In common with most industries, maintenance of solar photovoltaic (PV) plants is generally done using preventative or corrective methods. This means inspections and replacement of components at regular intervals or in response to a sudden fault. Most problems are caused by inverter failures.
Predictive maintenance solution
DNV and GreenPowerMonitor, a DNV company, have developed a predictive maintenance system for solar inverters that uses machine learning models to represent an inverter’s normal operation and to identify anomalous behaviour within new streaming data. The developed system has been successfully tested. However, to work properly it needs detailed maintenance logs which the majority of systems lack. This has prevented further development and validation of the method.
As a solution to the lack of detailed logs, DNV and GreenPowerMonitor have explored a more general anomaly detection approach which can be exploited when no maintenance logs for validation of the predictive maintenance system are available. The anomaly detection system relies on the concepts of density and reconstruction error to identify anomalies, and it is complemented with the use of explainability techniques to determine the potential causes of anomalies. The system has been tested using historical data for multiple inverters and results show that the implemented approach can identify sustained anomalies. Anomaly detection can also play an important role as a support tool for a predictive maintenance system but should not be seen as a substitute.