基于DE-CSO混合算法的多目标无功优化

Multi-objective reactive power optimization based on DE-CSO hybrid algorithm

将差分进化(Differential Evolution,DE)算法融入到纵横交叉(Crisscross Optimization,CSO)算法中,提出一种差分纵横交叉混合(Differential Crossover,DE-CSO)算法。混合算法通过将CSO算法纵横两个维度的交叉操作和DE算法变异交叉操作相结合,增强了算法的全局搜索能力,提高了种群粒子的多样性和算法的收敛精度。此外,针对多目标函数难以评价解的情况,应用多目标快速非支配排序和计算粒子拥挤距离的方法对产生的粒子进行评价,产生Pareto前沿分布。将所提算法应用于降低有功网损及电压偏移量、提高静态电压裕度的电力系统多目标无功优化中,通过在IEEE14、IEEE57节点系统中的仿真,验证了DE-CSO算法在多目标无功优化问题中较CSO算法有更好的性能。

Differential evolution(DE)algorithm is integrated into crisscross optimization(CSO)algorithm,and a differential crossover(DE-CSO)algorithm is proposed.By combining the cross operation of two dimensions of CSO algorithm and mutation cross operation of DE algorithm,the hybrid algorithm enhances the global search ability of the algorithm and improves the diversity of population particles and the convergence accuracy.In addition,for the case that multi-objective function is difficult to evaluate the solution,multi-objective fast non-dominated sorting and particle crowding distance calculating are used to evaluate the generated particles,and the Pareto frontier distribution is generated.The proposed algorithm is applied to the multi-objective reactive power optimization of power systemto reduce the active network loss and voltage offset and improve the static voltage margin.Through the simulation in IEEE14 and IEEE57 node systems,it is verified that the DE-CSO algorithm has better performance than CSO algorithm in multi-objective reactive power optimization.