Development of an Intelligent Energy Management System to Improve BESS State of Health

ABSTRACT The energy management system (EMS) is becoming a focal point of research in the renewable energy sector, especially when integrating PV solar systems, BESS, and a standby diesel generator of a microgrid (MG). The EMS controls, monitors, and manages the power dispatch of different integrated sources based on the strategy that balances the demand with the supply, with the available energy sources related to solar and BESS. The EMS application strategy directly affects the BESS SOH and, thus, increases its operational remaining useful lifetime (RUL). Therefore, this study develops an intelligent EMS (iEMS) implemented within the MG based on predictive artificial neural network (ANN) control power dispatch strategy. The proposed iEMS has proved to be effective and accurate (MAE = 6.43%) which improved and optimized the BESS SOH through the prosecution of a 6‐h prediction on blackout occurrence and noncritical load shedding. Consequently the iEMS preserved the SOH to 33% (an increase of 45% compared to classical EMS) and decreased the blackout occurrence by 56% (−5203 h) in contrast with the classical EMS where the SOH reached 20% and blackout occurrences totaled 11,899 h. This proves that the model is effective, and the control logic avoids high loads being dispatched from the BESS at critical time intervals where the AI model predicts a blackout occurrence.