适合于机组组合问题的贪婪随机自适应搜索模型  

A Greedy Randomized Adaptive Searching Model for Unit Commitment Problems

作　　者：黎静华[1] 韦化[1] 

机构地区：[1]广西大学电气工程学院,广西壮族自治区南宁市530004

出　　处：《电网技术》2010年第4期119-123,共5页Power System Technology

基　　金：广西大学科研基金项目(DD020015)的资助

摘　　要：提出了一种求解机组组合问题的方法——贪婪随机自适应搜索过程(greedy randomized adaptive search procedure,GRASP)。该方法迭代过程分为解的构造和局部搜索2阶段。在解的构造阶段,构造贪婪函数,形成带限制的候选列表,然后随机选择列表中的成分形成初始组合,作为局部搜索的起始解。在邻域搜索阶段,提出了一种适合于机组组合问题的变邻域算法,利用此法对邻域搜索得到局部最优解,取迭代过程中的最好局部解作为问题的最优解。算例仿真结果验证了所提方法的优越性。GRASP算法不仅为解决机组组合问题带来了新的思路和方法,同时给求解其它含复杂约束条件的组合优化问题带来了生机。A method to solve unit commitment, namely the greedy randomized adaptive searching procedure （GRASP）, is proposed. The iteration procedures of the proposed method divided into two stages, i.e., the construction of the solution and local searching. In the stage of contraction of the solution, the greedy function is constructed to form the candidate list with restriction, and then the initial combination is formed by randomized selection of the composition in the list and is taken as the starting solution of local searching. In the stage of neighborhood search a variable neighborhood algorithm suitable for unit commiment is put forward and local optimal solution is obtained by neighborhood search, then best local solution is taken as the optimal solution of the problem. The superiority of the proposed method is verified by the results of calculation example. The proposed GRASP algorithm brings a new thinking and approach for the solution of unit commitment, meanwhile it open up a new way for the solution of other unit commitment problems containing complex constraints.

关 键 词：混合整数规划 机组组合 优化运行 贪婪随机自适应搜索过程 

分 类 号：TM743[电气工程—电力系统及自动化]