The genetic algorithm was improved and applied to the optimization design of wireless sensor network, according to the actual environment of forest to establish an appropriate mathematical model, and on this basis the fitness function and the sensor network strategy was given. When anomalies occur, this algorithm can accurately and timely issue alerts and position information. For the genetic algorithm is easy to enter the errors of the local optimal solution, the simulated annealing operator join in the genetic algorithm, also based on the past, selected Crossover probability and Genetic probability of improperly, brought a significant impact to the optimization results, in this paper, in the optimization process dynamically adjust the crossover probability and mutation probability. MATLAB simulation results show that the improved genetic algorithm improved Algorithm optimization speed, overcome the local convergence of the errors, optimized Wireless sensor networks energy to achieve the longest life cycle of the network.