Compensation of Current Transformer Saturation Based on Improved Esprit Algorithm

Jun Guan,Zhiguo Hao,Wenzhe Chen,Zhiyuan Liu,Xiaojun Yu
DOI: https://doi.org/10.1109/eeeic.2016.7555447
2016-01-01
Abstract:Current transformer is prone to get into transient saturated during the fault in the power system. Operation performance of relay protection will be influenced severely due to serious distortion in the secondary side. An improved ESPRIT algorithm that can be applied in relay protection for anti-saturation is presented in this paper. The improved ESPRIT can accurately obtain the frequency, magnitude and phase angle of the fault current with unsaturated data segment of CTs' distorted secondary current in a relatively short time window. In order to reduce the calculation amount, singular value decomposition (SVD) is employed to determine the valid order of the signal. Moreover, the improved ESPRIT algorithm could reduce the computational load with little influence on calculative result through the dimension reduction of singular value decomposition. In order to obtain the magnitude and phase angle more accurate, AR model and least square method are employed to improve the robustness of algorithm. Finally, the accuracy of the improved ESPRIT is verified by CT's transient experiment data. The improved SEPRIT algorithm could be applied in the difference protection to avoid the malfunction by CTs' transient saturation and ensure the validity of protection.
What problem does this paper attempt to address?