Multi-robot Task Allocation And Path Planning Based On The Theory Of Immunity

Because of its incomparable advantages, multi-robot system has been widely studiedinstead of single robot system. At present multi-robot system is mainly used in multi-robotcooperation handling, multi-robot football match, factory automation, multi-robot operationexploration and map construction, etc. Multi-robot cooperation handling is one of the keyproblems of multi-robot system, in which task allocation and path planning is being makedfurther research of.To finish cooperation handling task in multi-robot system, task allocation between thebox and robot needed to be solved at first, in the process task allocation strategy is directlyrelated to the system efficiency. Inspired by the mechanism of idiotypic network hypothesis,the stimulation and suppression between the antigen and antibody by taking the box androbot behavior as antigen and antibody respectively are defined, and then task allocationmodel is established; Based on interferon regulatory mechanism, interferon regulatory factoris defined to distribute the robots and boxes according to the concentration between them, tooptimize the allocation strategy. The preliminary research results show that: the algorithmfor task allocation can solve the problem of long planning path, high loss of robot capacityand the larger number of robots that participate in the cooperation; Then the systemefficiency is improved, multi-robot cooperation task allocation strategy is fully embodiedand the algorithm shows strong planning ability.After getting the task allocation strategy, multi-robot start to carry the boxes accordingto the strategy, there may be some obstacles in the environment, so getting a optimal pathfrom the starting point to the target point without touching the obstacles is very necessary. Inorder to get planning path in the complicated environment, a new type of fireworksexplosive based on immune genetic algorithm is presented: in general immune geneticalgorithm the vaccination mechanism is included, immune algorithm is build wth theselection, crossover and mutation factor, vaccine extraction, vaccination; Inspired byfireworks explosion phenomenon, the fireworks explosion mechanism is put forward; whenin population is in certain generation, choose some outstanding individuals and expanedthem in their neighborhood, then the mobile robot path planning is realized in theobstacle-environment. The preliminary research results show that: the planning algorithmavoids the collision between the robot and obstacles, and can figure out a optimal path from start to finish.At the end, cooperation handling experiment is done based on the robot task allocationstrategy and path planning algorithm. Experiment results show that: the task allocationstrategy can optimize system efficiency and execution time, path planning algorithm canshorten path length; As a result, the cooperation handling task can be well completed.