Fault Diagnosis Technique for Hydroelectric Generators using Variational Mode Decomposition and Power Line Communications

Abstract

With the increasing proportion of Hydro generator units in the whole power grid and its aging infrastructure, it becomes increasingly important hot topic for our next generation smart grid to complement a real-time fault diagnosis in Hydroelectric Generators. In this paper, we propose a road-map towards this issue. To this end, Variational Mode Decomposition method and other signal processing algorithms, such as Multi-dimension Permutation Entropy and fuzzy C-means clustering algorithm and so on, are developed and power line communication technique is used. Computer simulations results show that our proposed diagnostics solutions successfully empower Hydroelectric Generators to independently predict and identify the three different most commonly faults.