Aiming at redundant nodes in wireless sensor networks and increased network costs, an improved coverage optimization method artificial fish swarm algorithm is proposed in this paper. First of all, mathematical model of nodes' utilization rate and coverage rate is established, and then artificial fish swarm algorithm is improved One, probability density function is used in the initial stage for the distribution of the initial location of individuals in the fish swarm so as to effectively avoid chaotic state of the fish swarm individual in the beginning; two, chaos algorithm is used in the foraging stage to interfere with the location of fish swarm individuals so as to effectively reduce the time of fish swarm individual getting close to the optimal solution; three, Gaussian mutation is used in the cluster behaviors so as to reduce the time of producing the optimal solutions. The improved artificial fish swarm can get the optimal coverage scheme while solving the model. Simulation experiments show that it can effectively improve the effect of network coverage as well as node's utilization rate, and reduce the network costs.