The study shows that, to save energy and improve bandwidth efficient cluster-based topology, and scalability of the network, but difficult to balance the network after the running of the period of time, the remaining energy of all nodes. In addition, with increasing WSN sensor nodes, the path of the search space to increase exponentially therefore seek effective heuristic algorithm to solve the clustering and routing in WSN is very important. Probabilistic approach in Reac-In algorithm is a new kind of optimization algorithm, which draws on the principles of Iterative change the cluster size and cluster head by optimization method.(1) broad access to information at home and abroad, an overview of the wireless sensor networks and its characteristics, summed up the WSN routing features and design requirements, analysis of current major the strengths and weaknesses of the routing protocols, and pointed out that the current research hotspots and unresolved problems. (2) We were studied on how to dynamically determine the optimal number of clusters, to analyze the factors affecting the optimal number of clusters. Simulations results show that Reac-In improve the optimal number of clusters compared to the life cycle of the network to be extended. (3) considering the degree of energy balance remaining node residual energy and network energy consumption is proposed based on the right to non-linear decreasing Presented with energy inspired based the priority encoding PSO routing algorithm to reduce the probability of an invalid path and improve the reliability of the network routing to provide new ideas for WSN routing optimization.