A Kind Of Multi-robot Path Planning Research And Simulation

This article has conducted the deep research on multi-robot path planning problem. The path planning algorithm combines the global path planning with local path planning. Path planning algorithm of multiple mobile robots is applicable to the environmental information partially known.Global path planning uses genetic algorithm combined with RRT. The RRT algorithm is the initial population of genetic algorithm. According to the specific problem and the particu-larity of the initial population, Genetic algorithm encoding, fitness function and genetic oper-ators implementations are given. In addition, arrcoring to the multi-robot system, the fitness functions and end condition in GA are changed. The improved genetic algorithm satisfied the relations of the cooperation among the robots. Simulation results show the feasibility of the algorithm.In order to improve the algorithm in real time and complete the global path planning, RRT algorithm adds adaptive step algorithm and regression and obstacles forecast algorithm. The simulation results show that the improved RRT algorithm accelerates the speed of origi-nal algorithm for path planning.The model of velocity obstacle has used in local path planning. The improved optimal reciprocal collision avoidance algorithm based on velocity obstacle is propsed. The priority rules and risk assessment adds to the optimal reciprocal collision avoidance algorithm. This algorithm can realize mutual collision between robots, and has fast calculation speed. Also the algorithm is not sensitive to system scale.