A novel digital twin for battery energy storage systems in micro-grids

Nowadays, traditional power grids, conventionally designed with a centralized paradigm, are evolving toward more decentralized configurations with numerous smaller nodes (prosumers), leading to increased interactions and complexity. In this context, digital twins (DTs) come in handy to replicate the behavior of a physical process in a fast, virtual, and safe way. This paper introduces a novel DT of a battery energy storage system (BESS), designed as a general-purpose framework adaptable for employment in complex scenarios. In addition to the DT design, this work presents a practical implementation of the proposed framework, rigorously validated through an extensive experimental campaign using real-world data. Specifically, the model is tested on realistic datasets, including a photovoltaic (PV) power profile, an experimental microgrid setup, and aged battery data. The results demonstrate the DT’s strong emulation capabilities, with a Mean Absolute Percentage Error (MAPE) consistently below 0.50% across all experiments.