基于二进制自适应微分进化算法的机组组合问题 - 电网技术.pdf

第34 卷 第4 期 电 网 技 术 Vol. 34 No. 4 2010 年4 月 Power System Technology Apr. 2010 文章编号：1000-3673 （2010 ）04-0110-05 中图分类号：TM 711 文献标志码：A 学科代码：470·4051 基于二进制自适应微分进化算法的机组组合问题 夏澍，张建华，刘自发，葛晓琳 （电力系统保护与动态安全监控教育部重点实验室(华北电力大学) ，北京市 昌平区 102206 ） A Binary Adaptive Differential Evolution Algorithm for Unit Commitment XIA Shu, ZHANG Jian-hua, LIU Zi-fa, GE Xiao-lin (Key Laboratory of Power System Protection and Dynamic Security Monitoring and Control (North China Electric Power University), Ministry of Education, Changping District, Beijing 102206, China) ABSTRACT: To solve the unit commitment that is a typical 题，处理起来非常复杂。传统的解决方法有优先顺 non-convex, nonlinear, high-dimensional and discrete 序法[1] [2] [3] 、动态规划法 、拉格朗日松弛法 等。优先 optimization problem, a binary adaptive differential evolution 顺序法计算速度快，但由于其过于简化，一般很难 (BADE) algorithm is proposed. Binary differential evolution 寻到最优解；动态规划法应用灵活，但面临维数灾 (BDE) algorithm possess good robustness and high searching 的问题；拉格朗日松弛法算法能够简化计算维数， efficiency, however it greatly depends on control parameters, therefore an adaptive adjusting strategy for control parameters 但存在对偶间隙问题。近年来，随着智能优化算法 is integrated with (BDE) algorithm to enhance its searching 的发展，很多学者将遗传算法[4-5](genetic algorithm， ability. Meanwhile, based on the characteristic of unit GA)、粒子群优化[6-8](particle swarm optimization， commitment, the individuals that do not conform to the PSO) 算法、社会演化 [9](social evolutionary constraint conditions are modified to make the algor