Implementation And Tuning a Model Predictive Control Scheme for a Single Ended Primary Inductor Converter


Creative Commons License

Rajaona A. M., AKYAZI Ö., DANAYİYEN Y.

2025 9th International Symposium on Electrical and Electronics Engineering (ISEEE), Galati, Romanya, 30 Ekim 2025, ss.1-6, (Tam Metin Bildiri)

Özet

 In this paper, an optimized Model Predictive Control (MPC) strategy is proposed for the Single-Ended Primary-Inductor Converter (SEPIC) operating under parameter and load variations. The controller is designed using a discrete-time state-space model to regulate the output voltage Vo, input voltage Vin, and load resistance Rload while satisfying duty-cycle constraints. Initially, random values of the control horizon Nc, prediction horizon Np, output error weight Q, and control effort weight Rmpc were applied. The system performance was then analyzed using the Integral of Time-weighted Absolute Error (ITAE) criterion, represented through a 3D surface plot of ITAE versus Q and Rmpc. This analysis enabled the identification of optimal parameters that improved system stability and dynamic performance. The optimized MPC was subsequently tested under variations in input voltage, output reference, and load resistance, maintaining accurate tracking, fast settling, and minimal overshoot. The results confirm that the proposed ITAEdriven MPC effectively enhances SEPIC converter performance under dynamically changing conditions.