基于数据驱动的电能质量分区治理策略

Regional Abatement Strategy for Power Quality Based on Data Driven

针对现代电网分布式电源及非线性负荷的高渗透性、分散性和动态时变性,传统治理方式难以有效控制全网电能质量的问题,该文从数据关联分析和污染全网治理的角度出发,提出一种基于数据驱动的电能质量分区治理策略。通过提取网络各节点电能质量时间序列中的重要特征点,配合互插值寻优法构建等时维序列数组。采用灰色关联理论分析序列数组内节点数据的相关性,依据节点间电能质量关联度将网络划分为多个控制区域,同时确定出各分区的主导节点,进而通过分区调控实现电能质量污染的全局治理。IEEE14节点仿真算例验证该文方法用于划分电能质量治理区域的准确性及分区改善全网供电质量的合理性。

For the high-permeability, dispersion, and time-variation of distributed generation and nonlinear load, it is difficult to effectively control power quality(PQ) of the whole network by traditional abatement, from the perspective of data relational analysis and whole network abatement, a regional abatement strategy of PQ based on data driven was proposed.The important feature points were extracted from PQ time series of each node, and interpolation optimization method was used to constructed a sequence array of equal time dimension. Data correlation of nodes was analyzed by grey relation theory, then the network was divided into multiple control regions based on relational degree, the best abatement point of each region was identified, thereby PQ abatement of the whole network was achieved by regional control. IEEE 14 nodes simulation result verifies the accuracy of dividing abatement regions and the rationality of improving PQ of whole network in each region.