Non-Linear Neuro Control for Active Steering for Various Road Conditions

To develop a front steering control system, a nonlinear steering control law is necessary, since the dynamical characteristics of cars are nonlinear at the situation of large side slip angle and high yaw rate. The robustness to the change of friction coefficient μ is required on the controller. In this study, an intelligent control of front wheel steering is presented. The controller consists of 2 neural network controllers for specific values of μ and an integrator under the Cubic Neural Network (CNN) architecture. The neural networks are designed with error back propagation learning. By using the CNN architecture, the controller can adapt to various values of μ. The effectiveness and the feasibility of the present active steering method are demonstrated by numerical simulations using simple 8DOF model and DADS full vehicle model.