融合鱼群和微分进化的蚁群算法的无功优化

对于求解电力系统无功优化问题,提出了一种融合鱼群和微分进化的蚁群优化算法（FDEACO）。受人工鱼群觅食、聚群和追尾行为的启发,在基本蚁群算法的基础上,应用人工鱼群算法的追尾行为对蚁群在可行域上搜索到的解进行改进,加快了向最优解收敛的速度。在信息素更新机制里,通过引入微分进化算法的发散项,增加一个随机扰动,减小了算法陷入局部最优的可能性。在IEEE30测试系统上对新提出的算法进行校验,并与其它算法比较,证明FDEACO算法收敛速度快、全局寻优能力强。

Reactive power optimization based on ant colony algorithm which combines artificial fish swarm algorithm and differential evolution algorithm

This paper proposes ant colony algorithm which combines artificial fish school algorithm and differential evolution algorithm(FDEACO) to solve reactive power optimization in electric power system.Enlightened by foraging,clustering and tailgating of fish school,on the basis of ant colony algorithm,solutions of ant colony algorithm found in feasible region are improved by applying tailgating of artificial fish school algorithm,which accelerates the convergence speed of optimal solution.In pheromone updating,the potential of local optimum is reduced by introducing divergence of differential evolution algorithm and adding random disturbance.The new algorithm is examined and compared in IEEE30 system.Results show that FDEACO has high convergence speed and strong global optimization.