A New Variational Mode Decomposition-Based Passive Islanding Detection Strategy for Hybrid Distributed Renewable Generations

In this article, we develop a novel passive islanding detection strategy for a hybrid distributed generation (DG) system that uses variational mode decomposition (VMD). The voltage signal's ripple content is evaluated at the point of common coupling (PCC). It splits the input voltage signal into its individual modes or intrinsic mode functions (IMFs). The upper envelope of first mode's IMF has been used for island detection. This strategy relies solely on internal, mutable, and variational processes. Results from the tests show that the suggested approach can reliably separate the islanding event from other events. Even with no power imbalance, the suggested approaches can identify islanding in about 0.15 s. It has been tested in a wide variety of operational scenarios and proved in non-island situations. The suggested solutions are easy to implement, since they do not require a classifier, also had no non-detection zone (NDZ), and function regardless of the amount and kind DG that might connect to the utility grid.

Automated summary

In this article, a novel passive islanding detection strategy for a hybrid distributed generation system is developed using variational mode decomposition (VMD). The strategy evaluates the ripple content of the voltage signal at the point of common coupling (PCC) and decomposes the input voltage signal into its individual modes or intrinsic mode functions (IMFs). The upper envelope of the first mode's IMF is used for island detection. The suggested approach shows reliable separation of the islanding event from other events and can identify islanding in about 0.15 s even without power imbalance. It has been tested in various operational scenarios and proved effective in non-island situations, making it an easy-to-implement solution for different types of distributed generation connected to the utility grid.