Remote Intelligent Diagnosis Technology for Leakage Current Based on Smart Measurement Devices

As user-side loads become more complex, the safety risks associated with leakage currents have increased. Neutral-wire miswiring, a common incorrect wiring method, can cause abnormal leakage current detection, leading to inaccurate leakage status readings. Most existing research focuses on leakage current analysis under normal wiring, with little attention to the effects of improper wiring. This paper examines the characteristics of neutral and live wire currents, as well as leakage currents, under three common types of improper wiring. Using the DBSCAN clustering algorithm, abnormal leakage current data is identified and categorized into individual and clustered abnormal users. For these abnormal user clusters, improper wiring is detected through the relationship between neutral and live wire currents and a multiple linear regression model. The proposed method is validated with experimental data from a single-phase smart IoT energy meter.