Novel GIS Mechanical Defect Simulation and Detection Method Based on Large Current Excitation With Variable Frequency

Abstract

The mechanical defects of gas-insulated switchgear (GIS) equipments seriously threaten their safe and stable operation, but there are detection problems such as latency and weak vibration response, especially for high-capacity GIS with large size. In this paper, a novel GIS mechanical defect simulation and detection method based on large current excitation with variable frequency is proposed to solve this problem. First, a 500 kV GIS mechanical defect simulation and detection system with variable frequency and large current loading capacity was designed by structural parameter adjustment, power matching and harmonic suppression based on the nonlinear vibration mechanism. Then, the vibration characteristics of normal and typical defects of GIS equipment with different excitation frequencies were experimentally studied. Finally, the evolution laws of vibration signals and features for mechanical defects with different severity levels at different currents were explored, and a defect identification model for different load intervals was established. Results show that the amplitude growth and superharmonic point frequencies of vibration signals are related to the defect types and excitation current frequencies. The nonlinear vibration signal features such as kurtosis and gravity center frequency of defective equipment are more significant when using the proposed detection method compared with single frequency excitation, especially at large currents above 3,000 A, which not only makes it easier to find internal defects at an early stage but also helps to overcome the missing detection problem of the traditional method, and the defect recognition accuracy based on vibration signals of variable frequency excitation is higher.