Mechanical defect identification for gas‐insulated switchgear equipment based on time‐frequency vibration signal analysis

Abstract

Abstract Mechanical defect is an important reason for the failure of gas‐insulated switchgear (GIS) equipment. Based on the time‐frequency characteristic vibration signal analysis on five kinds of mechanical defects, a novel intelligent algorithm model combining complementary ensemble empirical mode decomposition (CEEMD) and genetic algorithm improved kernel fuzzy mean clustering (GAKFCM) was proposed to identify the mechanical defect type. First, the mechanical defect platform and detection system were built. Then CEEMD and IMF sensitivity factors were used to analyse the time‐frequency signal of five kinds of vibration defects, and the feature extraction was performed on the processed vibration signals. Finally, the mechanical vibration defect recognition model was established based on the GAKFCM algorithm and its validity was verified. Results show that the developed detection system can detect mechanical vibration signals sensitively. Singular values, frequency band lines and entropy can reflect the energy attenuation and distribution differences for different type of mechanical defect vibration signals. The proposed GAKFCM clustering model combining the above vibration feature parameters can effectively find and diagnose the mechanical defect of GIS equipment. Its recognition accuracy reaches 96.74%, especially for the loose contact seat bolts and poor contact failures of the disconnector.

Keywords

• fault diagnosis
• feature extraction
• fuzzy set theory
• gas insulated switchgear
• genetic algorithms
• mechanical engineering computing