Local Path Planning Algorithm And Experiment Research For Wheeled Mobile Robot

This project is derived from the development project of an experimental platform for a non-holonomic constrained mobile robot Turtlebot3-Waffle,and studies the key technology of the navigation system,the path planning algorithm.Based on the design of path planning algorithm and the technical scheme of navigation system,this paper establishes a complete navigation system and carries out experimental verification which has a certain practical value and research significanceThe navigation system is developed with an indoor scene and incomplete environment information.It's based on hardware systems including laser radar,single board computer,control card and stepper motor,the navigation task is completely implemented with several functions including moving to target points which is assigned,avoiding obstacles,and path planning of smooth track,real-time track tracking and more.The path planning algorithm is based on computational geometry theory.The sixth-order Bezier curve is used for kinematic approximation in path planning.The algorithm has considered the velocity state,kinematics model and non-holonomic constraints,and the endpoint state constraints.The velocity optimization of the track planning is to generate a smooth trajectory and discrete time output of control stream.The trajectory tracking algorithm is based on the control theory of Various Structure Systems(VSS).The Integral Sliding Mode Controller is used to reduce the trajectory following error,and the saturation function is added to reduce the chattering phenomenon of the control system.Finally,the navigation system is implemented in the ROS operating system using the Python language.The virtual environment verification and real scene verification have proved the established navigation method is stable and effective,and can complete the navigation task based on single-line lidar sensing in indoor situations.The navigation system and wheeled mobile robot established in this project can be used as an experimental platform to support the teaching and experimental research of navigation system structure,planning algorithm and motion control content.