摘要
针对配电网络重构多为单一性能最优重构的问题,文章提出了使配电网线损、负荷均衡、供电电压质量最佳的多目标配网优化模型。结合GA中的进化思想和粒子群算法(PSO)中的群体智能技术,采用遗传粒子群混合算法寻优,通过随机权重方法来获得目标是Pareto前沿面的可搜索方向,体现出较GA和PSO更好的寻优性能。寻优过程中,部分个体以PSO方法迭代,其它个体进行GA中的选择、交叉和变异操作,整个群体信息共享,同时采用自适应参数机制和优胜劣汰的思想进化。在此基础上制定的配网优化方案能够在保证配网呈辐射状、满足馈线热容、电压降落要求和变压器容量等的前提下,最大限度地提高配电系统安全性和经济性。算例表明该算法在求解性能和效率两方面都有比较显著的优势。
According to the single performance of most distribution network reconfigurations （DNR）, this paper presents the multi-objective distribution network optimization model with the optimal network loss, load balancing, and power supply voltage. Combined with the evolution idea of genetic algorithm （GA） and population intellectual technology of particle swarm optimization （PSO） algorithm, it applies hybrid genetic particle swarm optimization algorithm （HGPSOA） to search the optimization. By the random-weighted method, it obtains the object that is the searching direction of Pareto front. During searching process, some individuals are iterated by PSO, the others follow the selection, crossover and mutation of GA, and the whole population information is shared by each agent. Simultaneously, it adopts the adaptive parameters mechanism and better fitness individuals surviving rules to evolve the population. Based on the above, distribution network optimization program can furthest enhance the security and the efficiency of distribution system, on the premise of ensuring that the distribution network is spokewise and also could satisfy heat capacity of feeder line, voltage reducing, transformer capacity and etc. Samples show that the algorithm has advantages both in effectiveness and efficiency.
出处
《华北电力技术》
CAS
2007年第9期1-7,共7页
North China Electric Power
基金
国家自然科学基金(70671039)
河北省自然科学基金(G2005000584)
关键词
多目标优化
混合算法
配电网络
重构
粒子群算法
multi-objective optimization
hybrid algorithm
distribution network
reconfiguration
PSO
作者简介
牛东晓（1962-），男，华北电力大学工商管理学院院长，MBA教育中心主任，教授，博士生导师，研究方向为电力市场技术经济、决策支持系统、电力负荷预测等方面。
