电力系统大停电后机组恢复的多目标优化方法

A Multi-Objective Optimization Method for Post-Blackout Unit Restoration

大停电后的机组恢复策略多为单目标优化,忽略了大量影响机组启动顺序的因素,为此提出了机组恢复的多目标优化策略。以机组启动后提供的发电量尽可能大、已恢复的电源点尽量在网架层面铺开、有利于后续厂站层机组和重要负荷的恢复为优化目标,建立机组恢复的多目标优化模型。先将恢复过程划分为一系列顺序执行的恢复时步,再将每一时步的优化问题转化成多目标"0/1背包"问题;结合最短路径法为机组恢复方案搜索恢复路径,采用快速非支配排序遗传算法进行求解;最后对每时步的Pareto最优方案排序,确定最优解。算例结果验证了该方法的有效性。

Most post-blackout unit restoration strategies are based on single-objective optimization and factors influencing the starting sequence of units are neglected.A multi-objective optimization strategy for post-blackout unit restoration is proposed and a multi-objective optimization model for unit restoration,in which following items are taken as optimization objectives： the generating capacity of restored units should be as much as possible,the restored units should be distributed in the network frame level as possible and the restored units are favorable to the restoration of important loads and follow-up units to be restored in power plant level,is built.Firstly,the restoration process is divided into a series of restoration steps to be executed sequentially;secondly,the optimization of each restoration step is turned into multi-objective 0/1 knapsack problem and combining with the shortest path method the restoration paths for unit restoration scheme are searched,and then the restoration paths are solved by fast and elitist non-dominated sorting genetic algorithm and classic Dijkstra algorithm is employed to select restoration paths;finally,the Pareto optimal solutions are sorted to determine the optimal solution.The effectiveness of the proposed method is verified by simulation results of New England 10-machine 39-bus system.