Abstract: In the dynamically perceived task environment with task-rich area,multi-robot foraging faces a lot of spatial conflictions caused by improper task allocation.In order to solve this problem,an autonomous task allocation method,which is based on an ant colony algorithm named repulsive pheromone ant colony algorithm(PR-ACA),is presented for the multi-robot system.Under the unknown unstructured environment,a simulation experiment of multi-robot cooperative foraging is performed.Experiment results show that the presented method can make multiple robots perform task allocation autonomously,and can effectively decrease spatial confliction among robots,especially when the group size is large.