COOPERATIVE BEHAVIOR LEARNING OF ROBOTS BY USING GA

It was tried to make a multi-robot cooperative system,which is learned by the genetic algo-rithm,to carry out maximum conveyance.The environment in which the multi-robot movement is carried out is expressed as a graph.The rule of movement is coded and optimized by the genetic algorithm.The times of reciprocating between two nodes that were specified beforehand are taken as the fitness.The environment was constructed on a computer,and the simulation was carried out.As the result of behavior learning the multi-robots give way mutually as circumstance demand.