Vehicle Chassis Decoupling Control Based on Neural Network Inverse Method

Abstract:

In order to eliminate the interference and coupling among automobile chassis electronic control subsystems, an integrated decoupling control approach based on neural network inverse method was proposed. The integration of active front steering (AFS), direct yaw moment control (DYC) and active suspension (ASS) was studied. According to Interactor algorithm, the reversibility of chassis system was analyzed and a BP neural network inverse system of the multivariable chassis system was obtained. Designing a close-loop controller and combining it with neural network inverse system, a compound controller was completed to improve dynamic performance. The simulation results showed that the proposed decoupling control approach could eliminate the interference and couple among automobile chassis electronic control subsystems, and improve the automobile handling and stability performance.