Detection for Abnormal Commutation Process State of Converter Based on Temporal and Amplitude Characteristics of AC Current

Commutation failure (CF) increases the current flowing through the converter and endangers the converter. To protect the converter, CF protection is used in high voltage direct current systems. The core of CF protection is accurate CF detection. Owing to the nonlinear characteristics of converter and the rapid control, the conventional differential current method used to detect CF faces challenges. This article proposes a method to detect CF directly by tracking commutation process. The study indicates that the characteristics of commutation process differ significantly between normal operation and CF. By extracting these characteristics, commutation process can be tracked, and CF can be detected. For improving reliability, the proposed method depends on the relative relationship between AC current and DC current, reducing the influence of current fluctuation. Temporal and amplitude characteristics are both extracted to judge commutation process state and detect CF. The performance of the proposed scheme is validated using an actual engineering model and recorded engineering data. The results indicate that the proposed method can effectively detect CF under different scenarios.