基于动态规划算法的故障恢复重构

摘要

A new method is presented to restore the power system based on dynamic programming （DP）. After isolated a fault, the out-of service area, which is constructed of load nodes could be divided into several groups, each has one contact branch used for load distribution. These groups sharply decrease the number of calculation loops. According to the different restore orders of the load nodes, operation sequences are produced, which turn the restore problem into a node-string choice. Based on a DP method, operation sequences are used to build a multiple-stage decision model, each contact branch is calculated as one stage, the algorithm is designed to transfer and restore the largest load capacity by optimal choice of the node-string. In each stage, the calculation process records the optimal state and lists the decision variable. The optimal result is obtained at the final stage. This method can be used to power service reconfiguration, when the power capacity is not enough to support all of the out-of service nodes, reducing the time of outage and improve the reliability of system. The proposed method is validated by simulation methods.%提出了一种基于动态规划算法的电网恢复重构方案，将配电网的故障恢复重构过程转化为一个多阶段决策过程并设计了动态规划算法进行求解。该方案以恢复最大容量的非故障失电负荷为目标，通过把非故障失电区的多个负荷节点分割为不同的恢复区，并根据每个分区中不同节点的恢复供电顺序生成计算序列组。以此将非故障失电负荷节点的恢复供电过程构建成适用于动态规划方法求解的重构计算模型。重构算法利用对分区方案的选择来降低整体计算复杂度，并采用有效节点串的提取来减少算法组合数和提高计算效率。文中还应用提出的算法对33节点和69节点的测试系统进行了恢复重构计算，给出了算例比较分析，验证了该方案的有效性。