梯级水电站水库群优化调度的网络规划分解协调法(NP-DC)

THE NETWORK PROGRAMMING DECOMPOSED COORDINATED METHOD FOR OPTIMAL MISPATCHING OF CASCADE HYDROPOWER STATION RESERVOIRS

本文提出了求解水库群优化调度的一种新方法——网格规划分解协调法(NP-DC)、其步骤是首先把库群调度的动态问题转化为一张静态网络图,然后按照大系统优化理论将网络沿时间座标分解为相互关联的子网络,并建立一个三级分解协调结构;子网络分别用OKA算法求解;引入到协调器中的关联变量采用POA法预测寻优。以实现系统的整体最优化。本方法可使网络流优化技术适用于多库、多目标、多时段以及目标函数不可微的复杂库群调度问题。

A new method for long-term optimal dispatching of reservoirs in series. 'The Network Programming—Decomposed Coordinated Method' (NP-DC) is presented in this paper. The move is: first, change the dynamic dispatching problem into a static network chart; second decompose the network into several incidence subsystems along the time axis according to the large scale system theory and estalish a three level decomposed coordinated structure; third, optimize each subsystem with Out-of-Kilter Algorithm(OKA); finally, lead the incidence variables into the upper level coordinator and optimize them with Progressive Optimal Algorithm (POA) to achieve global optimum. NP-DC is applicable in complex dispatching problem of multireservois, multipurpose multiperiod and having nondifferentiable objective function.