Research On The Fast Obstacle Avoidance Path Planning Algorithm Based On The Minkowski Sum

Mobile robot path planning is an important branch of robot study, which is widelyuesed in many areas such as assembly simulation, numerical control machines androbotics. It is still a hot and difficult problem to improve the instantaneity and efficiencyof collision detection for robot path planning in barrier environment, which will arrive atdestination by particular path. This paper studies and analyses the mobile robot pathplanning problem, mainly including collision detection method and path planningmethod.Firstly, this paper briefly introduces the research background and significance of themobile robot path planning, summarizes the research situations, research methods andcritical techniques of mobile robot path planning both at home and abroad, analyzes theirrespective advantages and shortcomings, which establishs a solid research foundation forthis paper.Secondly, this paper proposes a fast collision detection algorithm of polyhedra basedon the Minkowski Sum, gives the specific calculation method of the Minkowski Sum anda mapping strategy from the work space to the configuration space. Configuration spaceobstacle and free configuration space are structured quickly using the Minkowski sum.Consequently, the path planning problem of a polyhedron robot in the workspace wastransformed into that of a point representing the robot in configuration space. It haspreferable real-time capability and reduces the storage space effectively.Thirdly,this paper presents a fast path planning algorithm based on the Minkowskisum. The algorithm can change rotational motion into translational motion by slicingapproach with the aid of the Minkowski sum. And then adjacent roadmaps are chainedtogether to constitute the whole configuration space roadmap by the stair structure.Finally, this curve was mapped to the workspace. Thus, the algorithm improves thereal-time performance of the path planning.Finally, in C#environment, we realize the barriers avoidance of the robot in the pathplanning simulation system using OpenGL graphics library. The correctness and excute efficiency of the algorithm is verified by our experiment, and the detail results andanalysis is provided.