Demo: Intelligent Abnormal Detection for Electrical Equipment Based on Machine Learning

Electrical equipment is essential for factories, offices, and daily life. Neglecting its proper maintenance can result in fires or hazardous accidents. Currently, traditional electrical plugs can only confirm whether the electrical equipment is overloaded and cannot detect hazards or aging abnormalities in real-time. Therefore, an intelligent abnormal detection system for electrical equipment is required. Such a system needs to detect electrical information of the equipment, including voltage, current, power factor, and frequency, and combine it with AI techniques to dynamically detect abnormal conditions in the electrical equipment. In view of this, we proposed and developed an AIoT system with the following functional components: 1. Smart power sensing devices for collecting and monitoring abnormal events. 2. A real-time anomaly alerting app to receive abnormal notifications. 3. A highly efficient and fast anomaly recognition model, capable of identifying various abnormal conditions in electrical equipment. 4. A visualized electrical equipment analysis web platform for managing all electrical equipment in factories, buildings, and offices. Experimental results show that our AIoT system achieves accurate recognition results.