Detection of Abnormal Load Consumption in the Power Grid Using Clustering and Statistical Analysis

摘要: Nowadays, the electricity load profiles of customers (consumers and prosumers) are changing as new technologies being developed, therefore it is necessary to correctly identify trends, changes anomalies in data. Anomalies consumption can be caused by abnormal behavior or a failure smart meters grid. Accurate identification such crucial for maintaining stability grid reduce loss distribution companies. Smart produce huge amounts measurements every day analyzing all computationally expensive very inefficient. Therefore, aim this work propose an anomaly detection method, that addresses issue. Our proposed method firstly narrows down potential anomalous large datasets clustering discretized time series, then analyses selected using statistical S-H-ESD calculate final score. We evaluated compared our four state-of-the-art methods on created synthetic dataset series containing collective anomalies. outperformed other terms accuracy.