EXPERIMENTAL EVALUATION OF CLASSICAL BACKSTEPPING AND RBF NEURAL NETWORK BASED BACKSTEPPING ON AN INVERTED PENDULUM SYSTEM
DOI:
https://doi.org/10.62985/j.huit_ojs.vol26.no2E.424Keywords:
Backstepping, RBF neural networks, nonlinear control, inverted pendulumAbstract
This paper presents an experimental comparison between the classical Backstepping controller and an improved Backstepping design incorporating Radial Basis Function (RBF) neural networks on an inverted pendulum system. While the conventional Backstepping method ensures stability based on a known mathematical model, its performance is sensitive to parameter uncertainties and unmodeled nonlinearities. The RBF-enhanced Backstepping controller addresses these limitations by compensating for unknown dynamics through real-time neural network approximation. Experimental results on an STM32F4 embedded platform indicate that the Backstepping-RBF controller achieves faster stabilization, reduced oscillations, and smoother control signals compared with the classical Backstepping method.
References
[1] F.-K. Tsai and J.-S. Lin, “Nonlinear control design of 360-degree inverted pendulum systems,” in Proc. 4th Int. Conf. Control Autom. (ICCA), Montreal, Canada, 2003, pp. 634–638, doi: https://doi.org/10.1109/ICCA.2003.1595099.
[2] L. B. Prasad, B. Tyagi, and H. O. Gupta, “Optimal control of nonlinear inverted pendulum system using PID controller and LQR: Performance analysis without and with disturbance input,” Int. J. Autom. Comput., vol. 11, no. 6, pp. 661–670, 2014, doi: https://doi.org/10.1007/s11633-014-0818-1
[3] T.-B. Dang et al., “PID control for cart and pole system: Simulation and experiment,” J. Fuzzy Syst. Control, vol. 2, no. 1, pp. 29–35, 2024, doi: https://doi.org/10.59247/jfsc.v2i1.165
[4] D.-P. Hoang, “A survey of experimental LQR for cart and pole,” J. Fuzzy Syst. Control, vol. 2, no. 2, pp. 97–103, 2024, doi: https://doi.org/10.59247/jfsc.v2i2.211
[5] N. X. Chiem and H. N. Phan, “Design controller of the quasi-time optimization approach for stabilizing and trajectory tracking of inverted pendulum,” MATEC Web Conf., vol. 226, p. 02007, 2018, doi: https://doi.org/10.1051/matecconf/201822602007
[6] S. Irfan, A. Mehmood, M. T. Razzaq, and J. Iqbal, “Advanced sliding mode control techniques for inverted pendulum: Modelling and simulation,” Eng. Sci. Technol. Int. J., vol. 21, no. 4, pp. 753–759, 2018, doi: https://doi.org/10.1016/j.jestch.2018.06.010
[7] M. Mahmoud, R. Saleh, and A. Ma’arif, “Stabilizing of inverted pendulum system using robust sliding mode control,” Int. J. Robot. Control Syst., vol. 2, no. 2, pp. 230–239, 2022, doi: https://doi.org/10.31763/ijrcs.v2i2.594
[8] H.-G.-B. Pham, “Trajectories tracking control for rotary inverted pendulum using backstepping method,” J. Fuzzy Syst. Control, vol. 3, no. 1, pp. 57–63, 2025, doi: https://doi.org/10.59247/jfsc.v3i1.276
[9] J. Liu, Radial Basis Function (RBF) Neural Network Control for Mechanical Systems. Berlin, Germany: Springer, 2013, doi: https://doi.org/10.1007/978-3-642-34816-7
[10] H. N. Phan and C. X. Nguyen, “Building embedded quasi-time-optimal controller for two-wheeled self-balancing robot,” MATEC Web Conf., vol. 132, p. 02005, 2017, doi: https://doi.org/10.1051/matecconf/201713202005
[11] H. V. Khuong, N. X. Chiem, and A. Obukhov, “Nonlinear control law design for inverted pendulum systems via RBF neural networks,” J. Fuzzy Syst. Control, vol. 3, no. 2, pp. 164–169, 2025, doi: https://doi.org/10.59247/jfsc.v3i2.314
[12] U. Yildiran, “Adaptive control of an inverted pendulum by a reinforcement learning-based LQR method,” Sakarya Univ. J. Sci., 2023, doi: https://doi.org/10.16984/saufenbilder.1286391
[13] S. Irfan, L. Zhao, S. Ullah, A. Mehmood, and M. F. Butt, “Control strategies for inverted pendulum: A comparative analysis of linear, nonlinear, and artificial intelligence approaches,” PLOS ONE, vol. 19, no. 3, Art. no. e0298093, 2024, doi: https://doi.org/10.1371/journal.pone.0298093
