Research On Multi Robot Path Planning Based On Coevolution Genetic Algorithms

Multi-robot system has many advantages that can not be compared to a single robot system, it is gradually becoming an important direction of the robot research. In multi-robot system, path planning is one of the key issues. Path planning is the basis of the robot system to perform various tasks, reflecting the robot system in motion with the surrounding environment interaction. Multi-robot system in a complex environment of the path planning is very relevant. Path planning methods described herein for multi-robot systems, path planning task collaborative research fleet. The main points are as follows:(1)In this paper, since the co-evolution, population affinity not only with the individual where relevant, need and collaborative behavior of individuals in other populations to be evaluated, usually coevolutionary algorithm for establishing cooperative behavior only in the evaluation, but not a note of the successful collaboration of individual portfolio, so the next evolution, each population remains independent evolution, again selected collaborators with the last probably not the same, so that the destruction of the original winning combination, you can not be on successful combination times preserved immune proposed coevolutionary algorithm, full use of the characteristics of immune memory, to build the global memory set, note the success of collaborative behavior, so that populations evolve.(2)In this paper, with co-evolutionary algorithm ICCEA improved to solve the multi-machine path planning in a static obstacle environment. Based on multi-robot path planning method is the use of coevolution coevolution of multiple robots to optimize the overall ideological path planning. This algorithm overcomes the traditional path planning algorithm is slow evolutionary disadvantage, accelerated the overall planning of cooperative multi-robot path, maintaining the diversity of the population evolution, path planning in order to avoid falling into local optima. The ICCEA used to solve multi-robot path planning and simulation environment static obstacles to achieve the three robot obstacle avoidance, collision avoidance shortest path.(3) In this paper, based on the application ICCEA solve the multi-robot path planning. Using the Leader-Follower centralized structure formation, in the multi-robot path planning algorithm coevolution added feature to control the controller uses a triangle formation, formation control in order to achieve the task.