发输电协调检修计划的主从规划模型与分区搜索算法  

A Bilevel Programming Model and Partition Searching Algorithm for Integrated Maintenance Scheduling

作　　者：于大洋[1] 韩学山[1] 赵建国[2] 

机构地区：[1]山东大学电气工程学院,山东省济南市250061 [2]国家电网技术学院,山东省济南市250002

出　　处：《电网技术》2010年第4期88-93,共6页Power System Technology

基　　金：国家自然科学基金资助项目(50677036)~~

摘　　要：提出了针对发输电协调检修计划的主从规划模型,主问题以总成本最小为目标,主、从问题分别采用分区禁忌搜索(partition tabu search,PTS)法和线性规划方法求解。为解决主问题解空间的"平坦区域"内寻优效率低的问题,PTS算法将主问题的解空间分为发电和输电2个子空间,并按一定概率对子空间分别进行搜索,每次迭代中,PTS算法根据子空间的邻域特征,调整概率分布参数以减少对"平坦区域"的搜索,提高了寻优效率。算例结果验证了模型与算法的有效性。A bilevel programming model and a new partition tabu searching algorithm for integrated maintenance scheduling are presented. The leader problem takes the minimum of total cost as the objective; and the leader problem and the follower problem are solved by partition tabu searching （PTS） algorithm and linear programming respectively. To solve the low search efficiency in the flat region of leader problem solution space, the PTS algorithm divides the solution space of leader problem into two subspaces, i.e., the generation and transmission, and with a certain probability the two subspaces are searched respectively. After each time of iteration, the PTS algorithm adjusts the probability distribution parameters according to the neighbourhood features of the two subspaces to reduce the search times in flat regions, thus the search efficiency can be increased. Calculation results of IEEE RTS show that the proposed model and algorithm are effective.

关 键 词：发输电协调检修计划 主从规划 分区禁忌搜索 平坦区域 

分 类 号：TM726[电气工程—电力系统及自动化]