Trajectory Tracking And Obstacle Avoidance Of Unmanned Vehicle Based On Model Predictive Control

Unmanned vehicle trajectory tracking and obstacle avoidance have been hot issues in unmanned driving systems,and are the core research content of unmanned driving technology,and are of great significance for improving road driving safety.Model Predictive Control(MPC),due to its distinctive features of model prediction,rolling optimization,and feedback correction,makes it currently recognized as an effective algorithm for handling multi-variable constrained control of complex systems.Unmanned vehicle system is a typical multi-variable and stong-coupling system Compared with other control algorithms,MPC has natural advantages.Therefore,in this paper,under the framework of model predictive control,the trajectory tracking and obstacle avoidance control of unmanned vehicles are studied.The main research results are as follows:(1)An unmanned vehicle dynamics model under a small angle assumption is established,and an unmanned vehicle trajectory tracking control algorithm based on nonlinear predictive control is designed using this model.The experimental results show that the MPC algorithm can track the predetermined trajectory well(2)Aiming at the problem of time-consuming calculation of unmanned vehicle trajectory tracking algorithm based on nonlinear control,an unmanned vehicle trajectory tracking algorithm based on multi-step linearized predictive control is proposed.The algorithm uses a multi-step linearization strategy to calculate the optimal control sequence near the nominal trajectory of the unmanned vehicle,and uses an iterative strategy to estimate the nominal trajectory of the unmanned vehicle.Compared with non-linear MPC and single-step linearization strategies,the simulation results show this method can greatly reduce the algorithm operation time and has a high tracking accuracy.(3)In view of the complex design of traditional unmanned double-layer obstacle avoidance control algorithm and the large amount of computing resources,a single layer unmanned vehicle obstacle avoidance strategy based on model predictive control is proposed,and obstacle information is added to the optimization proposition in the form of a penalty function In this paper,an obstacle avoidance algorithm for autonomous vehicles based on Taylor expansion and an obstacle avoidance algorithm for autonomous vehicles based on bilinear interpolation are designed.Experimental results show that the single-layer unmanned vehicle obstacle avoidance algorithm can smoothly avoid obstacles on the target trajectory without deviating from the original trajectory as much as possible.