An Advanced Signal Decomposition Technique for Islanding Detection in DG System

This article presents a new passive islanding detection method by using an advanced signal decomposition technique, i.e., time-varying filter-based empirical mode decomposition (TVF-EMD). In the TVF-EMD, the adaptively tuned and predefined parameters and variation of the cutoff filter frequencies with respect to time make it more desirable over other decomposition techniques for enhanced resolution-based time-frequency analysis of nonstationary signals. The voltage signal measured at the distributed generation is processed through the TVF-EMD to decompose the signal into various intrinsic mode functions (IMF). The energy of the IMF is further extracted through the Teager energy operator. The computed energy is then used to identify the islanding scenario. In order to assess the performance of the proposed TVF-EMD, a number of simulation studies are performed on two standard test systems under varying islanding (i.e., different power mismatches and load quality factors) and critical nonislanding (i.e., capacitor switching and nonlinear load switching) conditions. The obtained results prove the efficacy of the proposed method.