An Estimation Method for Soft Fault Reflection Coefficient of Power Cable Based on Sliding-Window TLS-ESPRIT

Abstract

Reflection coefficient is important to evaluate the characteristics of power cable soft faults. Thus, a reflection coefficient estimation method based on sliding-window total least squares (TLS) estimation of signal parameters via rotational invariance techniques (ESPRIT) is proposed. The idea of this method is to transform random signals of cable transfer function (CTF) into array signals with deterministic auto-correlation function, then precisely locate and evaluate soft faults by using TLS-ESPRIT estimation of CTF parameters. First, the two intrinsic mode functions (IMF) signals, which are the reflected waves of the test terminal and the cable body, are separated from the CTF by variational mode decomposition (VMD). Then, the IMFs were intercepted into continuous subintervals, and the frequency and attenuation factor contained in each subinterval were estimated by ESPRIT, and the amplitude and phase were estimated by TLS. Next, the reflection coefficient is calculated by decoupling the amplitude and phase. Finally, in the simulation, the amplitude relative error and phase relative error of the soft fault reflection coefficient estimated by the proposed method are less than 6% and 5%, respectively. And the feasibility of the proposed method is verified by field cases.