Power transformers must be reliable otherwise the modern power system will break down due to lack of supply. Most monitoring techniques do not monitor data in real time or analyze and predict potential equipment failures in advance. This work proposes a cloud-based framework for monitoring transformer health and predicting failure. This system uses sensors enabled by IoT to continuously capture other parameters like oil temperature, winding temperature, load current, voltage and dissolved gases. Information is sent to the cloud through gateways and the software embedded within does analysis and machine learning on it to predict faults.
M et al. (Wed,) studied this question.