Multiple Robots Autonomous Task Planning Based on Improved Genetic Algorithm

This paper presents an autonomous task planning method for multiple collecting robots on improved genetic algorithm. Considering the limited capacity of collecting robots and the suitability of hopping robots in the microgravity environment of small celestial bodies, we developed a task planning model based on the Capacitated Vehicle Routing Problem (CVRP) and proposed an improved genetic algorithm to simultaneously allocate tasks and plan paths for multiple collecting robots. During robot movement, the A ∗ algorithm is employed for active obstacle avoidance when unknown obstacles are encountered. Experimental results demonstrated that our algorithm, can efficiently allocate tasks for the robot group, and each robot actively avoiding obstacles during movement.