The Study Of Four-Wheel Steering Simulation Based On MATLAB/Simulink And Neural Network

With the development of vehicle technology and the remarkable improvement of people's living standard, people's requirements of vehicle stability are also increasing, and making the vehicle handling and stability as an important direction of the car. Compared with the Front-Wheel Steering system, Four-Wheel Steering system has the following advantages:1.At the low speed, the 4WS can reduce the turning radius, improve the mobility, and make it easier to access to paking lots or maneuvere in narrow road;2.At the high speed, the 4WS can change lanes quickly, ensure that the sideslip angle is basicly zero, improve the tracking capacity and reduce the skid and drift phenomena, so as to improve vehicle handling and stability and enhance security.This paper presents a new control strategy, study the 4WS control method by introducing neural network control theory. It's main contents are as follows:1.Introduces the working principle and the development process of Four-Wheel Steering, models the two degress of freedom linear and three degress of freedom linear Four-Wheel Steering vehicle model, analyses the nonlinear factors of 4WS, focus on the tire sideslip which can infact the steering characteristics and stability of vehicle,and introduces some common tire model, studies the"semi-empirical tire model". Through MATLAB programming fits the relationship between sideslip force and sideslip angle, model the three degress of freedom nonlinear Four-Wheel Steering vehicle based on Simulink, and compared with the Front Wheel Steering system;2.Analyses the stability of 4WS based on evaluation indicator, put forward the control strategy of 4WS, that is, at high-speed small steering angle, ensure the sideslip angel is zero, at large steering angle, can resist skip, at low speed, ensure vehicle mobility;3.Design nonlinear data acquisition module based on control objectives and control strategy, uses the data to train BP neural network, determines the optimal network structure and parameter values, compared with Front-Wheel Steering system, the simulation results shows that the 4WS with neural network can improve handling stability;4. Presentes a new control method, simulates neural network control method and other Four-Wheel Steering control method, the results shows that the new control method can improve the handling stability and tracking capability greatly, validate the research study of Four-Wheel Steering has some reference value.