Precursor Detection to Avoid Serious Incidents in Energy Storage System
To assess the health status of batteries in Energy Storage System (ESS) and detect precursors to negative incidents that can deteriorate the safety and reliability of renewable power supply
As serious incidents happen rarely in ESS, we tackled the problem with novelty detection relying on abundance of normal operation data. Our method is also equipped with interpretability to provide what features or components may have been mainly responsible for the anomalous incident.
Expected to support safe operation and increase the reliability of energy storage and its management