A Bi-level Economic Dispatch considering Battery Degradation Prediction Model for Microgrids with Hybrid Energy Storage Systems

The integration of renewable energy sources (RES) and energy storage systems (ESS) offers significant potential benefits to users. However, the intermittency of RES introduces volatility and uncertainty to power systems. To address the limitation of traditional models, which often focus solely on short-term economic benefits and fail to effectively integrate long- and short-term energy management, we propose a two-layer microgrid economic dispatch model based on a hybrid energy storage system (HESS-TEM) comprising battery energy storage (BESS) and compressed air energy storage (CAES). In this model, CAES is leveraged for long-term energy management, while BESS provides short-term power compensation to minimize prediction errors. Additionally, a neural network-based battery degradation cost prediction model (NN-BDC) is incorporated to capture the nonlinear degradation dynamics of batteries by factoring in key indicators, including environmental temperature, charge-discharge rate, state of charge, state of health, and depth of discharge. Experimental results demonstrate that HESS-TEM achieves strong economic performance and stability under various pricing schemes, with NN-BDC showing a notable advantage in accurately capturing complex battery degradation behaviors. This model provides an effective solution to enhance the economic viability and stability of microgrids with high RES penetration.