基于动态规划法和遗传算法的配电网无功优化  

Reactive Power Optimization in Distribution Power Network Based on Dynamic Programming and Genetic Algorithm

作　　者：赵昆[1] 耿光飞[1] 

机构地区：[1]中国农业大学信息与电气工程学院,北京100083

出　　处：《沈阳农业大学学报》2010年第2期251-253,共3页Journal of Shenyang Agricultural University

基　　金："十一五"国家科技支撑计划资助项目(2006BAJ04B06)

摘　　要：对配电网无功优化规划问题进行研究,针对配电网的辐射型结构,在无功二次精确矩选择补偿点方法的基础上,提出了配电网无功优化分区的方法,并在选点时考虑已选补偿点对后续优化选点的影响,运用动态规划法来确定最优补偿位置;用改进遗传算法求解补偿点的最优补偿量,对选择、交叉、变异等算子进行改进,每一代群体进行分组进化,并采用自适应罚因子、交叉率和变异率,以提高收敛速度和解的质量。采用IEEE 33节点系统验证了本算法的有效性和实用性。本研究提出的算法能使配电网各分区得到合理的无功补偿,从而提高配电系统经济性和电压质量。The reactive power optimization in distribution network was studied. A power system division method based on reactive quadric accurate moment was proposed with consideration of the radial structure of distribution network. Dynamic programming was used to ascertain the optimal compensation nodes due to the influence of the nodes that had been compensated on the follow-up nodes being selected. Then an improved genetic algorithm was proposed to determine the optimal compensation capacity. The genetic algorithm operations such as selection, crossing and mutation were improved and all individuals of every generation were divided into two groups in the evolution. Adaptive penalty factor, crossing rate and mutation rate were used to enhance the convergence speed and the quality of solution. The IEEE33-bus system test showed the effectiveness and practicality of the proposed methods. The rational reactive power compensation of the distribution network was achieved using the methods mentioned above and thus economical benefits and good power quality were obtained.

关 键 词：无功优化 电力系统分区 动态规划法 改进遗传算法 

分 类 号：TM7[电气工程—电力系统及自动化]