【摘要】 针对有功网损、电压偏差和静态电压稳定裕度的多目标无功优化问题.提出了一种用于多目标的改进纵横交叉算法。该算法采用纵向和横向两个不同方向的搜索机制,摒弃基础纵横交叉算法的竞争算子。在纵向交叉后采用目标序列排序建立虚拟个体适应度,再对粒子虚拟适应度大小进行比较选择精英粒子,而横向交叉之后不进行粒子比较。与传统的NSGA-II相比,该算法能够使粒子在收敛的过程跳出局部最优,提高粒子多样性,同时提高效率。最后引进非支配选择和拥挤距离计算,产生Pareto前沿。通过和NSGA-II在IEEE 14节点系统、IEEE 57节点系统多目标无功优化上仿真对比,结果验证了所提方法收敛精度更高,其Pareto前沿分布更加均匀,范围更加广泛,能够很好的解决电力系统多目标无功优化问题。更多还原

【Abstract】 In allusion to multi-objective reactive power optimization with active power loss,voltage deviation and static voltage stability margin,an improved algorithm of crisscross optimization algorithm is proposed.The improved algorithm retains the search mechanism of vertical cross and horizontal cross in the original algorithm. After the vertical cross,instead of utilizing the competitive operator, the improved algorithm choose the elitist particle by establishing virtual individual fitness based on the rule of target sequence sorting.While the horizontal cross is without this operation.Compared with the traditional NSGA- II,the algorithm can make the particle in the process of convergence to jump out of the local optimum,improve the diversity of particles and the efficiency of the algorithm. Finally,Producing Pareto frontier by introducing non- dominated selection and crowding distance calculation.Compared with the simulation result of NSGA-II in multi objective IEEE 14 node system and IEEE 57 node system reactive power optimization,It shows that the proposed method is higher convergence precision,the Pareto front distribution is more uniform,more wide to solve multi-objective reactive power optimization problem of power system.更多还原