Active disturbance rejection control based on BP neural network for suspension system of electromagnetic suspension vehicle

An electromagnetic suspension (EMS) vehicle achieves stable suspension through active feedback control of electromagnetic force. However, the vehicle’s suspension stability can be affected by various external disturbances. To improve the stability and robustness of the EMS vehicle’s suspension system, an active disturbance rejection control algorithm which is an adaptive parameter optimized by a backpropagation neural network (BP_ADRC) is proposed. Firstly, the dynamic model of the EMS vehicle’s suspension system is derived. Secondly, a nonlinear tracking differentiator and a nonlinear extended state observer are established. On this basis, a nonlinear state error feedback control law is designed, and the ADRC algorithm of the suspension system is constructed. Subsequently, an adaptive optimization algorithm using a BP neural network is given, and it can adaptively adjust the gain parameters of the nonlinear state error feedback control law according to the system changes. Finally, the effectiveness of the proposed BP_ ADRC algorithm is verified through simulations and experimental methods. The results indicate that the BP_ADRC algorithm exhibits good control performance under various disturbances, thereby effectively improving the stability of the EMS vehicle’s suspension system.