基于混合双种群差分进化的电力系统经济负荷分配  

Economic distribution of power systems based on hybrid differential evolution with double populations

作　　者：王凌[1,2] 黄付卓[1,2] 李灵坡[1,2] 

机构地区：[1]清华大学自动化系 [2]清华大学信息科学与技术国家实验室,北京100084

出　　处：《控制与决策》2009年第8期1156-1160,1166,共6页Control and Decision

基　　金：国家自然科学基金项目(60834004;70871065;60774082)

摘　　要：针对电力系统经济负荷分配本质上的非线性约束优化问题,提出一种双种群混合差分进化算法.采用两个种群且以较小的计算量实现目标函数的寻优并保持解的可行性,同时引入单纯型法来提高算法的局部搜索能力.基于典型算例对该算法的进化行为进行测试,并通过仿真和比较验证了所提出算法的有效性.Economic distribution of power systems can be regarded as a non-linear constrained optimization problem. In this paper, a hybrid double-population differential evolution is proposed, which uses two populations to optimize objective function and to maintain feasibility of solutions with a little computational effort. In addition, simplex method is embedded to enhance the exploitation ability. The evolving behavior of the algorithm is tested based on a typical instance. Simulation and comparison show the effectiveness of the proposed algorithm.

关 键 词：经济负荷分配 约束处理 差分进化 双种群 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]