Compensation of Distorted Secondary Current Caused by Saturation and Remanence in a Current Transformer

A large primary current and DC component in a current transformer (CT) will result in a distorted secondary current if the CT becomes saturated. It may lead to a malfunction in the protective relay because of a smaller rms current resulting from the distorted current during fault occurrence. Hence, detection and correction of the distorted current caused by the saturated CT is crucial. This paper employed wavelet transform to detect the time for fault occurrence and identify all saturation periods. Five features of the fault current were extracted. Fuzzy-c-means was utilized to partition all possible fault currents into 42 clusters according to their features through statistic tests. Forty-two nonlinear regression formulas were developed individually to obtain compensated secondary currents at the initial stage. Takagi-Sugeno-Kang fuzzy rules were then employed to integrate output currents obtained by all the nonlinear regression formulas into the final compensated secondary current. The proposed method was verified by Simulink/Visual C and was also implemented in Field Programmable Gate Array (FPGA, Altera Stratix EP1S25F780C5 development board) chip. The test results show the applicability of the proposed method.