Abstract: In order to improve the dynamic performance and optimization control of active suspension, the radial basis function neural network algorithm was applied in inversion of sliding mode control strategy. The C-class random road excitation and 2 degrees with 1/4 vehicle suspension system model were established, and the equation of state of motion of active suspension system was established. The evaluation indexes of vehicle vertical acceleration, suspension dynamic deflection and tire dynamic displacement were studied, the RBF inversion of sliding mode control strategy was designed. The 2 degrees with 1/4 vehicle suspension system model and the RBF inversion of sliding mode control strategy were verified using the Matlab/CARSIM software. The results show that under the condition of C-class random road excitation, this control strategy reduced the acceleration of car body by 34.7%, the dynamic deflection of suspension by 28%, and the dynamic displacement of tire by 28.7%. It improves the dynamic performance and control performance of the suspension.
Key words: control strategy; inversion sliding mode; radial basis function（RBF）neural network; active suspension; Matlab/CARSIM