Research On Robot Path-planning In An Unknown Dynamic Environment

For the diversity and complexity of the environment, path planning has being an important aspect in the mobile robotic research field and also a crucial technology to intelligentize the robot. This thesis has made a deep and systematic study of the path planning problem. The following aspects are investigated and discussed: Firstly, the thesis briefly introduces the history of the mobile robotic development and summarizes different research aspects of the robot. Secondly, after analyzed and compared different research development level of foreign and domestic, the thesis proposes a solution to the robot path planning in dynamic environment. To practical purpose, the solution tries to resolve the path planning problem of a certain kind of environment, which has static global environment with unknown-obstacles. The solution can be divided into two levels:the global path-planning level and the local path-planning level. The main purpose of the first level is to lay out a global path in the static global environment which satisfies the optimal condition. In this level, the MAKLINK graph is built to describe the static environment of the robot, and then the Ford algorithm is used to obtain the shortest path between any points in the environment, which is optimized by the genetic algorithm finally. The experiment's results verify that the above solution is effective. According to the local environment detected by the sensors, the second level is to plan a local path for robot when the unknown-obstacles appear. The local planning level can resolve not only the obstacle avoidance, but also the path planning problem. Thirdly, based on the above, the thesis discusses the local path-planning problem for robot to avoid dynamic obstacles with fixed track in detail. Finally, the full text is summarized and multi-robot path-planning problem is prospected.