Research On Trajectory Tracking Control Of Intelligent Vehicle

With the development of society,the number of automobiles is increasing day by day,and the resulting problems of energy consumption,traffic congestion and traffic accidents are seriously affecting people’s lives.Intelligent vehicles can partially or fully replace human drivers and can avoid traffic congestion and traffic accidents caused by driver factors.Trajectory tracking control is one of the key points in the study of intelligent vehicles and an important part in the realization of intelligent transportation,and the study of trajectory tracking control problem is of great practical significance for the realization of intelligent vehicles driving automatically.The model predictive control algorithm can deal with the multi-constraint problem well and has unique advantages for studying the trajectory tracking control problem.In this thesis,the trajectory tracking control problem of intelligent vehicles is studied based on model predictive control algorithm.According to the needs of the system model,a simplified vehicle three-degree-offreedom dynamics model is established considering the vehicle’s motion in three directions: lateral,longitudinal and transverse pendulum,combined with the vehicle tire model based on the magic formula to further simplify the vehicle model on the basis of small-angle steering,reducing the complexity of the vehicle model and laying the foundation for the subsequent research on trajectory tracking control.The nonlinear model predictive control is studied,and the linearization process is transformed into a linear model predictive control which is convenient for solving operations,and the detailed linearization process transformation process is studied,and the prediction model and objective function are established.The linearized error model,vehicle dynamics constraints and quadratic programming problem are studied,the model prediction trajectory tracking controller is established by combining the vehicle simulation model,the influence of the prediction time domain,an important parameter in the model prediction control,on the system is studied,the fuzzy control rules are formulated by the two inputs of speed and adhesion coefficient,the optimal prediction time domain parameters are obtained,and the variable time domain adaptive trajectory tracking controller is designed by combining the fuzzy control tracking controller combined with fuzzy control.A joint simulation platform is built using MATLAB/Simulink and Carsim software,and the effectiveness of the designed controller is verified under various working conditions.Considering the existence of obstacles in the reference trajectory,the obstacle avoidance trajectory tracking control is studied,and a two-layer system of trajectory replanning and trajectory tracking is established.The upper layer trajectory planning system receives the obstacle information and calculates the local reference trajectory according to the model prediction control algorithm solution,and the lower layer trajectory tracking control system receives the information transmitted from the upper layer and outputs the front wheel turning angle to control the vehicle driving.A joint simulation platform is built to conduct obstacle avoidance trajectory tracking tests on single and double obstacles on dry and slippery roads respectively,and the results show that the vehicle can adapt to different speeds and successfully avoid obstacles to complete trajectory tracking.