| The autonomous mobile robots are finding their way into every aspect of our daily lives,from intelligent navigation to logistics,providing immense convenience.This thesis aims to investigate the path planning algorithm of robots in the field of automatic driving.The main contributions are as follows:· Global Path Smoothing:A path smoothing algorithm is presented that adheres to the kinematic characteristics of the given robot.This results in smoother robot movements,ultimately improving motion stability.· Local Path Planning Algorithm:A bounded-DWA(Bounded Dynamic Window Approach),a new and efficient local motion planning algorithm for mobile vehicles equipped with Ackermann steering geometry is presented.The new algorithm im-proves the DWA by incorporating Ackermann steering kinematic constraints to re-strict the sampling space and improve sampling efficiency.The new algorithm also includes a new evaluation function that selects trajectories that quickly approach the target location while being as close as possible to the target pose.Furthermore,a fuzzy logic controller is used to adjust weighting coefficients in real-time.These improvements address the problems in local motion planning algorithms,making the algorithm a valuable contribution to mobile vehicles with Ackermann steering structures.· Intelligent Automatic Parking Simulation System:Using ROS and Gazebo,this thesis presents a simulation system for intelligent automatic parking,including GUI,vehicle/environment modeling and path planning/smoothing algorithms.The system facilitates the comparison of multiple path planning algorithms,and enables swift modification and testing of parking algorithms.Moreover,it provides the ability to directly observe the impact of various algorithms. |
