求解机组组合问题的领域搜索法 - 中国电机工程学会.pdf

第28 卷 第13 期 中 国 电 机 工 程 学 报 Vol.28 No.13 May 5, 2008 2008 年5 月5 日 Proceedings of the CSEE ©2008 Chin.Soc.for Elec.Eng. 33 0258-8013 (2008) 13-0033-08 TM 76 A 470 40 文章编号： 中图分类号： 文献标识码： 学科分类号： ⋅ 求解机组组合问题的领域搜索法 黎静华，韦 化 (广西大学电气工程学院，广西壮族自治区 南宁市 530004) Unit Commitment via Local Search Point Method LI Jing-hua, WEI Hua (College of Electrical Engineering, Guangxi University, Nanning 530004, Guangxi Zhuang Autonomous Region, China) ABSTRACT: Unit commitment is an important and difficulty 其计算速度。对 100 台机组24 时段仿真结果表明，CPU 计 aspect of optimal operation of power systems. Because it 算时间仅为原来的4 s，所耗费用却大大降低；同时该方法收 belongs to be mixed variables programming that is difficult to 敛速度快、精度高，尤其适合于求解大规模机组的组合问题。 find the optimal solution in mathematics. This paper presented 关键词：混合整数规划；机组组合；内点法；优化运行；电 a new method combining of neighbor search and interior point 力系统 to solve unit commitment (UC) nondeterministic polynomial 0 引言 (NP)-hard problem. The structure of neighborhood is defined and a new way to adjust neighborhood was put forward in 从传统电力体制到电力市场，UC 问题一直是 which various restrictions can be dealt with to ensure the result 电力系统优化运行的一个重要方面。几十年来，电 feasibility. The large and complex discrete space which is even 力科学工作者几乎尝试应用了所有优化算法解决 difficult to obtain solutions will be replaced by a very small [1-7] UC 问题 。但至今为止，仍未能找到一种理想的 solution neighborhood. The advantages of interior point 解决办法。究其根本原因，UC 问题是一个既包含