摘要
为城镇配电网经济性优化,利用经济电流密度概念和多目标规划理论,提出计入导线截面策略因子投资费用和年网损费用最小的多目标函数。采用遗传算法和退火动态罚因子研究实数编码并行优化、动态罚函数约束、配电网路径与潮流自识别解码方法。10 kV配电网实例验证表明,随着种群数目的增多,所提模型的综合费用趋向于1 200万元左右,低于不计导线截面策略因子的多目标优化方案及单目标优化方案,所提模型与算法能够实现线路长度和导线截面策略因子并行优化,提高解的搜索精度。
A multi-objective function is proposed for the economic optimization of urban distribution network, which,based on the economic current density and multiple objective planning theory,takes the minimum investment cost considering conductor section strategy factor and the minimum yearly network loss cost as its objectives. The genetic algorithm and the annealing dynamic penalty factor are applied to study the methods of real-number-coding for parallel optimization,dynamic penalty function restrictions,self-recognition decoding for power flow and path. Case study for a 10 kV distribution network shows that,the comprehensive cost of the proposed model tends to 12 million yuan along with the increase of selected population,less than that of multi-objective optimization scheme or single objective optimization scheme without conductor section strategy factor. The proposed model and algorithm realize the parallel optimization of line length and conductor section strategy factor,and improve the searching precision of solution.
出处
《电力自动化设备》
EI
CSCD
北大核心
2012年第8期48-53,共6页
Electric Power Automation Equipment
关键词
配电网
多目标规划
遗传算法
经济性
模型
优化
electric power distribution
multiple objective planning
genetic algorithms
economics
models
optimization
