Robot Path Planning Based On Artificial Potential Field Method

Nowadays, the technology of robots is quickly developing and researchers constantly improve the level of intelligence of the robot. Besides, the autonomous navigation performance of mobile robots will reflect the level of intelligence of robots. Most of the mobile robots help people explore unknown regions and engage in dangerous operations,so path-planning-capability of robots in unknown environment is very important which means the robots need good environment adaptability and robustness. On the other hand,in order to regularly monitor the robots’ operational status and surrounding environment,the building of remote control system is also very necessary.Firstly, the article describes the structure of the mechanical system AS-R robot which is used in the experiments. On this basis, it builds a more accurate two-wheel differential drive mobile robot kinematic model, and it is derived by equation. In addition to, the simplification of working environment of the robot which turns into a two-dimensional coordinate system is also written in the article. It calculates the pose of the robot by reckoning principle.Secondly, the perception system of the robot and the analyzing of ranging principle and performance of the infrared and sonar sensors. In the laboratory environment, we analyze how the mobile robot selects safe distance in the state of motion. On the basis of PSD sensor and sonar sensor, it proposes two obstacle avoidance algorithms. The practicality of these two algorithms is verified by calling the library function of AS-R robot.Thirdly, this article introduces the classical principles of artificial potential field. As for local minimum trap, on the basis of improving artificial potential field method, it modifies the repulsive force direction and introduces the border following behavior, which can ensure the force robot feeled before reaching the target position is not zero. Then the mobile robot path planning process is divided into three behavior modules: tending to target, obstacle avoidance and border following behavior. The robot choose theappropriate behavior based on real-time environment and robot itself, ultimately, to reach the target position.Finally, in order to improve the robot autonomous navigation function and to monitor the surrounding environment, it design a set of robot remote contrlo system which based on wireless LAN technology. Experimenter can visit the client on the PC,which can achieve real-time monitoring of mobile robots.