Identification of Sub/Super-synchronous Oscillation Parameters of Wind Power System Based on INMF Combined Blind Signal Separation

With the development of power electronics technology, multi-machine coupling sub/super-synchronous oscillation accidents in wind power systems occur from time to time. Most of the common methods are aimed at the detection and identification of a single oscillation mode and are susceptible to noise interference. Mode aliasing will occur in oscillation identification when coupling is considered. In order to solve these problems, this paper proposes a sub/super-synchronous oscillation detection and parameter identification method based on improved non-negative matrix factorization (NMF) combined with blind signal separation, using synchronous squeeze transformation to improve the decomposition results of LOFAR spectra and enhance the aggregation degree within a short-time data window. Then, the blind signal separation algorithm is used to classify and sort the basis vectors obtained by decomposition, and the target signal is reconstructed. The parameters of the reconstructed signal are identified using the least squares rotation invariant technique (TLS-ESPRIT). Finally, the effectiveness of the method proposed in this paper is verified through the composite signal testing and PSCAD/EMTDC electromagnetic simulation. The results show that the method proposed in this paper does not require additional noise reduction means, can more accurately extract coupled sub/super-synchronous oscillation(SSSO) signals, and has higher reliability in parameter identification results. Meanwhile, it provides help and reference for broadband oscillation detection and traceability of renewable energy systems.