ACVT Harmonic Recovery Method Based on Deconvolution and Local Least 2-Norm Regression

A harmonic voltage recovery method for Capacitive Voltage Transformer (CVT) based on deconvolution and local least 2-norm regression is proposed. The proposed method takes advantage of the periodic consistency of CVT’s primary and second side voltage. It uses deconvolution and local least square norm regression methods, respectively, under periodic and aperiodic conditions to calculate the first side voltage signals of CVT, which can avoid the influence of aperiodic sampling on the deconvolution method and reduce the cumulative error. Simulation results show that the proposed method can accurately recover steady harmonic voltage, harmonic amplification voltage, transient overvoltage, and other aperiodic voltages.