结合网络重构的主动配电网日前无功电压双层优化

Bi-level optimization of day-ahead reactive-voltage in active distribution network with network reconfiguration

高渗透率接入的分布式光伏出力的不确定性增加了配电网无功电压控制的控制难度,尤其是对日前的无功电压控制,若控制数量过多,会给运行人员带来繁重的操作任务。为此,建立了一种主动配电网日前无功电压控制数学模型,该模型将配电网网络重构技术与无功电压优化方法相结合,根据配电网的实际运行情况,实现网络重构和无功电压的协调优化。为求解上述模型,提出了一种双层优化的控制策略:第一层建立了以日前系统网损最小为目标的网络重构数学模型,为减少日前重构次数,根据极限净负荷对电压分布的影响,基于层次聚类方法对日前24 h进行时段合并,对各时段分别进行网络重构,用以应对分布式光伏出力和负荷大幅随机变化;第二层建立了以电压偏差最小为目标的无功电压优化模型,通过第一层优化后的网络拓扑结构与各小时级无功电压优化配合,以应对分布式光伏出力和负荷的小幅波动。最后,在IEEE 33节点系统上进行仿真算例分析,验证了所提方法的有效性。

The uncertainty of distributed photovoltaic (PV) power output with high penetration access makes it very difficult to control the reactive voltage,especially for the day-ahead reactive-voltage control.Too many controls bring heavy operating tasks for the operators.To this end,a day-ahead reactive-voltage of the active distribution network model is presented in this paper which combines the network reconfiguration technologies and reactive-voltage optimization method.The coordinated optimization of network reconfiguration and reactive-voltage control is realized according to the operating condition in the practice of distributed network.To solve the above model,a bi-level optimization strategy is proposed in this paper.In the first level,a day-ahead network reconfiguration model is proposed aiming at minimizing the total network loss.In order to reduce the number of reconfiguration,according to the influence of the limit net load on the voltage distribution,the time periods of the day ahead 24 hours are combined based on the hierarchical clustering method,and the network reconfiguration of each time period is carried out respectively to cope with the large random changes in the output and load of distributed photovoltaics.In the second level,the reactive-voltage optimization model is established aiming at the minimum voltage deviation.The optimized network topology of the first level is cooperated with the optimization of reactive-voltage at each hour level to cope with small fluctuations in distributed photovoltaic output and load.Finally,the effectiveness of the proposed method is verified by the simulation arithmetic analysis on the IEEE 33-node power system.