Development of Machine Learning Based Control System for Vehicle Active Suspension System

In this paper, Gaussian process (GP) algorithm, which is one of the machine learning methods, is designed to control the vehicle active suspension system (VASS). Experimental data were trained by supervised learning method (regression method). The data were obtained from an optimal linear quadratic controller tuned based on a full state feedback optimal control approach. The results demonstrated that the proposed machine learning (ML) based ground-penetrating radar (GPR) controller outperforms the optimal controller under uncertainties in terms of reducing the oscillation in sprung mass position with a 15% and 21.64% reduction for square and random road conditions, respectively.

PDF

___

[1] L. Sitnik, M. Magdziak-Tokłowicz, R. Wróbel, and P. Kardasz, “Vehicle Vibration in Human Health,” J. KONES. Powertrain Transp., vol. 20, no. 4, pp. 411–418, 2015, doi: 10.5604/12314005.1137854.
[2] Y. Shahid and M. Wei, “Comparative analysis of different model-based controllers using active vehicle suspension system,” Algorithms, vol. 13, no. 1, Jan. 2020, doi: 10.3390/a13010010.
[3] J. Watton, K. M. Holford, and P. Surawattanawan, “The application of a programmable servo controller to state control of an electrohydraulic active suspension,” Proc. Inst. Mech. Eng. Part D J. Automob. Eng., vol. 218, no. 12, pp. 1367–1377, Dec. 2004, doi: 10.1243/0954407042707650.
[4] H. Pang, X. Zhang, J. Yang, and Y. Shang, “Adaptive backstepping-based control design for uncertain nonlinear active suspension system with input delay,” Int. J. Robust Nonlinear Control, vol. 29, no. 16, pp. 5781–5800, Nov. 2019, doi: 10.1002/rnc.4695.
[5] L. Ovalle, H. Ríos, and H. Ahmed, “Robust Control for an Active Suspension System via Continuous Sliding-Mode Controllers,” Eng. Sci. Technol. an Int. J., 2021, doi: 10.1016/j.jestch.2021.06.006.
[6] W. Sun, H. Gao, and O. Kaynak, “Adaptive backstepping control for active suspension systems with hard constraints,” IEEE/ASME Trans. Mechatronics, vol. 18, no. 3, pp. 1072–1079, 2013, doi: 10.1109/TMECH.2012.2204765.
[7] N. Yagiz and Y. Hacioglu, “Backstepping control of a vehicle with active suspensions,” Control Eng. Pract., vol. 16, no. 12, pp. 1457–1467, Dec. 2008, doi: 10.1016/j.conengprac.2008.04.003.
[8] S. Kilicaslan, “Control of active suspension system considering nonlinear actuator dynamics,” Nonlinear Dyn., vol. 91, no. 2, pp. 1383–1394, Jan. 2018, doi: 10.1007/S11071-017-3951-X.
[9] I. Fialho and G. J. Balas, “Road adaptive active suspension design using linear parameter-varying gain- scheduling,” IEEE Trans. Control Syst. Technol., vol. 10, no. 1, pp. 43–54, Jan. 2002, doi: 10.1109/87.974337.
[10] L.-X. Guo, L.-P. Zhang, L.-X. Guo, and L.-P. Zhang, “Robust H∞ control of active vehicle suspension under non-stationary running,” JSV, vol. 331, no. 26, pp. 5824–5837, Dec. 2012, doi: 10.1016/J.JSV.2012.07.042.
[11] Y. Taskin, Y. Hacioglu, and N. Yagiz, “Experimental evaluation of a fuzzy logic controller on a quarter car test rig,” J. Brazilian Soc. Mech. Sci. Eng., vol. 39, no. 7, pp. 2433–2445, Jul. 2017, doi: 10.1007/s40430-016-0637-0.
[12] U. Rashid, M. Jamil, S. O. Gilani, and I. K. Niazi, “LQR based training of adaptive neuro-fuzzy controller,” in Smart Innovation, Systems and Technologies, 2016, vol. 54, pp. 311–322, doi: 10.1007/978-3-319-33747-0_31.
[13] S. J. Huang and W. C. Lin, “A neural network based sliding mode controller for active vehicle suspension,” Proc. Inst. Mech. Eng. Part D J. Automob. Eng., vol. 221, no. 11, pp. 1381–1397, Nov. 2007, doi: 10.1243/09544070JAUTO242.
[14] M. Heidari and H. Homaei, “Design a PID controller for suspension system by back propagation neural network,” J. Eng. (United Kingdom), vol. 2013, 2013, doi: 10.1155/2013/421543.
[15] S. L. Brunton and J. N. Kutz, “Data Driven Science & Engineering - Machine Learning, Dynamical Systems, and Control,” p. 572, 2017, Accessed: Sep. 06, 2021. [Online]. Available: databook.uw.edu.
[16] “Active Suspension - Quanser.” https://www.quanser.com/products/active-suspension/ (accessed Sep. 08, 2021).
[17] G. Kopsiaftis, E. Protopapadakis, A. Voulodimos, N. Doulamis, and A. Mantoglou, “Gaussian process regression tuned by Bayesian optimization for seawater intrusion prediction,” Comput. Intell. Neurosci., vol. 2019, 2019, doi: 10.1155/2019/2859429.
[18] J. Zhang, W. Li, L. Zeng, and L. Wu, “An adaptive Gaussian process-based method for efficient Bayesian experimental design in groundwater contaminant source identification problems,” Water Resour. Res., vol. 52, no. 8, pp. 5971–5984, Aug. 2016, doi: 10.1002/2016WR018598.
[19] X. Chen, Y. Tian, T. Zhang, and J. Gao, “Differential Evolution Based Manifold Gaussian Process Machine Learning for Microwave Filter’s Parameter Extraction,” IEEE Access, vol. 8, pp. 146450– 146462, 2020, doi: 10.1109/ACCESS.2020.3015043.