基于变量降维的大电网经济调度优化方法

摘要

With the development of power grid's scale,because of the increase of variables and huge dimensions,it will be increasingly difficult to solve the economic dispatch problem.In order to overcome the difficulty,the Benders decomposition method is used to decompose the security constrained economic dispatch problem into the main optimization problem without security constraints and the subordinate power flow problem with security constraints.The security constraints on the subproblem are returned to the main problem by model reconstruction.And the original problem can be solved through coordination between main problem and sub-problem.In the process of solving the optimization problem,the adjusted values of units are distinguished according to current operation information.The units with small safety and economic contributions are centrally addressed.At the same time,the optimization time intervals are tested according to the physical characteristics of power flow and load change on the important section,and the unimportant neighbouring time intervals are simplified and combined.Then the number of unit variables and time intervals can be reduced,so as to lower the goal of solving dimensions.Finally,the results show that the proposed optimization method can greatly reduce the optimization time while retaining a high precision required by engineering calculation.%随着电网规模扩大,经济调度问题因变量增多,维度巨大,造成求解困难.文中利用Benders分解法,将大电网安全约束经济调度问题分解为无安全约束的优化主问题和网络潮流安全约束子问题,将子问题安全约束返回主问题进行模型重构,通过主子问题协调最终实现问题求解.优化问题求解过程中,根据当前运行信息,对机组调整价值进行甄别,将其中对安全和经济贡献小的机组进行集中处理,同时根据重要断面潮流负荷变化的物理特征对研究时段进行检验,将非重点关注的相邻时段进行简化合并,实现压缩机组变量和时段数量、降低求解维度的目的.最后,算例验证表明基于变量降维的优化方法可以减少大电网优化时间,同时保持较高的求解精度,满足工程化计算要求.