Power Transformers are considered critical and costly equipment in electrical power systems. Sudden transformer failures can cause considerable economic losses. Conventional transformer monitoring systems have used periodic manual inspection for transformer monitoring. Such methods may be ineffective for detecting faults at the primary stage. This paper proposes an IoT-based transformer health monitoring system in real-time using ESP32 and Raspberry Pi Zero 2W. The system monitors various critical transformer parameters in real time, including temperature, gas concentrations (Hydrogen, Methane), oil level, load currents, operating voltage, and ultrasonic fault signals. The ESP32 module will be used to collect the transformer parameters and transmit the values to Firebase for cloud storage. In addition, Machine learning (ML) models like Random Forest, KNN, Gradient Boosting, and Support Vector Machine (SVM) are used for fault prediction. The collected data is sent to Firebase, where a web-based dashboard built with ReactJS provides visualization and overall analysis of the system. This system enables early fault detection and extends the transformer's lifespan.
Kalyani et al. (Thu,) studied this question.