Research On Obstacle Avoidance Path Planning For Six-dof Manipulator

Manipulator is increasingly common for industrial production, detection, explosive ordnance disposal, collection and other fields nowadays. It brings great help for the production life of the people. However,the use and development of manipulator is seriously restricted by the problem of obstacle avoidance of the work environment. So research on obstacle avoidance path planning for manipulator has important practical significance and use value. Because of the typical character of flexible, stable, operable and complex structure about the six-dof manipulator, this article carry on a research on obstacle avoidance path planning for six-dof manipulator.This paper studies form three aspects: manipulator modeling and construction of virtual environment, collision detection between manipulator and space obstacle, obstacle avoidance path planning method.DH model to do kinematic modeling for the six-dof manipulator is proposed in this work after In-depth understanding of mathematical modeling of the robot, DH model, the principles of the robot positive solution and inverse solution. And JAVA3D and VRML have been decided to use for constructing the virtual environment via the introduction of virtual simulation technology.Analyzing advantages and disadvantages of collision detection algorithm among bounding sphere, AABB, OBB, K-Dops, then proposed a ellipsoidal bounding box collision detection method based on the idea of bounding box. This method has the advantage of highly surrounded and simply established. It could achieve very good results when used into path planning for obstacle avoidance.A detailed analysis and research is carried out on the typical path planning method as inverse algorithm, artificial potential field, A* algorithm. Decided to use artificial potential field method, and then re-set the artificial potential field function. This method is possibly lead to local minimum problems which seriously affect the performance of manipulator obstacle avoidance. This paper presents a virtual target point method, which effectively solved the problem.This paper validates the method using AS-MrobotE modular robot and virtual simulation system. It is been proven that the proposed method can quickly and effectively planning path on the condition of obstacles or none and reach the desired research goals.