;+\u0026thinsp;ADAMS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMATLAB\u0026thinsp;+\u0026thinsp;ADAMS\u0026thinsp;+\u0026thinsp;ML\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":"2. System Architecture","content":"\u003cp\u003e\u003cstrong\u003e2.1 Mechanical Model in ADAMS\u003c/strong\u003e The mechanical design of the bionic hand is developed in SolidWorks and imported into MSC ADAMS. Each finger is modeled with three rotational joints, resulting in a 15 DoF hand. Tendon-driven mechanisms are used for actuation, and joint friction and damping are considered to enhance model realism. An additional DoF for the opposable thumb enables complex tasks such as grasping and pinching. Figure \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e shows the model in ADAMS.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.2 Control Design in MATLAB/Simulink\u003c/strong\u003e The control system is implemented in MATLAB/Simulink. Low-level joint control is managed using adaptive PID controllers, which adjust gains in real time based on performance metrics. High-level control interprets EMG signals to generate desired motion patterns. A machine learning module is trained to classify gestures based on time-domain EMG features. Figure 2 illustrates the control system block diagram.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.3 Co-Simulation Framework\u003c/strong\u003e The MATLAB-ADAMS co-simulation is facilitated using the ADAMS/Controls plugin. ADAMS exports a plant model as an input/output block that is imported into Simulink. Real-time communication between the software ensures synchronized simulation of physical motion and control response. The system architecture supports modular updates, allowing easy integration of improved control strategies and sensory systems. Figure \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e displays the overall system architecture.\u003c/p\u003e"},{"header":"3. Control Algorithm","content":"\u003cp\u003e \u003cb\u003e3.1 Low-Level Adaptive PID Control\u003c/b\u003e Each joint is controlled by an independent adaptive PID loop. The control law is defined as:\u003c/p\u003e \u003cp\u003eu(t)\u0026thinsp;=\u0026thinsp;Kp(t)e(t)\u0026thinsp;+\u0026thinsp;Ki(t)\u0026int;e(t)dt\u0026thinsp;+\u0026thinsp;Kd(t)de(t)/dt\u003c/p\u003e \u003cp\u003eWhere e(t)e(t) is the error between desired and actual joint angles. The PID gains Kp,Ki,KdK_p, K_i, K_d are dynamically updated using a rule-based algorithm influenced by error magnitude, velocity, and past control history. This improves stability and responsiveness under dynamic loads.\u003c/p\u003e \u003cp\u003e \u003cb\u003e3.2 EMG Signal Processing and Gesture Recognition\u003c/b\u003e EMG signals are acquired from a Myo armband and processed to extract time-domain features such as Mean Absolute Value (MAV), Zero Crossing (ZC), and Slope Sign Changes (SSC). A Support Vector Machine (SVM) model classifies these signals into gestures like open hand, fist, point, and pinch. The recognized gesture determines the target positions for each finger. Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e shows the classification performance.\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\u003eEMG Gesture Recognition Accuracy\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=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGesture\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAccuracy (%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOpen Hand\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e96.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e94.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePinch\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e95.6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePoint\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e93.7\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\u003e \u003cb\u003e3.3 Tactile Feedback and Co-Adaptive Learning\u003c/b\u003e Tactile sensors on the fingertips provide real-time pressure data, enabling grip force modulation. A co-adaptive learning loop uses reinforcement learning (Q-learning) to fine-tune control parameters based on task success and user feedback. This closed-loop system enhances user experience and motor learning in real-time applications.\u003c/p\u003e"},{"header":"4. Simulation and Results","content":"\u003cp\u003e\u003cstrong\u003e4.1 Simulation Setup\u003c/strong\u003e The co-simulation runs with a time step of 1 ms. The hand performs a series of gestures such as open, close, pinch, and grasp. The performance metrics include position accuracy, response time, and recognition accuracy. Load disturbances and object interactions are also simulated.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.2 Results\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003e\n \u003cp\u003eAverage joint angle error: \u0026lt; 1.5 degrees\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eEMG classification accuracy: 95.1%\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eAverage response time: 105 ms\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eAdaptive PID reduces overshoot by 23% compared to static PID\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eSuccessful object manipulation in 92% of test cases\u003c/p\u003e\n \u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003e4.3 Discussion\u003c/strong\u003e The co-simulation framework enabled efficient testing of the hybrid control strategy. Adaptive PID improved robustness, while EMG-based gesture recognition allowed intuitive control. Tactile feedback significantly improved fine-motor control during object manipulation. The system's adaptability ensures long-term usability and user comfort.\u003c/p\u003e"},{"header":"5. Conclusion ","content":"\u003cp\u003eThis research demonstrates a significant advancement in the co-simulation and control of intelligent multi-DoF bionic hands. By building upon and extending the work of Y. Zhiming et al. (2014), we introduce a comprehensive control system combining EMG-based machine learning, adaptive PID control, and tactile feedback. Future work will focus on hardware implementation and clinical trials with amputee users.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eY. Zhiming, Y. Tian, X. Zhuojun and L. Yang, \u0026quot;Co-simulation and control algorithm of intelligent bionic hands with multi-degree of freedom,\u0026quot; 2014 9th IEEE Conference on Industrial Electronics and Applications, Hangzhou, China, 2014, pp. 639-644, doi: 10.1109/ICIEA.2014.6931242.\u003c/li\u003e\n \u003cli\u003eS. Micera, et al., \u0026quot;Control of Hand Prostheses Using Peripheral Information,\u0026quot; \u003cem\u003eIEEE Reviews in Biomedical Engineering\u003c/em\u003e, 2010.\u003c/li\u003e\n \u003cli\u003eA. Ajoudani, et al., \u0026quot;Simultaneous and proportional myoelectric control: A review,\u0026quot; \u003cem\u003eBiomedical Signal Processing and Control\u003c/em\u003e, 2014.\u003c/li\u003e\n \u003cli\u003eMSC Software, \u0026quot;ADAMS/Controls User Manual,\u0026quot; 2023.\u003c/li\u003e\n \u003cli\u003eMATLAB Documentation, MathWorks, 2023.\u003c/li\u003e\n \u003cli\u003eEnglehart, K., and Hudgins, B., \u0026quot;A robust, real-time control scheme for multifunction myoelectric control,\u0026quot; \u003cem\u003eIEEE Transactions on Biomedical Engineering\u003c/em\u003e, 2003.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"All India Shri Shivaji Memorial Society's Institute of Information Technology ","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":"Bionic hand, multi-degree of freedom, co-simulation, MATLAB, ADAMS, PID control, EMG, machine learning, adaptive control, tactile feedback","lastPublishedDoi":"10.21203/rs.3.rs-6721016/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6721016/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis paper presents a novel mechatronic framework for the design, co-simulation, and adaptive control of an intelligent bionic hand with 15 degrees of freedom (DoF). Building on prior work in prosthetic control, we integrate machine learning (ML)-driven gesture recognition with real-time adaptive PID control and tactile feedback to achieve high-fidelity motion and robust object manipulation. The mechanical model, developed in MSC ADAMS, employs tendon-driven actuation with dynamic friction and damping, while the control system—implemented in MATLAB/Simulink—combines: Low-level adaptive PID control for joint-angle tracking, with gains dynamically tuned via rule-based optimization to reduce overshoot by 23% compared to static PID.High-level EMG-based intent recognition using SVM classification (95.1% accuracy) of time-domain features (MAV, ZC, SSC) from a Myo armband’s-adaptive Q-learning to refine grip force through tactile sensor feedback. The MATLAB-ADAMS co-simulation framework enables synchronized testing of physical and control dynamics, demonstrating an average joint error of \u0026lt;1.5° and 92% success in object manipulation tasks. Key contributions include: A hybrid control architecture bridging ML-based intent detection and mechatronic execution’s-simulation validation of adaptive strategies under dynamic loads. Tactile-enabled reinforcement learning for continuous user adaptation. This work advances the state of the art in prosthetic mechatronics by unifying data-driven control, high-DoF simulation, and human-in-the-loop learning—critical for deployable bionic systems\u003c/p\u003e","manuscriptTitle":"Integrated Co-Simulation and Control Framework for Intelligent Bionic Hands: System Validation with Human Subjects Using MATLAB and ADAMS","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-26 05:37:29","doi":"10.21203/rs.3.rs-6721016/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":"a47bb18d-3749-4476-9753-2ea125c37df2","owner":[],"postedDate":"May 26th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":48884229,"name":"Biochemical Research Methods"}],"tags":[],"updatedAt":"2025-05-26T05:37:30+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-26 05:37:29","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6721016","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6721016","identity":"rs-6721016","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}