Research On Trajectory Optimization And Tracking Control Of Mobile Robot

With the development of science and technology,mobile robots have been widely used in all aspects of society,such as household sweeping robots,logistics robots in logistics companies,mobile robots in museums and medical robots in hospitals.Robot technology has gradually become an assistant to human life and work.Path planning technology is the basis of robot autonomous navigation,and the trajectory tracking technology of mobile robot is the guarantee that the robot can run safely to the target point,and whether the mobile robot can run smoothly to the target point depends on the smoothness of the planned path.Therefore,in the process of moving forward,the trajectory consistent with the kinematics of the vehicle is generated through the calculation of the system,and the advancing Angle of the mobile robot is adjusted during the moving process,so that the mobile robot can be stably and safely tracked.This is the key research content of this paper.Based on static raster map,this paper analyzes and compares five kinds of path planning algorithms.By comparing the advantages and disadvantages of different algorithms,A~*algorithm is selected as the path planning algorithm in this paper.In the aspect of trajectory tracking control,the model predictive control algorithm is selected as the trajectory tracking control algorithm by comparing the trajectory tracking error of pure tracking method,Stanley method and model predictive control algorithm.The main contents of this paper are as follows.(1)The global path trajectory planned by the A~*algorithm has many corners and is not smooth,and the movement direction of the mobile robot changes greatly during operation,which will affect the control accuracy of the robot.To solve this problem,this paper proposes an improved A~*trajectory planning algorithm with Minimum jerk,and compares it with the improved A~*algorithm with Bessel curve.The algorithm adopts time homogenization of discrete points in path planning,constructs multi-segment polynomial function description,and uses Minimum jerk to build its optimization function to form feasible interval of trajectory.The smoothness constraint and inequality constraint of adjacent trajectories are added to the traditional kinematics model of mobile robot,and the optimal trajectories are solved.Through simulation tests,the proposed algorithm shorts the path length by 1.56%compared with the path planned by A~*algorithm.On the other hand,the maximum tracking error of the road left side and the maximum tracking error of the road right side after tracking trajectory optimization is reduced by 94%and 89.5%.(2)Trajectory tracking control takes model predictive control as the core,builds vehicle kinematics model to analyze future dynamics,performs mathematical solution to optimization problems,and applies the first control quantity obtained from the solution to the system.This paper built a joint simulation platform based on Carsim and MATLAB/Simulink.The experimental platform adopted the joint difference between horizontal and vertical as the tracking error to realize the system’s tracking of roads,ensure the tracking speed of mobile robots,and improve the control accuracy of trajectory tracking.On the other hand,the semi-physical co-simulation experiment is realized.(3)Model-based predictive control algorithm has a small tracking error,so this algorithm is introduced into ROS and a local path planning algorithm for model-based predictive control is designed.Simulation results show that the TurtleBot3 burger robot model is adopted in this paper.Compared with the Dynamic Window Approach(DWA),the model predictive control algorithm is more consistent with the trajectory of global path planning in its own environment map and the self-built environment map.The control precision of mobile robot trajectory tracking is improved.