Search

To improve the robustness and adaptive capability of electronic stability control(ESC) ,the joint extended Kalman filtering technique is adopted to estimate tire coming stiffness with a 2 DOF model improved. On this basis, a joint control algorithm of linear quadratic regulator( LQR) and radial basis function neural network (RBFNN) PID control is proposed. A two-layer controller is designed, in which the upper layer uses LQR to get active yaw moment, while the lower layer controls the wheel torque allocation and variable slip ratio by RBFNN PID. 15 DOF vehicle model is used for simulation verification, and the results show that the control strategy designed better meets the control requirements of ESC and has good self-learning function and transient control capability. [ABSTRACT FROM AUTHOR]