Artificial fish swarm algorithm (AFSA) has important theoretical research value and practical significance in solving VRP. The traditional AFSA which does not consider the structural features of VRP will lead to too complex for the process to solve problems, too much time to search optimal solution and too low computational accuracy. In this paper, the traditional method is improved that neighborhood search which are more efficient for VRP are used in the three behaviors of AF swarm, and discretize the three behaviors. The improvement optimizes...