Earth Leakage Detection and Earth Resistance Monitoring Using Isolation Tree Model

Electric street poles installed in parking areas and station areas, the lights are equipped with high protection switch gear even though when the protection gear get any malfunction such as electric shock, fire hazards and equipment damage it is difficult to identify and requires large manual power for inspection. For this in this project a sensor based machine learning model is used to detect earth leakage current, earth continuity and earth resistance. The proposed solution includes Isolation Forest technique to detect anomalies and integrate with cloud such as ThingSpeak which is used for collecting data from sensor, data visualization and remote monitoring. The system also include automated street lightning and shutdown the system if faults occur in the system.