Application of Multisynchrosqueezing Transform for Subsynchronous Oscillation Detection Using PMU Data
Yue Ma,Qi Huang,Zhenyuan Zhang,Dongsheng Cai
DOI: https://doi.org/10.1109/tia.2021.3057313
IF: 4.079
2021-05-01
IEEE Transactions on Industry Applications
Abstract:This study proposes a new technique using multisynchrosqueezing transform (MSST) for subsynchronous oscillation (SSO) detection and modal parameter identification. First, MSST is used to process the real-time data of voltage or current collected by the phasor measurement unit (PMU) to obtain the timefrequency spectrum representation of the signal. Nonzero MSST coefficients are used to represent the frequency components of the original signal at each time. Then, for the unknown number of subsynchronous frequency range components, the mean-shift algorithm is applied to the nonzero MSST coefficients to determine the number of SSO modes contained in the signal and the mode type to which nonzero MSST coefficients belong. Subsequently, the corresponding MSST coefficients of each mode exceeding the threshold are used to reconstruct the components of the original signal. Finally, the least-squares estimation method is used on the reconstructed signal to obtain the parameters of each mode, such as amplitude, phase angle, and damping factor. Through simulation case studies and comparison with existing methods, it is verified that the proposed method utilizes the abundant information contained in the PMU data and demonstrates superiority and effectiveness in multimodal signal timefrequency analysis. Moreover, practical issues and potential applications of this technique are systematically discussed.
engineering, electrical & electronic, multidisciplinary