Research on rapid parameter identification in DC-DC converter digital twin systems considering phase deviation

Abstract

This paper addresses the issue of lag caused by the sampling phase deviation in BUCK converter digital twin-parameter identification, and proposes a phase compensation method to improve digital twin-parameter identification. After sampling the characteristic signals of the BUCK converter, the proposed method introduces a Hilbert transform to compensate for the phase deviation in both the physical converter and the digital twin converter signals. By performing phase compensation before calculating the fitting degree of the digital twin, this method reduces the impact of phase deviation on the fitting degree calculation, allowing the fitting degree to more accurately reflect the true differences among converter signals. This reduces the computational redundancy caused by phase deviation and enables the digital twin converter to promptly update parameters. In addition, a BUCK converter digital twin platform is designed for experimental validation. Experimental results demonstrate that the proposed improved digital twin-parameter identification method saves 50% of optimization iterations when compared to traditional methods without compromising accuracy. This indicates that parameter identification can achieve accurate results with fewer optimization iterations, effectively enhancing the speed of digital twin converter parameter identification and facilitating subsequent converter-field coupling analysis.