Robust gain-scheduled feedforward compensator with quasi-LPV model for buck converters in abrupt input voltage disturbances.

This paper presents a robust control strategy for DC-DC Buck converters operating under sudden input voltage disturbances. A novel quasi-Linear Parameter Varying (LPV) model is constructed by incorporating parasitic elements and expressing system dynamics as functions of the duty cycle to capture real-world dynamics more accurately. Based on this model, a gain-scheduled LPV feedforward compensator is synthesized using affine parameter-dependent Linear Matrix Inequalities (LMIs) to achieve H_∞ disturbance attenuation performance. The proposed feedforward compensator adapts continuously to duty cycle variations using only input voltage measurement, thereby simplifying implementation without compromising control performance. This feedforward structure is integrated with a generic feedback controller, enhancing the system's ability to reject abrupt input disturbances effectively. Experimental results demonstrate superior performance. Compared to a conventional PI controller, our proposed method achieves a 22.46 % higher Disturbance Rejection Ratio (DRR), 65 times higher Attenuation Ratio (AR), 83.49 dB higher Power Supply Rejection Ratio (PSRR), and an impressive 98.46 % reduction in output voltage ripple (from 38.235 % to just 0.588 %). Furthermore, it allows only 0.35 % of the disturbance to pass to the output (Disturbance Transmission Gain - DTG). Comparative analysis confirms this approach outperforms other state-of-the-art methods across all disturbance rejection metrics, delivering significantly higher AR and PSRR, and substantially lower voltage ripple. Our controller maintains a 99.65 % DRR even under a challenging 146.15 % input voltage disturbance, highlighting that it exhibits superior robustness. This makes it highly applicable for demanding power electronics systems in areas like renewable energy, electric vehicles, and industrial power converters.