Research On Lateral Control Of All Terrain Vehicle Based On SPSA-LQR

The path tracking control technology is an important part of automatic driving technology.As the key technology of path tracking control,the lateral path tracking control technology of vehicle has gradually become the research focus of experts and engineers.In this thesis,a new solution to the problem of automatic driving vehicle lateral path tracking is proposed,and the results are as follows.Firstly,according to Newton's second law of motion and vehicle dynamics principle,the vehicle dynamics model based on the simplified bicycle model is constructed.Because the state variables in the model are difficult to measure,and in order to better express the relative position between the actual vehicle position and the planned path,a lateral dynamic model with error as the state quantity is constructed.Through the analysis of MATLAB simulation software,the performance and accuracy of the dynamic model are evaluated,which proves that the model can accurately express the real vehicle.Secondly,in the design of the lateral controller,in view of the difficulty in determining the parameters of the linear quadratic regulator algorithm,the simultaneous perturbation stochastic approximation algorithm(SPSA)is introduced to optimize the parameters of the linear quadratic regulator.A lateral controller based on SPSA-LQR is designed,and the performance of the traditional linear quadratic regulator and the lateral controller based on SPSA-LQR is compared by simulation software.It is proved that the lateral controller based on SPSA-LQR has better performance in tracking performance and energy consumption.Thirdly,aiming at vehicle active steering system,a current loop regulator based on traditional PID is designed.Then the membership function of input and output of Fuzzy PID parameters is designed,and the fuzzy inference rules from input to output are designed according to expert experience.The position loop controller based on Fuzzy PID is realized.Then the bottom steering system of the vehicle is tested.The experiment shows that the Fuzzy PID control has a certain system adaptive ability,and can improve the rapidity and accuracy of the response of the active steering system.Finally,the test platform of all terrain vehicle is introduced,including the whole vehicle structure,steering system,sensor controller system,vehicle communication architecture,etc.At the same time,the software system running on the controller is introduced.According to the road conditions that all terrain vehicle can meet,two reasonable experimental conditions are designed.The lateral control method based on SPSA-LQR and the bottom active steering system proposed in this thesis are verified and the test results are given.The results show that the lateral control algorithm designed in this thesis can cover most of the driving conditions of the all terrain vehicle,and can track the planning trajectory better than the traditional LQR controller.