Unsupervised Monitoring System for Predictive Maintenance of High Voltage Apparatus

Abstract

The online monitoring of a high voltage apparatus is a crucial aspect for a predictive maintenance program. The insulation system of an electrical machine is affected by partial discharges (PDs) phenomena that—in the long term—can lead to the breakdown. This in turn may bring about a significant economic loss; wind turbines provide an excellent example. Thus, it is necessary to adopt embedded solutions for monitoring the insulation status. This paper introduces an online system that exploit fully unsupervised methodologies to assess in real-time the condition of the monitored machine. Accordingly, the monitoring process does not rely on any prior knowledge about the apparatus. Nonetheless, the proposed system can identify the relevant drifts in the machine status. Notably, the system is designed to run on low-cost embedded devices.