基于改进多种群遗传算法的节能环保多目标优化模型  

Multi-Objective Optimization Model Based on Improved Multiple Population Genetic Algorithm Considering Energy Conservation and Environmental Protection

作　　者：韦杏秋[1,2] 陈碧云[1,2] 陈绍南[1,2] 

机构地区：[1]广西大学电力系统最优化研究所,广西南宁530004 [2]广西电力系统最优化与节能技术重点实验室,广西南宁530004

出　　处：《电网与清洁能源》2013年第12期22-29,共8页Power System and Clean Energy

基　　金：国家自然科学基金项目(51107011)~~

摘　　要：电力系统多目标经济负荷分配问题是个非线性、高维的复杂优化问题。提出基于交互式的改进多种群遗传算法,通过引入精英策略和移民策略的多种群遗传算法可以有效地克服标准遗传算法容易陷入局部最优解、易早熟的缺陷。针对文中提出的煤耗和排放2个目标函数,提出了基于目标满意度和总体协调度的交互式多目标处理方法,通过寻求向量空间内满足总体协调度的最短"欧氏距离",来贴近决策者满意的理想值,解决了各目标函数之间最优解的相互冲突,达到协调好各个目标函数的目的,充分体现了决策者的意愿。试验算例表明,该算法能够有效地解决电力系统多目标经济负荷分配问题。The economic load dispatch in the power system is a nonlinear and complicated question of high dimension. This paper proposes the improvement based on the interactive genetic algorithm method. With introduction of the elite strategy and immigration strategies, the muhiple population genetic algorithms can efficiently overcome the problems of local optimal solution and early maturation which the standard genetic algorithm tends to fall into. For the two target functions of coal consumption and emission as mentioned in the paper, an interactive multi-objective processing method based on the target satisfaction and overall coordination degree is proposed. By seeking the vector space to meet the shortest ＂euclidean distance＂ of the overall coordination degree, the ideal value satisfactory to the decision maker can be reached and conflicts between the objective function and the optimal solution be solved so as to coordinate the each objective function and embody the decision makers＇ intentions. The test examples show that this algorithm can effectively solve the power system muhi-objeetive eeonomic load dispatch problems.

关 键 词：节能减排 经济负荷分配 改进多种群遗传算法 多目标计算 

分 类 号：TM732[电气工程—电力系统及自动化]