Research On Smooth Path Planning And Trajectory Tracking Of Mobile Robot

Mobile robots have been widely used in medical service,catering,logistics transportation and other fields,and manual operations are gradually replaced by advanced robotic technology.Path planning and trajectory tracking control are the key technologies of autonomous robot navigation.The smoothness of the path affects the movement effect of the robot.The basis of autonomous navigation is the ability to plan a path that conforms to the kinematics of the vehicle and track it stably.In this thesis,an experimental platform was built with Song Ling SCOUT chassis aiming at the actual problems of mobile robots.It installs the Ubuntu system on the onboard computer and it is equipped with 16-line Lidar and IMU sensors.The plan is independently designed for mobile robot with ROS communication mechanism to realize the planning movement.The surrounding environment and obstacle information are obtained by fusion processing of the data that are collected by Lidar and IMU,then generate a navigation map.The mobile robot performs Hybrid A~* path planning and model predictive control based on map information and sensor data.The main contents of this thesis are as follows:(1)Aiming at the problem of continuous two-step expansion,the more variation in the robot's movement direction will affect the accuracy of tracking control.This thesis proposes a cost penalty based on the curvature difference as a solution.Analyzing the curvature of adjacent nodes,it adds the curvature information of the current node and the parent node to the node attributes.The expansion cost function is optimized by the sum of the adjacent curvature differences of consecutive steps,and the key value of the node expansion is updated.The change in the curvature of the trajectory is significantly slowed down.The planned path after smoothing is more suitable for trajectory tracking,which reduces the performance requirements of tracking control to a certain extent.(2)The iterative calculation of gradient descent method is used to optimize the feasible path.The smoother is composed of obstacle items,curvature items,smoothing items and Voronoi items.Iterative optimization of planned path points can obtain a smoother path point sequence.The smoothness of the robot's moving trajectory is guaranteed,and the kinematics needs of the robot are met.(3)The tracking controller takes model predictive control as the core,and constructs predictive models to analyze future dynamics.It mathematically solves the open-loop optimization problem,and applies the first variation of the solution to the system.In this thesis,constructing the kinematics model of the mobile robot,a speed constraint method based on the curvature of the trajectory is designed coupled with the analysis and calculation of the dynamic relationship.At the same time,the output speed must meet the curvature speed constraint and the maximum speed constraints.It ensures that the tracking speed of the mobile robot meets performance requirements,while improving the safety of the tracking algorithm.Finally,building up the SCOUT robot platform,the Hybrid A~* path planning and model predictive control modules are completed through program design.The experimental environment map is established to verify the planning module and the control module.There are 78 figures,11 tables and 56 references in this thesis.