有抽水蓄能电站的联合电力系统优化调度模型和算法

Model and Approach of Optimal Operation for Hydrothermal Power System with Pumped-Storage Stations

本文介绍了一组有抽水蓄能电厂的电力系统优化调度的模型和算法，包括机组的最优起停。用Ｌａｇｒａｎｇｉａｎ松弛法将问题分解为多个子系统，对协调级建议一种灵敏系数法代替常规的次梯度法、速度快适应性好；火电机组组合采用分解的方法，对抽水蓄能电厂的模型作了详细的描述，考虑了上、下游水库水位变化的影响。提出一种最大增益网的算法，可最优选择日抽水量。相应的模型和算法已成功地用于实际系统，并用于运行计划及效益分析的计算．证明了其合理性及有效性。

In this paper,the model and approach of optimal operation for power system with pumped-storage plants are presented,including unit commitment. A sensitivity coefficient coordination algorithm between two levels for Lagrangian relaxation method is proposed instead of conventional subgradient algorithm. The model of pumped-storage is described accurately considering level variations of upstream and downstream reservoirs. A maximum gain-newtwork algorithm is given for optimal choosing the volume of daily pumped water. The model and algorithm have been applied in an actual large-scale system,and proved to be efficient and practical.