Large-scale performance evaluation of energy management strategies for hybrid energy storage systems with a focus on system sizing

This study presents a quantitative performance evaluation of energy management strategies (EMS) for hybrid energy storage systems (HESS) with a focus on system sizing. It addresses the lack of systematic and comparable studies across EMS types.A technology-agnostic linear storage model is employed. Four representative EMS are evaluated, namely filter-based, deadzone-based, model predictive control (MPC)-based, and fuzzy logic-based approaches. These are benchmarked against a reference dispatch optimization with perfect forecast. The study is conducted across multiple storage configurations and representative load profiles, resulting in approximately 10000 investigated setups. For each setup, EMS parameters and system size are optimized using evolutionary algorithms.A clear performance ranking of the EMS emerges. Deadzone-based EMS perform best, followed by fuzzy logic-based, MPC-based, and filter-based EMS. Performance differences are small for most EMS, with required relative system overdimensioning typically between 3% and 10%. EMS performance is governed by load profile characteristics. Load profile variability is identified as the dominant predictor. All EMS exhibit worse sizing performance with increasing variability. Different dispatch strategies yield similar system sizing, indicating low sensitivity of sizing to dispatch details. Optimized EMS parameters show strong clustering, indicating robustness and enabling the derivation of heuristics for EMS parameter selection.The results imply that EMS tuning and system sizing can be decoupled with limited performance loss under the adopted model and sizing objective. This enables a separation of the HESS design process into sizing and EMS design, improving transparency, explainability, and modularity. Deadzone-based EMS provide strong baseline performance and constitute a benchmark for future EMS development.