Power distribution network topology identification method and system based on approximate Markov blanket

The invention provides a power distribution network topology identification method and system based on an approximate Markov blanket. The method comprises the steps that original voltage and power data of a power distribution network system are acquired and preprocessed; establishing an approximate Markov blanket topology identification model based on the obtained original voltage and power data of the power distribution network system and the candidate power distribution network topology structure parameter set; and solving the approximate Markov blanket topology recognition model by adopting an expectation maximum value algorithm to obtain the current power distribution network topology structure parameters. And accurate and rapid identification of different types of power distribution network topological structures can be realized. 本发明提出了一种基于近似马尔科夫毯的配电网拓扑识别方法及系统，包括：获取配电网系统原始电压、功率数据并进行预处理；基于所获取的配电网系统原始电压、功率数据和候选配电网拓扑结构参数集合建立近似马尔科夫毯拓扑识别模型；采用期望最大值算法对近似马尔科夫毯拓扑识别模型进行求解，得到当前配电网拓扑结构参数。可实现对不同类型配电网拓扑结构进行准确和快速的识别。