Nonparametric Frequency Response Identification for Dc-Dc Converters Based on Spectral Analysis with Automatic Determination of the Perturbation Amplitude

Abstract

Digital control for high switching frequency converter enables new features on DC-DC power conversion for a minimum cost. Frequency response identification is one such enabled functionality used in auto tunning, measurement of components to assess the converter’s state of health, or system stability monitoring. High accuracy, flexibility to operate in open or closed loop, and minimum impact in the converter’s regular operation are the frequency response identification system’s goals. We propose in this paper a nonparametric identification system addressing these main goals. First, it can autoadjust the perturbation size to reduce the perturbation’s impact on the converter’s output quantities. Second, as it is based on spectral analysis, it is suitable for open and closed-loop operation. Third, we demonstrate the identification system’s high accuracy, achieving a very low difference between the experimental measurements and the discrete model used as reference.