Research Of Obstacles Avoidance For Mobile Robot Based On Artificial Potential Field

Since the beginning of the21th century, with the continuous progress of science and technology, the pace of development of robot technology is getting faster and faster, more and more important for the future of the people’s life and social development, the emergence of the root makes people more dependent on help complete types of dangerous, heavy, complex work. The mobile robot of autonomously avoid obstacles has become the most important research to scholars at home and abroad.In this thesis, The existing robot obstacle avoidance planning algorithms has been studied and summarized, focus on the method of obstacle avoidance in the unknown geographical environment, propose a new obstacle avoidance algorithm based on artificial potential filed method, and the algorithm has been discoursed and proved, the contents are as follows:(1) Method of obstacle avoidance based on the reaction typeThe method used the local path planning based on real-time access to information from sensors, the existing robot obstacle avoidance planning algorithms has been analyzed and summarized, propose a new obstacle avoidance algorithm based on reaction, the mobile robot operating environment is defined as different scenes, the running direction of robot is determined by the corresponding actions has been created according to the type of scenes. The method can effectively solve the problem of anti—oscillation in a narrow space and avoid the U-shaped trap.(2) Research of obstacle avoidance for mobile robot based on improved Artificial Potential FieldBecause the robot is running in an unknown geographical information environment in the most of cases, the real-time environment information needs access from sensors, in order to adapt to the dynamic environment, the dynamic artificial potential filed method has obtain to prove the traditional artificial potential filed method, the corresponding strategies have been propose to solve the target point unreachable and local minimum problem; a new "rescue" planning of create a new intermediate target point combined with the idea based on obstacle avoidance planning of reaction type for the local minimum point, in order to prevent the robot into a local point repeated, propose a "memory" selection planning, thus ensuring the robot is effective, fast and safe to move the target point.