基于多变量维纳滤波的中低压电网拓扑学习方法

Topology learning method for medium and low voltage power grid based on multivariate Wiener filtering

为了实现电网的最优市场结算和可靠的动态运行,需要对电网的运行拓扑情况进行估计.提出了一种新的通用电网拓扑估计框架,利用摆动动力学引起的节点电压相位角的时间序列测量进行拓扑估计.利用多元维纳滤波来分解不同节点电压角波动之间的相互作用,并通过考虑多元维纳滤波器元件的相位响应来识别操作边.在标准IEEE测试用例上的仿真测试结果表明,该学习框架的性能良好,达到了预期的设计目标.

In order to achieve optimal market settlement and reliable dynamic operation of power grid,it is necessary to estimate the operation topology of power grid.A new topology estimation framework for general power grid was proposed,and the time series measurement of node voltage phase angle caused by swing dynamics was used to estimate the topology.The multivariate Wiener filtering was adopted to decompose the interaction among voltage angle fluctuations at different nodes,and the operating edge was identified by considering the phase response of multivariate Wiener filtering elements.The simulation results for standard IEEE test cases show that the as-proposed learning framework has good performance and achieves the expected design goals.