基于群体智能成果的路径规划程序自动生成系统

Path planning algorithms are widely used in various motion planning tasks, such as robot motion and autonomous driving. So far, many excellent path planning algorithms have been proposed for applications in different fields. For a specific task environment, choosing the appropriate path planning algorithm can plan a better path that satisfies the constraints more efficiently. Based on the results of swarm intelligence, the adaptability and path planning efficiency of rapidly-exploring random tree (RRT) path planning algorithm and its variants RRT-Star path planning algorithm and RRT-Star-Smart path planning algorithm under different task environments were studied. Using genetic programming algorithm as a framework to design a system, which could automatically analyze the map features of the current environment and combine the characteristics of RRT path planning algorithm and its variants to generate new path planning algorithms that were more suitable for the current environment. The generated path planning algorithm can efficiently plan a feasible path from the starting point to the target point.