Particle swarm optimization based on Gaussian mutation and its application to wind farm micro-siting

摘要

In this paper, a particle swarm optimization algorithm with Gaussian mutations, denoted by GPSO, is proposed to solve constrained optimization problems. Two Gaussian mutation operators are employed to search the promising regions for better solutions. One operator is for the region between the personal best position and the global best one. The other operator is for the region around the global best position. The Gaussian mutations help the population jump out of local optima and find better solutions with more probability. The feasibility-based method compares the performance of different particles. Evaluated by three typical optimization problems, GPSO is more accurate, robust and efficient for locating global optima. The GPSO method is applied to a wind-farm micro-siting problem. Simulation results demonstrate that the power generation of the wind farm is further improved while the execution time is substantially reduced.