GIS vibration defect feature extraction method based on fourier analysis band coding

Abstract

The application provides a GIS vibration defect feature extraction method based on Fourier analysis band coding, and belongs to the field of vibration feature extraction, and solves the problem that the prior art pays less attention to typical mechanical vibration defects of GIS and lacks analysis methods; the method comprises the following steps: constructing a vibration detection device and a mechanical defect simulation platform, obtaining a time domain vibration waveform through simulation, and obtaining a frequency domain vibration waveform through Fourier transform; adopting a band energy feature extraction method, obtaining a power spectrum, performing band coding operation, respectively calculating the energy and energy proportion of multiple bands, and obtaining a vibration feature information table; repeatedly simulating, reclassifying, establishing a vibration feature information library, detecting actual equipment, comparing with mechanical defect features in the vibration feature information library, and obtaining actual possible defect types; the application can detect and identify GIS mechanical defects, discover slight mechanical defects in time, guide maintenance, and provide protection for safe operation of GIS equipment.

Description

Technical Field

[0001] This invention belongs to the field of vibration feature extraction technology and is applied to GIS equipment in power systems. Specifically, it is a GIS vibration defect feature extraction method based on Fourier analysis frequency band coding. Background Technology

[0002] GIS equipment, short for gas-insulated fully enclosed combined electrical equipment, has unique advantages such as small footprint, large transmission capacity, high reliability and good environmental compatibility, and is therefore widely used. Moreover, the stable operation of GIS equipment is directly related to the safety of the power system.

[0003] Traditional studies on the reliability of GIS equipment often focus excessively on insulation faults. However, in recent years, GIS equipment in operation has frequently experienced failures caused by latent mechanical defects, especially 110kV and above. These mechanical defects include abnormal contact in conductive circuits, unbalanced casing connections, fatigue loosening of components, or loose bolts. GIS equipment with these mechanical defects will generate abnormal mechanical vibrations under the influence of alternating electrodynamic forces from load currents and mechanical forces from switch operations. Abnormal mechanical vibrations in GIS equipment are extremely harmful, potentially causing damage to basin insulators or insulating supports, gas leaks, gas pressure drops, floating grounding points, and increased circuit temperature rise. In severe cases, they can even lead to insulation accidents. Therefore, strengthening the detection of mechanical vibration faults in GIS equipment is a crucial means to ensure its safe operation.

[0004] In existing research on mechanical defects in GIS equipment, the focus is often on transient abnormal vibrations during the opening and closing of circuit breaker operating mechanisms. However, less attention is paid to latent mechanical defects such as loosening, fatigue, and wear in the main circuit and internal components of GIS equipment. In recent years, some scholars have conducted vibration characteristic studies on the periodic abnormal vibrations of one or two simple defects within GIS equipment, achieving a preliminary understanding of the vibration patterns of mechanical defects in GIS equipment. However, a systematic study of the vibration patterns of mechanical defects in multiple components within GIS equipment is still lacking, and further analysis of the high-frequency, weak-response detailed vibration characteristics of GIS equipment is even more scarce. Summary of the Invention

[0005] To fully or at least partially solve the problems mentioned in the background art, this invention proposes a GIS vibration defect feature extraction method based on Fourier analysis frequency band coding. This method can effectively construct a precise quantitative vibration feature information database of typical defects, which can be used for mechanical defect detection and type identification of GIS equipment on site. It can also promptly detect early faults caused by minor mechanical defects, thereby guiding maintenance work and providing important protection for the safe operation of GIS equipment.

[0006] The present invention employs the following technical solutions to achieve its objective:

[0007] A method for extracting vibration defect features in GIS based on Fourier analysis frequency band coding includes:

[0008] Analyze the vibration theory of GIS equipment and construct a vibration detection device;

[0009] A GIS equipment mechanical defect simulation platform is constructed. Using the GIS equipment mechanical defect simulation platform, typical mechanical vibration defects of GIS equipment are simulated to obtain the time-domain vibration waveform of the corresponding vibration signal.

[0010] The time-domain vibration waveform is transformed using Fourier analysis to obtain the frequency-domain vibration waveform;

[0011] The frequency domain vibration waveform is subjected to a frequency band energy feature extraction method to obtain the power spectrum of the corresponding vibration signal. Frequency band encoding is performed to divide the power spectrum into multiple frequency bands, and the energy of each frequency band and its proportion relative to the total energy are calculated to obtain the vibration feature information table of the corresponding vibration signal.

[0012] The simulation process of typical mechanical vibration defects is repeated to obtain a certain number of vibration characteristic information tables of various types of vibration signals. According to the classification of typical mechanical vibration defects, a vibration characteristic information database is established.

[0013] Using the vibration detection device, the actual GIS equipment is tested to obtain a vibration characteristic information table of the actual GIS equipment. This table is then compared with the typical mechanical vibration defect characteristic information in the vibration characteristic information database to determine the types of mechanical vibration defects that may exist in the actual GIS equipment.

[0014] Specifically, the vibration detection device is constructed by sequentially connecting a vibration sensor, a current adapter, a data acquisition card, and a PC-based host computer. This device has the advantages of being integrated and portable, and can easily obtain the time-domain vibration waveform during the actual operation of GIS equipment.

[0015] Furthermore, typical mechanical vibration defects of GIS equipment include poor contact of disconnecting switches, loose long conductor contacts, loose shielding covers, loose molecular sieve containers, and bent guide rods. These five mechanical vibration defects are typical in the operation of GIS equipment. The most important function of the GIS equipment mechanical defect simulation platform is to simulate the time-domain vibration waveforms of these five typical defects.

[0016] Specifically, the GIS equipment mechanical defect simulation platform includes a main circuit, a vibration and noise defect simulation module, a resonant voltage boosting and induction current boosting module, a vibration characteristic detection module, a measurement and operation control module, and a grounding and overcurrent protection module.

[0017] Specifically, the GIS equipment mechanical defect simulation platform applies a voltage range of 0 to 220kV and an applied current range of 0 to 3150A.

[0018] Furthermore, the time-domain vibration waveform is subjected to spectral analysis using the Fast Fourier Transform (FFT) method to obtain the frequency-domain vibration waveform. Frequency-domain analysis provides accurate analytical basis because it is based on spectral analysis, which decomposes a complex signal into a superposition of simple signals in the frequency domain. These simple signals correspond to various frequency components and simultaneously reflect the relationship between amplitude, phase, power, energy, and frequency. Currently, the most widely used method in spectral analysis is the Fast Fourier Transform (FFT) method.

[0019] Furthermore, a frequency band energy feature extraction method is used on the frequency domain vibration waveform. Based on the Wiener-Khinchin theorem, the frequency domain vibration waveform signal is regarded as stationary and random. Therefore, the power spectral density is the Fourier transform of the autocorrelation function of the frequency domain vibration waveform signal. It is found that the total power under the power spectral density is equal to the corresponding total average signal power of the frequency domain vibration waveform, and thus the power spectrum of the corresponding vibration signal is obtained.

[0020] Furthermore, since the vibration signal of GIS equipment is an electromechanical vibration signal with a wide response frequency band, typically concentrated between 0.2 and 3000 Hz, after obtaining the power spectrum of the corresponding vibration signal, the power spectrum is divided into 6 frequency bands within the main frequency range of the GIS equipment vibration signal, with each band consisting of 500 Hz. The main frequency range is from 0 to 3000 Hz. The 6 frequency bands are encoded so that each frequency band code corresponds to a different frequency band energy and frequency band energy ratio, serving as a standard for establishing a vibration characteristic information database.

[0021] After obtaining the frequency band energy of the vibration signal by the frequency band energy feature extraction method, it is encoded, and then a vibration feature information database under various mechanical vibration defect conditions is obtained through a large amount of simulation test data.

[0022] Specifically, the vibration characteristic information table obtained from the actual GIS equipment is compared with the frequency band energy and frequency band energy ratio information in the vibration characteristic information database. A comprehensive judgment is then made to determine the possible types of mechanical vibration defects in the actual GIS equipment.

[0023] In summary, due to the adoption of this technical solution, the beneficial effects of this invention are as follows:

[0024] This invention focuses on the loosening, fatigue, and wear-related mechanical defects of the main circuit and internal components of GIS equipment. It can effectively represent the vibration patterns of these latent defects and compare and apply them in actual maintenance. Based on Fourier transform theory and combined with frequency band coding, it effectively constructs a precise quantitative vibration characteristic information database of typical defects, which can promptly detect early faults caused by minor mechanical defects, thereby guiding maintenance work and providing important protection for the safe operation of GIS equipment in power systems. Attached Figure Description

[0025] Figure 1 This is a schematic flowchart of the method of the present invention;

[0026] Figure 2 This is a schematic diagram of the force analysis on the contact surface of the contactor;

[0027] Figure 3 This is a schematic diagram of the vibration detection device for GIS equipment.

[0028] Figure 4 The waveform test results are shown for GIS equipment under normal operating conditions.

[0029] Figure 5 The waveform test results of GIS equipment with loose screws are shown in the figure.

[0030] Figure 6 This is a schematic diagram illustrating the classification of typical mechanical vibration defects.

[0031] Figure 7 The waveforms are typical frequency domain vibration waveforms based on Fourier transform.

[0032] Figure 8 This is a schematic diagram of the software display interface for various results in the mechanical defect vibration characteristic information database. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0034] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0035] Before providing a detailed description of the methods involved in this embodiment, it is necessary to analyze the vibration theory of GIS equipment to make this embodiment clearer.

[0036] Depending on the configuration of the main circuit, GIS equipment can be divided into two types: separate phase type and three-phase common type.

[0037] For phase-separated GIS equipment, when an alternating current flows through the central conductor, a circulating current will be formed in the metal casing of the GIS equipment. It will be subjected to electromagnetic force in the alternating magnetic field generated by the current in the central conductor, which will cause it to expand and contract, resulting in vibration.

[0038] For three-phase integrated GIS equipment, under ideal conditions, the three-phase currents are balanced and the vector sum of the induced currents in the casing is zero. In actual operation, an unbalanced induced current will always be generated in the casing, and its value is usually very small. The three-phase conductor currents will generate an alternating electromagnetic field in the casing, producing regular vibrations.

[0039] The casing of GIS equipment is mostly made of aluminum alloy, which has the characteristics of low magnetic permeability and small eddy currents. Therefore, the electromagnetic force is relatively small and no obvious vibration will be generated under normal circumstances. Even if a weak vibration signal is generated, its vibration frequency is mainly 100Hz.

[0040] GIS equipment contains components such as busbars, circuit breakers, and disconnectors, all connected by guide rods and contacts. Due to frequent operation, contact spots form on the surface of the contacts. When current flows through them, the current lines contract near the contact surface and, under the influence of their own magnetic field, generate an electrodynamic force, causing vibration. The current lines on the contact surface and the resulting electrodynamic force... Figure 2 As shown.

[0041] The electrodynamic force between the contacts is:

[0042]

[0043]

[0044] In the above formula, F is the electrodynamic force between the contacts; μ0 is the free magnetic permeability; I0cos(ωt) is the current flowing through the contacts; a is the radius of the contact spot; D is the diameter of the contact surface; ζ is the deformation coefficient of the material; F j H represents the initial pressure on the contact; b The Brinell hardness is the material's stiffness.

[0045] In actual operation, alternating current flows through the conductor. When the current is not zero, the electrodynamic force of the contacts is repulsive. When the current through the contacts is zero, the electrodynamic force of the contacts is zero. Therefore, when alternating current flows through the conductor, periodically changing electrodynamic force is generated on the contacts, causing the contacts to vibrate. The vibration frequency is 100Hz.

[0046] The vibration signal of the GIS equipment casing contains a wealth of information about the GIS's operating status. Even minor changes in the internal structure of the GIS equipment can be reflected in the vibration signal. Vibration is the response of various electrical components inside the GIS equipment to multi-source excitation. Vibration also exists under normal operating conditions of the GIS equipment. However, mechanical defects such as loose fasteners, loose shielding covers, and poor contact of disconnecting switches can lead to changes in the vibration frequency distribution and amplitude, resulting in significant changes in the vibration signal. Therefore, analyzing the vibration signal characteristics of the GIS equipment casing is the theoretical basis for electrical equipment fault diagnosis.

[0047] The following details the method of this embodiment, which enables the extraction of vibration defect features of GIS equipment and their practical detection and application to ensure the safe operation of GIS equipment in the power system.

[0048] A method for extracting vibration defect features in GIS based on Fourier analysis frequency band coding includes:

[0049] Analyze the vibration theory of GIS equipment and construct a vibration detection device;

[0050] A GIS equipment mechanical defect simulation platform is constructed. Using the GIS equipment mechanical defect simulation platform, typical mechanical vibration defects of GIS equipment are simulated to obtain the time-domain vibration waveform of the corresponding vibration signal.

[0051] The frequency domain vibration waveform is obtained by transforming the time-domain vibration waveform using Fourier analysis.

[0052] The frequency domain vibration waveform is subjected to a frequency band energy feature extraction method to obtain the power spectrum of the corresponding vibration signal. Frequency band coding is then performed to divide the power spectrum into multiple frequency bands. The energy of each frequency band and its proportion relative to the total energy are calculated to obtain the vibration feature information table of the corresponding vibration signal.

[0053] The simulation process of typical mechanical vibration defects is repeated to obtain a certain number of vibration characteristic information tables of various types of vibration signals. According to the classification of typical mechanical vibration defects, a vibration characteristic information database is established.

[0054] A vibration detection device is used to detect the actual GIS equipment, and a vibration characteristic information table of the actual GIS equipment is obtained. This table is then compared with the typical mechanical vibration defect characteristic information in the vibration characteristic information database to determine the types of mechanical vibration defects that may exist in the actual GIS equipment.

[0055] In this embodiment, the vibration detection device is constructed by sequentially connecting a vibration sensor, a current adapter, a data acquisition card, and a PC-based host computer; please refer to the schematic diagram of this device. Figure 3The current adapter includes a current adapter circuit board and a current adapter power board, and is connected to a power adapter and a gain switch. In actual use, the number of components can be configured as follows: 4 vibration sensors, 4 current adapters, and 1 data acquisition card for integration. This device has the advantages of integration and portability, and can easily obtain the time-domain vibration waveform during the actual operation of GIS equipment.

[0056] Next, vibration waveform tests were conducted, with the test subjects being the normal state and those with loose screws. Figure 4 This is the waveform under normal conditions. Figure 5 The waveforms are shown in the image, indicating a loose screw defect. Test results show that vibration signals outside the GIS equipment housing can be detected by the vibration sensor. During a loose screw fault, the vibration amplitude is significantly larger, the waveform distortion is severe, and the characteristics are clearly different. Therefore, it is believed that defect identification can be achieved by comparing the characteristics of different vibration defect waveforms.

[0057] In this embodiment, a high-voltage, high-current GIS equipment mechanical defect simulation platform was established based on a 220kV-level GIS equipment real-model platform. The platform includes a main circuit, a vibration and abnormal noise defect simulation module, a resonant voltage boosting and induction current boosting module, a vibration characteristic detection module, a measurement and operation control module, and a grounding and overcurrent protection module. The voltage range that can be applied is 0 to 220kV, and the current range is 0 to 3150A.

[0058] The vibration and abnormal noise defect simulation module in the GIS equipment mechanical defect simulation platform can simulate typical mechanical vibration defects of GIS equipment, such as... Figure 6 As shown, the five mechanical vibration defects are typical of GIS equipment operation, including poor contact of disconnecting switches, loose long conductor contacts, loose shielding, loose molecular sieve containers, and bent guide rods. The most important function of the GIS equipment mechanical defect simulation platform is to simulate the time-domain vibration waveforms of these five typical defects.

[0059] The following section details the methods related to Fourier analysis in this embodiment.

[0060] For mechanical vibration defect identification, the amount of information that time domain analysis can provide is very limited; time domain analysis can often only roughly answer whether GIS equipment has defects, and sometimes it can also provide information on the severity of defects, but it cannot provide information such as the location of the defects.

[0061] As the most important and commonly used signal processing analysis method in mechanical defect diagnosis, frequency domain analysis can evaluate the status of GIS equipment by understanding the dynamic characteristics of the test object, accurately and effectively diagnose equipment defects and locate defects, and thus provide an analytical basis for preventing failures.

[0062] Frequency domain analysis is based on spectrum analysis, which decomposes a complex signal into a superposition of simple signals in the frequency domain. These simple signals correspond to various frequency components and simultaneously reflect the relationship between amplitude, phase, power, energy and frequency. Currently, the most widely used method in spectrum analysis is the Fast Fourier Transform (FFT).

[0063] For the time-domain vibration waveform x(t), the Fourier transform X(f) is as follows:

[0064]

[0065] In the formula, X(f) is the amplitude spectrum, f is the frequency, and t is the time.

[0066] Typical frequency domain vibration waveforms of Fourier transform are as follows: Figure 7 As shown.

[0067] Based on the Fourier analysis method, this embodiment then introduces a method for extracting frequency band energy features.

[0068] The power spectral density (PSD) function defines how the power of a signal or time series changes with frequency; this instantaneous power can be expressed as:

[0069] P = x(t) 2

[0070] Since signals with non-zero average values ​​cannot be integrally squared, a Fourier transform cannot be performed in this case. The Wiener-Khinchin theorem provides a simplified alternative: if the signal is considered stationary and random, the power spectral density is simply the Fourier transform of the signal's autocorrelation function; the power spectral function is:

[0071]

[0072] ω=2πf

[0073] The power spectral density of a signal exists if and only if the signal is a generalized stationary process. If the signal is not a stationary process, then the autocorrelation function must be a function of two variables, and thus the power spectral density does not exist. However, time-varying spectral densities can be estimated using similar techniques.

[0074] The spectral density and autocorrelation of the time-domain vibration waveform x(t) form a Fourier transform pair (different autocorrelation functions are used for power spectral density and energy spectral density).

[0075] The spectral density is usually estimated using Fourier transform techniques, but techniques such as Welch's method and maximum entropy can also be used.

[0076] One of the results of Fourier analysis is Parseval's theorem, which states that the sum (or integral) of the squares of a function, that is, its energy, is equal to the sum (or integral) of the squares of its Fourier transforms, as shown in the following equation:

[0077]

[0078] In the formula, X(f)=FT{x(t)} is the continuous Fourier transform of x(t), and f is the frequency component of x.

[0079] The above theorem also holds true in the discrete case (DTFT and DFT). Another conclusion is that the total power at the power spectral density is equal to the corresponding total average signal power, which is an autocorrelation function that gradually approaches zero.

[0080] After obtaining the power spectrum of the vibration signal using the above principles and methods, in this embodiment, since the vibration signal of the GIS equipment is an electromechanical vibration signal with a wide response bandwidth, typically concentrated between 0.2 and 3000 Hz, the main frequency range of the vibration signal (0 to 3000 Hz) is divided into six frequency bands, each lasting 500 Hz: 0-500 Hz, 500-1000 Hz, 1000-1500 Hz, 1500-2000 Hz, 2000-2500 Hz, and 2500-3000 Hz. The energy and energy percentage of each of the six frequency bands are then calculated.

[0081] After obtaining the frequency band energy of the vibration signal by the frequency band energy feature extraction method, it is encoded in the form of Table 1 below. Through a large amount of experimental data, a vibration feature information database under various mechanical vibration defect conditions was obtained.

[0082] Table 1 Six-band coding standard table

[0083] coding Bandwidth energy Bandwidth energy ratio 1 <![CDATA[0-1E -9 ]]> 0-1％ 2 <![CDATA[1E -9 -1E -8 ]]> 1％-10％ 3 <![CDATA[1E -8 -1E -7 ]]> 10％-30％ 4 <![CDATA[1E -7 -1E -6 ]]> 30％-60％ 5 <![CDATA[1E -6 -1E -5 ]]> 60％-90％ 6 <![CDATA[≥1E -5 ]]> 90％-100％

[0084] Based on the frequency band division method and the coding rules in Table 1 above, after averaging multiple measurements of simulated typical defects, a database of abnormal noise vibration characteristics based on the six-band energy coding method can be obtained, along with waveform coding results under correct conditions. Please refer to [the table]. Figure 8 The illustration.

[0085] By obtaining a database of vibration characteristic information under various mechanical vibration defect conditions, defect types can be identified and distinguished; such as Figure 8 As shown, the different defect characteristics are as follows:

[0086] (1) Normal state. Under normal circumstances, the frequency band with the highest energy is the first frequency band, 0-500Hz, and the energy ratio of the frequency band is basically above 90%, while the energy ratio of other frequency bands is less than 10%, or even less than 1%.

[0087] (2) Defect of poor contact of disconnecting switch. When the disconnecting switch has poor contact, the energy of the first frequency band is significantly reduced, while the energy of other frequency bands is significantly increased, and the proportion of frequency band energy is also greatly increased. That is, when the disconnecting switch has poor contact, the vibration signal has a relatively strong high-frequency component.

[0088] (3) Defects of loose molecular sieve containers. The most obvious distinguishing feature of loose molecular sieve fasteners is that the energy ratio of the 2nd to 6th frequency bands is relatively low, less than 1%, which is due to the excessively strong amplitude of the 100Hz component in this type of defect.

[0089] (4) Shielding cover loosening defect. When the shielding cover is loose, there is a relatively strong high frequency component under a low load current. As the load current increases, the 100Hz frequency component increases, and the high frequency energy is relatively less, but there is still a relatively strong frequency component in the 2-3 frequency band.

[0090] (5) Loose long conductor contacts. The characteristics of loose conductor contacts are similar to those of loose shielding, but the fault information is stronger than that of loose shielding. This is because both are high-potential components that are loose and have similar characteristics. If such characteristics are present, it can be identified as a loose high-potential fastener, and then a careful comparison can be made to determine which type of defect is more similar.

[0091] (6) Guide rod bending defect. The bending of the guide rod is completely the same as the normal situation, and it can be considered that there is no obvious defect and it does not affect the safe and stable operation of the GIS equipment.

[0092] Due to the complex on-site environment, various defect and normal state standard samples in the vibration characteristic information database can be used as reference cards. The vibration signal characteristic information on site is obtained by testing with a vibration detection device. The reference cards with 4×6 frequency band energy and 4×6 frequency band energy ratio are compared respectively to make a comprehensive judgment and determine the possible defect type.

[0093] Compared to normal conditions, poor contact caused by incomplete closing of the disconnecting switch is the most serious fault. The database shows significant differences from normal conditions, indicating a clear defect, consistent with field operating experience. Compared to poor contact of the disconnecting switch, the remaining three types of defects are relatively less noticeable and less distinguishable from normal conditions. Furthermore, these three defects often do not manifest as discharge faults in field operation experience and do not affect the normal operation of GIS equipment in the short term; therefore, the defects are not obvious. However, in GIS equipment operation experience, many discharge faults are caused by minor early mechanical defects. Studying the characteristics of abnormal vibration signals can effectively identify early defects, providing an important guarantee for the safe operation of GIS equipment.

Claims

1. A method for extracting vibration defect features in GIS based on Fourier analysis frequency band coding, characterized in that, include: Analyze the vibration theory of GIS equipment and construct a vibration detection device; A GIS equipment mechanical defect simulation platform is constructed. Using the GIS equipment mechanical defect simulation platform, typical mechanical vibration defects of GIS equipment are simulated to obtain the time-domain vibration waveform of the corresponding vibration signal. The time-domain vibration waveform is transformed using Fourier analysis to obtain the frequency-domain vibration waveform; The frequency domain vibration waveform is subjected to a frequency band energy feature extraction method to obtain the power spectrum of the corresponding vibration signal. Frequency band encoding is performed to divide the power spectrum into multiple frequency bands, and the energy of each frequency band and its proportion relative to the total energy are calculated to obtain the vibration feature information table of the corresponding vibration signal. After obtaining the power spectrum of the corresponding vibration signal, within the main frequency range of the GIS equipment vibration signal, the power spectrum is divided into 6 frequency bands, each 500Hz. The main frequency range is from 0 to 3000Hz. The 6 frequency bands are encoded so that each band corresponds to a different band energy and band energy ratio, serving as the standard for establishing a vibration characteristic information database. The frequency band energy range corresponding to frequency band code 1 is 0-1E. -9 The frequency band energy ratio range is 0%-1%; The frequency band energy range corresponding to frequency band code 2 is 1E. -9 -1E -8 The frequency band energy ratio ranges from 1% to 10%. The frequency band energy range corresponding to frequency band coding 3 is 1E. -8 -1E -7 The frequency band energy ratio ranges from 10% to 30%. The frequency band energy range corresponding to frequency band coding 4 is 1E. -7 -1E -6 The frequency band energy ratio ranges from 30% to 60%. The frequency band energy range corresponding to frequency band code 5 is 1E. -6 -1E -5 The bandwidth energy ratio ranges from 60% to 90%. The frequency band energy corresponding to frequency band code 6 is greater than or equal to 1E. -5 The bandwidth energy ratio ranges from 90% to 100%. The simulation process of typical mechanical vibration defects is repeated to obtain a certain number of vibration characteristic information tables of various types of vibration signals. According to the classification of typical mechanical vibration defects, a vibration characteristic information database is established. Using the vibration detection device, the actual GIS equipment is tested to obtain a vibration characteristic information table of the actual GIS equipment. This table is then compared with the typical mechanical vibration defect characteristic information in the vibration characteristic information database to determine the possible types of mechanical vibration defects in the actual GIS equipment. The vibration characteristic information table obtained from the actual GIS equipment is compared with the frequency band energy and frequency band energy ratio information in the vibration characteristic information database. A comprehensive judgment is then made to determine the possible types of mechanical vibration defects in the actual GIS equipment.

2. The GIS vibration defect feature extraction method based on Fourier analysis frequency band coding according to claim 1, characterized in that: The vibration detection device is constructed by sequentially connecting a vibration sensor, a current adapter, a data acquisition card, and a PC-based host computer.

3. The GIS vibration defect feature extraction method based on Fourier analysis frequency band coding according to claim 1, characterized in that: Typical mechanical vibration defects in GIS equipment include poor contact of disconnecting switches, loose long conductor contacts, loose shielding, loose molecular sieve containers, and bent guide rods.

4. The GIS vibration defect feature extraction method based on Fourier analysis frequency band coding according to claim 1, characterized in that: The GIS equipment mechanical defect simulation platform includes a main circuit, a vibration and noise defect simulation module, a resonant voltage boosting and inductive current boosting module, a vibration characteristic detection module, a measurement and operation control module, and a grounding and overcurrent protection module.

5. The GIS vibration defect feature extraction method based on Fourier analysis frequency band coding according to claim 4, characterized in that: The mechanical defect simulation platform for GIS equipment applies a voltage range of 0 to 220kV and an applied current range of 0 to 3150A.

6. The GIS vibration defect feature extraction method based on Fourier analysis frequency band coding according to claim 1, characterized in that: The time-domain vibration waveform is subjected to spectral analysis, and the frequency-domain vibration waveform is obtained by fast Fourier transform.

7. The GIS vibration defect feature extraction method based on Fourier analysis frequency band coding according to claim 6, characterized in that: The frequency domain vibration waveform is subjected to a frequency band energy feature extraction method. Based on the Wiener-Khinchin theorem, the frequency domain vibration waveform signal is regarded as stationary and random. Therefore, the power spectral density is the Fourier transform of the autocorrelation function of the frequency domain vibration waveform signal. It is found that the total power under the power spectral density is equal to the corresponding total average signal power of the frequency domain vibration waveform, and thus the power spectrum of the corresponding vibration signal is obtained.

Images