Method for diagnosis of gas insulated switchgear based on VMD parameters optimized by Chaotic Cuckoo algorithm

With the further improvement of the national electricity consumption, the fault accidents of the power grid are frequent. As an important equipment, GIS may cause a large scale of power outages, resulting in a heavy loss of power grid. This paper is devoted to studying two common mechanical failures of GIS, screw loosening, and metal particle vibration. The key part of research is to extract feature information corresponding to faults from vibration signals. Taking part of the GIS in the laboratory as the object of study, two common GIS failures were constructed: screw loosening and metal particle vibration. A GIS mechanical fault diagnosis method based on the Chaotic Cuckoo algorithm to optimize Variational Mode Decomposition(VMD) parameters is proposed. First, the Chaos Cuckoo algorithm was used to optimize the parameters of VMD. And the normal vibration signal and the fault signal are decomposed by VMD. Feature vectors of different types of faults signal can be extracted. Finally, the VMD entropy, the RMS, and the minimum Euclidean distance principle are used to identify the feature vectors of the sample. Then the diagnosis of the GIS mechanical fault is realized.