Path optimization algorithm for agents based on artificial immune and emotional learning

This paper combines artificial immune and emotional learning methods to solve the path optimization problems in complex, dynamic and real-time multi-agent systems. In artificial immune algorithm, path metric is defined as the affinity function between antigen and antibody, namely, the matching degree between optimal path and candidate paths. At the same time, emotional learning method is used to train the weight factors, which will affect path choosing; so that the weight factors in path metric can be updated in real-time, and the optimum path can be got. The validity of the proposed algorithm is proved through applied in CSU_YunLu RoboCupRescue simulation team.