Kernel-based Approximate Dynamic Programming for Autonomous Vehicle Stability Control

Aiming at the stability control problem of the non-linear dynamic characteristics of vehicles, this paper proposes a kernel based dual heuristic programming(KD HP) intelligent vehicle stability control strategy, and establishes a two-degree-of-freedom vehicle model considering tire nonlinearity. The method of approximate linear correlation analysis(ALD) is used to sparse the sampled data, and a kernel dictionary is established. Carsim/Simulink joint simulation shows that under the double lane changing condition and the double lane emer-gency lane changing condition, the KDHP controller has better tracking accuracy than the dual heuristic programming(DHP) controller, and can better adapt to the nonlinear system.