Active Explicit Model Predictive Current Control of Power Conversion System in Grid-Connected BESS

Abstract

An active explicit model predictive current controller (APCC) is developed for the power conversion system (PCS) in grid-connected battery energy storage systems (BESS). The accurate discrete-time model of the PCS is derived, with both the system and input matrices expressed as scaled rotation matrices, significantly simplifying controller design. The current reference tracking problem, considering BESS dc-link voltage variations, is formulated as a multiparametric quadratic program, where the feasible current domain is divided into multiple critical regions. The optimal control law in each region is shown to be a piecewise affine function of the current tracking error and can be pre-computed explicitly, enabling efficient real-time implementation in industrial power drives. A disturbance observer is combined with the explicit predictive controller to reject lumped disturbances, and the offset-free tracking capability of the APCC has been theoretically proven. The proposed controller is validated on a 20 kVA experimental platform using an industrial DSP (TMS320F28335). Experimental results demonstrate that the APCC achieves prompt transient response, offset-free tracking performance in steady state, and enhanced robustness against grid variations and system uncertainties.