Fast compensation of current transformer saturation

This paper presents a novel method for current transformer (CT) saturation detection and compensation. In the method, CT saturation characteristics are represented by a partial nonlinear regression model. The parameters of this model can be estimated rapidly and accurately using both unsaturated and saturated sections of the waveform of a secondary fault current, even in a heavy saturation scenario, where the fault waveform is severely distorted. To reduce the computational load of the estimation, separable nonlinear least squares (SNLLS) method is used to separates the linear and nonlinear parameters of the model, thus reducing the complexity of the estimation. With the fast and accurate parameter estimation, the method can be used for real-time protective relaying. The method has been evaluated on the data obtained from PSCAD/EMTDC simulation. The analysis of the test results shows that the proposed method has superior performance than that offered by conventional compensation algorithms.