In order to reduce the impact of the wake effect on the wind farm power generation, to improve the utilization rate of wind energy, an optimization algorithm-adaptive weight Genetic-Particle Swarm Optimization Algorithm (GA-PSO) is proposed. Firstly, the cost per unit of produced energy and coordinates of wind turbines were regarded as the target function and variable respectively, inertial weight was added to the speed update of optimization variable to change the search speed. Secondly, the layout of wind turbines was optimized based on micro-site selection of WASP. Eventually, the calculation results were also compared with the genetic algorithm(GA), firefly algorithm(FA) and particle swarm algorithm(PSO) optimization algorithms. The results show that the cost per unit of produced energy in the wind farm optimized by PGOA algorithm is 0.2016 ten thousand yuan/GWh , reducing 0.0232 ten thousand yuan/GWh and the annual power generation is 82.633 GWh, increased by 8.538 GWh compared with before optimizing, meanwhile the wake loss is decreased by 1.12%. The research results are guiding significance to the future micro-site selection of wind farms.