Research On The Stability Of The Semi-trailer Based On Neural Network Control

With the continuous development of science and technology, highway transportation has become a dominant force to promote the development of social economy. Semi-trailer train car because of its large volume, high efficiency advantages, widely applies in the middle of the road transportation. Meaning-while, as the complex construction, which has paid big attention to it due to traffic accident. So this article take the semi-trailer train as the research object, analyze its yawing stability and roll stability, and design the neural network controller, the stability of the control is realized. In this paper, the main work is as follows:Firstly, analyzed the semi-trailer train yawing stability and roll stability, the instability chance is the research target, by simplifying, install three axis semi-trailer train nonlinear tire model.Secondly, the meaning of biological neural network and artificial neural network was introduced and analyzed the structure and principle of BP(Back Propagation)neural network, through the learning of BP neural network standard, put forward five ways to enhance learning rules, make a chance for the system identification and controller design.Once again, expounded the meaning of the system identification, analyzed its principle and process, including training date acquisition and processing and learning algorithm selection and parameters determination, according to the nonlinear vehicle identification model NNI(Neural Network Identification), design the neural network controller NNC(Neural Network Controller), which took the lateral acceleration as the control target.Finally, the dynamic simulation select the hook and double line, two typical working condition in Trucksim, and the high and low adhesion coefficient road with Matlab/Simulink. The simulation results shows the Neural Network Controller can effectively control the stability of semi-trailer train, realized the research goal.