99 / 2025-04-16 20:21:29

Enhanced Model Predictive Control for Hybrid Energy Storage Systems in Microgrids Using ESO for Real-Time Disturbance Compensation

Model Predictive Control, Extended State Ob-server, Current ripple, Disturbance Observation

Hybrid Energy Storage Systems (HESS) coordinate the operation of batteries and supercapacitors, enabling rapid energy absorption and release in microgrids. By effectively smoothing ffuctuations between renewable generation and load demand, they ensure a stable and reliable power supply. Consequently, efffcient control strategies for HESS are critical to enhancing overall microgrid performance. Model Predictive Control (MPC), with its ability to handle multivariable constraints and predict future behavior, has demonstrated excellent performance in HESS control. However, conventional MPC typically assumes that the battery voltage remains constant throughout the prediction horizon, leading to accumulated model errors and inaccurate predictions. To address this issue, this paper proposes an Extended State Observer Model Predictive Control (ESOMPC) method. In this approach, the unmodeled battery voltage dynamics are treated as an unknown disturbance, with ESO used for real-time estimation and compensation, thereby updating the system state and improving control accuracy. Simulation results show that the ESO-MPC method signiffcantly reduces DC bus voltage ripple and battery current ripple in an HESS microgrid, thus enhancing system stability and power quality.