Data-driven friction modeling and compensation for rotary servo actuators

Abstract

This study proposes a data-driven friction modeling and compensation method aimed at solving the problem of servo performance degradation caused by friction in rotary servo actuators. First, a data-driven friction modeling method is proposed on the basis of the physics-informed neural network (PINN) and the LuGre model. The constructed friction model consists of sliding regime, static regime, and presliding regime, which extends the variables of the friction model to include velocity and position. The data-driven friction model not only retains the accuracy of the LuGre model in describing the dynamic behavior of friction at zero velocity but also improves the accuracy and convergence speed of the model through the powerful learning ability of PINN, which is verified in the two examples of constructing friction test data. Second, on the basis of the data-driven friction model, a composite compensation strategy centered on friction compensation is proposed. The friction compensator is used to compensate the internal friction of the actuator, and the extended Kalman filter is used to suppress the random disturbance to achieve the precise control of the servo actuator. Experimental validation of the proposed compensation strategy against three traditional control methods demonstrates its superiority, with average improvements of 49.5%, 30.4%, and 32.7% in velocity tracking accuracy, respectively, while ensuring consistent accuracy across different positions. The proposed data-driven friction modeling and compensation method provides a new perspective and method for overcoming the effect of friction.

Graphical abstract

Keywords

rotary servo actuator / physics-informed neural network / LuGre model / data-driven friction model / friction compensation

References

[1] Liu Y L, Chen X C, Yu Z G, Yu H, Meng L B, Yokoi H. High-precision dynamic torque control of high stiffness actuator for humanoids. ISA Transactions, 2023, 141: 401–413

[2] Yuan S S, Deng W X, Liang X L, Yao J Y, Yang G L. Nonlinear robust adaptive precision motion control of motor servo systems with unknown actuator backlash compensation. ISA Transactions, 2023, 137: 349–357

[3] Wolf S, Grioli G, Eiberger O, Friedl W, Grebenstein M, Höppner H, Burdet E, Caldwell D G, Carloni R, Catalano M G, Lefeber D, Stramigioli S, Tsagarakis N, Van Damme M, Van Ham R, Vanderborght B, Visser L C, Bicchi A, Albu-Schäffer A. Variable stiffness actuators: review on design and components. IEEE/ASME Transactions on Mechatronics, 2016, 21(5): 2418–2430

[4] Wensing P M, Wang A, Seok S, Otten D, Lang J, Kim S. Proprioceptive actuator design in the MIT cheetah: impact mitigation and high-bandwidth physical interaction for dynamic legged robots. IEEE Transactions on Robotics, 2017, 33(3): 509–522

[5] Zheng Y Q, Zhang G S. Quantized adaptive tracking control for nonlinear systems with actuator backlash compensation. Journal of the Franklin Institute, 2019, 356(15): 8484–8506

[6] Asignacion A, Haninger K, Oh S, Lee H. High-stiffness control of series elastic actuators using a noise reduction disturbance observer. IEEE Transactions on Industrial Electronics, 2022, 69(8): 8212–8219

[7] Liu X, Li Y G, Cheng Y H, Cai Y. Sparse identification for ball-screw drives considering position-dependent dynamics and nonlinear friction. Robotics and Computer-Integrated Manufacturing, 2023, 81: 102486

[8] Tu X, Zhou Y F, Zhao P, Cheng X. Modeling the static friction in a robot joint by genetically optimized BP neural network. Journal of Intelligent & Robotic Systems, 2019, 94(1): 29–41

[9] Yin N, Xing Z G, He K, Zhang Z N. Tribo-informatics approaches in tribology research: a review. Friction, 2023, 11(1): 1–22

[10] Bo L C, Pavelescu D. The friction-speed relation and its influence on the critical velocity of stick-slip motion. Wear, 1982, 82(3): 277–289

[11] Canudas de WitC, OlssonH, AstromK J, LischinskyP. Dynamic friction models and control design. In: Proceedings of 1993 American Control Conference. San Francisco: IEEE, 1993, 1920–1926

[12] Canudas de Wit C, Olsson H, Astrom K J, Lischinsky P. A new model for control of systems with friction. IEEE Transactions on Automatic Control, 1995, 40(3): 419–425

[13] Huang S N, Liang W Y, Tan K K. Intelligent friction compensation: a review. IEEE/ASME Transactions on Mechatronics, 2019, 24(4): 1763–1774

[14] Marton L, Lantos B. Modeling, identification, and compensation of stick-slip friction. IEEE Transactions on Industrial Electronics, 2007, 54(1): 511–521

[15] BruntonS L, Kutz J N. Data-Driven Science and Engineering: Machine Learning, Dynamical Systems, and Control. Cambridge: Cambridge University Press, 2019, 161–170

[16] Peng H J, Song N N, Li F, Tang S. A mechanistic-based data-driven approach for general friction modeling in complex mechanical system. Journal of Applied Mechanics, 2022, 89(7): 071005

[17] TrinhM, Schwiedernoch R, GründelL, StormsS, Brecher C. Friction modeling for structured learning of robot dynamics. In: Liewald M, Verl A, Bauernhansl T, Möhring H, eds. Production at the Leading Edge of Technology. Cham: Springer International Publishing, 2023, 396–406

[18] Baş H, Karabacak Y E. Machine learning-based prediction of friction torque and friction coefficient in statically loaded radial journal bearings. Tribology International, 2023, 186: 108592

[19] Baş H, Karabacak Y E. Triboinformatic modeling of the friction force and friction coefficient in a cam-follower contact using machine learning algorithms. Tribology International, 2023, 181: 108336

[20] Olejnik P, Ayankoso S. Friction modelling and the use of a physics-informed neural network for estimating frictional torque characteristics. Meccanica, 2023, 58(10): 1885–1908

[21] Raissi M, Perdikaris P, Karniadakis G E. Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 2019, 378: 686–707

[22] Xia Y C, Meng Y G. Physics-informed neural network (PINN) for solving frictional contact temperature and inversely evaluating relevant input parameters. Lubricants, 2024, 12(2): 62

[23] Zhai W D, Tao D W, Bao Y Q. Parameter estimation and modeling of nonlinear dynamical systems based on Runge–Kutta physics-informed neural network. Nonlinear Dynamics, 2023, 111(22): 21117–21130

[24] Bolandi H, Sreekumar G, Li X Y, Lajnef N, Boddeti V N. Physics informed neural network for dynamic stress prediction. Applied Intelligence, 2023, 53(22): 26313–26328

[25] Yang X Y, Du Y X, Li L H, Zhou Z X, Zhang X P. Physics-informed neural network for model prediction and dynamics parameter identification of collaborative robot joints. IEEE Robotics and Automation Letters, 2023, 8(12): 8462–8469

[26] Gu W B, Primatesta S, Rizzo A. Physics-informed neural network for quadrotor dynamical modeling. Robotics and Autonomous Systems, 2024, 171: 104569

[27] Wang C, Peng J D, Pan J F. A novel friction compensation method based on stribeck model with fuzzy filter for PMSM servo systems. IEEE Transactions on Industrial Electronics, 2023, 70(12): 12124–12133

[28] Feng H, Qiao W H, Yin C B, Yu H F, Cao D H. Identification and compensation of non-linear friction for a electro-hydraulic system. Mechanism and Machine Theory, 2019, 141: 1–13

[29] Thenozhi S, Sánchez A C, Rodríguez-Reséndiz J. A contraction theory-based tracking control design with friction identification and compensation. IEEE Transactions on Industrial Electronics, 2022, 69(6): 6111–6120

[30] Wan M, Dai J, Zhang W H, Xiao Q B, Qin X B. Adaptive feed-forward friction compensation through developing an asymmetrical dynamic friction model. Mechanism and Machine Theory, 2022, 170: 104691

[31] Wu Y F, Wang Z H, Li Y Y, Chen W, Du R H, Chen Q W. Characteristic modeling and control of servo systems with backlash and friction. Mathematical Problems in Engineering, 2014, 2014(1): 328450

[32] Marques F, Woliński Ł, Wojtyra M, Flores P, Lankarani H M. An investigation of a novel LuGre-based friction force model. Mechanism and Machine Theory, 2021, 166: 104493

[33] Feng H, Yin C B, Cao D H. Trajectory tracking of an electro-hydraulic servo system with an new friction model-based compensation. IEEE/ASME Transactions on Mechatronics, 2023, 28(1): 473–482

[34] Liu Y D, Feng X Y, Li P G, Li Y F, Su Z, Liu H Y, Lu Z T, Yao M. Modeling, identification, and compensation control of friction for a novel dual-drive hydrostatic lead screw micro-feed system. Machines, 2022, 10(10): 914

[35] Yu H T, Gao H B, Deng H C, Yuan S, Zhang L X. Synchronization control with adaptive friction compensation of treadmill-based testing apparatus for wheeled planetary rover. IEEE Transactions on Industrial Electronics, 2022, 69(1): 592–603

[36] Zhang W J, Li M Y, Gao Y P, Chen Y Q. Periodic adaptive learning control of PMSM servo system with LuGre model-based friction compensation. Mechanism and Machine Theory, 2022, 167: 104561

[37] Hu X Y, Han S J, Liu Y Y, Wang H R. Two-axis optoelectronic stabilized platform based on active disturbance rejection controller with LuGre friction model. Electronics, 2023, 12(5): 1261

[38] Yao J Y, Deng W X, Jiao Z X. Adaptive control of hydraulic actuators with LuGre model-based friction compensation. IEEE Transactions on Industrial Electronics, 2015, 62(10): 6469–6477

[39] Li Z Q, Zhou Q K, Zhang Z Y, Zhang L C, Fan D P. Prestiction friction compensation in direct-drive mechatronics systems. Journal of Central South University, 2013, 20(11): 3031–3041

[40] Jiang X L, Fan D P, Fan S X, Xie X, Chen N. High-precision gyro-stabilized control of a gear-driven platform with a floating gear tension device. Frontiers of Mechanical Engineering, 2021, 16(3): 487–503

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grant Nos. 52305079 and U19A2072).

Conflict of Interest

Dapeng FAN is a member of the Editorial Board of Frontiers of Mechanical Engineering, who was excluded from the peer review process and all editorial decisions related to the acceptance and publication of this article. Peer review was handled independently by the other editors to minimize bias.

RIGHTS & PERMISSIONS

2024 Higher Education Press