基于场景分析的电力系统日前调度及其快速求解方法

Day-ahead dispatch and its fast solution method of power system based on scenario analysis

在含风光电力系统规模逐渐增大的背景下,提出一种基于场景分析的电力系统日前调度快速求解方法。考虑到不同时刻风光出力均具有显著相关性,基于多元正态分布和蒙特卡罗采样生成大量具有时间相关性的原始场景。利用改进的k-means聚类算法对原始场景进行预分类,并采用基于Kantorovich概率距离的同步回代缩减算法对处理后的场景进行缩减,缩短场景分析的计算时间。建立基于场景分析的电力系统日前优化调度模型。为提高模型求解效率,引入包含风光预测误差向量信息的边界场景,在调度模型中考虑上下边界场景的备用容量约束,并建立考虑边界场景备用容量约束的优化调度模型。以某省级电网实测数据进行仿真分析,验证了所提模型及方法的有效性。

Under the background of gradual increase in the scale of power system with wind and photovoltaic power,a fast solution method of day-ahead dispatch for power system is proposed based on scenario analysis.Considering the wind and photovoltaic power at different times are of significant correlation,a large number of original scenarios with time correlation are generated based on multivariate normal distribution and Monte Carlo sampling.An improved k-means clustering algorithm is used to pre-classify the original scenarios,and the simultaneous backward reduction algorithm based on Kantorovich probability distance is adopted to reduce the processed scenarios,which reduces the calculation time of scenario analysis.A dayahead optimal dispatch model of power system based on scenario analysis is established.In order to improve the solution efficiency of the model,the boundary scenarios containing predicted error vector information of wind and photovoltaic power are introduced,the reserve capacity constraints of upper and lower boundary scenarios are considered in the dispatch model,and an optimal dispatch model considering the reserve capacity constraints of boundary scenarios is established.The measured data of a provincial power grid is taken for simulation and analysis,verifying the effectiveness of the proposed model and method.