Neural network-based adaptive inverse fuzzy control for steer-by-wire system considering steering hysteresis

Realizing the steady tracking of steering angle for steer-by-wire (SBW) system is essential for the vehicle driving safety. However, tire torques act on the tracking control loop in the form of additional disturbances of SBW system, causing the transient output torque of steering motor to deviate from the expected value and thereby leading the steering hysteresis, which degrades the stability and accuracy of angle tracking. To solve this issue, a neural network-based adaptive fuzzy control method with hysteresis inverse compensation is proposed to enhance the tracking accuracy of steering angle and anti-disturbance ability for SBW system. Firstly, a hysteresis operator is constructed from the steering hysteresis dynamic model of SBW system, and an inverse compensator is designed using inverse multiplicative structure to dynamically compensate the control command of steering angle, which reduces the angle tracking hysteresis. Then, an adaptive fuzzy control method using general regression neural network is constructed. The fuzzy mapping is reconstructed into interpretable geometric parameters, which can be dynamically adjusted through a neural network-based parameter adaptive regulator to obtain the desired control characteristic of fuzzy controller, thereby optimizing the output torque of steering controller to enhance the anti-disturbance tracking performance for SBW system. Finally, the effectiveness of proposed method is verified on an experimental vehicle equipped with SBW system. Experimental results show that the proposed method effectively improve the tracking accuracy of steering angle, enhance the anti-disturbance ability and realize the steady angle tracking for SBW system.