Abstract
Faults in distribution and power transformers jeopardize stability of the power network. Hence, various diagnosis techniques are implemented in order to prevent or at least detect transformer integrity violations. The majority of diagnosis techniques are functioning off-line and requires transformer disconnection from the power line. This is certainly undesirable for utility management and customer. Therefore, on-line or online diagnosis is more preferable and faster than off-line monitoring procedure. The aim of this study is to implement transformer real-time diagnosis technique based on the analysis of the vibrational signal spectrum. It is supposed that vibrational signature of the transformer is transferred and processed over the cloud environment using Internet of Things (IoT), and then the prognosis algorithm is executed over portable device.