Method and system for judging contact quality of contact finger of GIS (Gas Insulated Switchgear) disconnecting switch

Abstract

The invention discloses a method and system for judging the contact quality of a contact finger of a GIS disconnecting switch, and the method comprises the steps: carrying out the real-time collection of switching-on instantaneous vibration through employing optical fiber EFPI low-frequency / high-frequency dual-channel sensors which are distributed at the rigid coupling part of a contact finger mechanical link, and extracting the time domain, frequency domain and time frequency characteristics; and the contact resistance and the thermal risk index of the contact finger are estimated through a pre-trained multi-channel model, so that high-precision contact finger in-place and thermal safety early warning without electric quantity and contact is realized. In this way, the vibration response characteristic of the contact finger at the closing moment is sensed through the EFPI optical fiber sensor, and judgment of the contact quality of the contact finger under the non-visual condition is achieved; when the contact resistance of a certain contact finger is increased, early abnormity identification of a single contact finger, local contact degradation trend tracking and early warning before formation of potential hot spots can be realized through an EFPI vibration signal, and the operation safety and the maintenance efficiency of the GIS isolation switch are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of power technology, and in particular to a method and system for judging the contact quality of GIS disconnect switches. Background Technology

[0002] Gas-insulated metal-enclosed switchgear (GIS) is a key piece of equipment widely used in high-voltage power transmission and transformation systems. Its disconnector contact assemblies are completely sealed within a metal housing and SF6 insulation medium. In this equipment, the switch's closed-position state cannot usually be confirmed directly by visual inspection or temperature measurement. Current technologies generally rely on the following indirect methods for judgment: (1) Closing position display technology based on mechanical indicating mechanism: Existing GIS disconnect switches are generally equipped with mechanical closing indicating mechanism, which synchronizes the output displacement of the operating mechanism to the indicating window on the outer casing through transmission linkage, shift fork or gear set. The operator judges whether the disconnect switch has been operated in place by relying on the "closing / opening" position displayed in the indicating window.

[0003] However, it has the following shortcomings: Mechanical chain errors cannot reflect the actual contact state of the contact fingers. The transmission link consists of multiple mechanisms, which are subject to wear, loosening, and backlash accumulation, causing the indicated position to potentially differ from the actual contact position of the contact fingers. The indicated position only reflects the mechanism's action, not the electrical contact performance. Even if the mechanical indication indicates closing, the actual contact fingers may have insufficient contact pressure, eccentric contact, or some contact fingers may not be in contact. Changes in contact resistance cannot be detected. There is no one-to-one correspondence between the mechanical displacement signal and the electrical contact state. Therefore, the mechanical indication method cannot accurately determine the actual contact quality of the disconnector switch contact fingers.

[0004] (2) Closing status judgment technology based on electrical quantity parameters: Electrical equipment judges whether the contacts are closed by the changes in operating current and voltage waveforms. However, in GIS disconnect switches, since their working characteristics are "no-load operation", they do not carry load current after closing, and traditional electrical parameter monitoring methods cannot provide an effective basis for judgment.

[0005] The limitations of this technology include: the normal operating current after the disconnector is closed is close to zero, and even if the contacts are not fully in contact, the current measurement cannot provide differentiation. It is difficult to reverse-calculate the contact state when the current distribution of the contact group is uneven, and traditional methods cannot distinguish contact abnormalities in a single contact in a multi-contact structure. Therefore, traditional power monitoring methods cannot meet the requirements of GIS disconnectors for judging "contact quality" and "contact pressure".

[0006] (3) Contact status judgment technology based on temperature monitoring: Existing switchgear often uses fiber optic temperature measurement, infrared temperature measurement and other methods to monitor the temperature rise of contacts in order to identify abnormal heating caused by poor contact. Fiber optic cables cannot be laid inside the completely sealed SF6 metal cavity at the contact finger position; the metal shell completely shields infrared radiation, and infrared imaging cannot penetrate the shell; the contact surface of the contact finger is located inside the multi-contact structure, and the true temperature cannot be obtained by inverting the external temperature field.

[0007] The technical limitations of this technology are as follows: It exhibits a significant temperature hysteresis effect, meaning the temperature rise only occurs after heat generation, making it impossible to identify early risks caused by increased contact resistance. It cannot distinguish the temperature rise of individual contacts; temperature measurements primarily reflect the overall situation and cannot pinpoint specific abnormal contacts. The enclosed structure renders temperature measurement methods completely ineffective; even if localized overheating occurs, the surface temperature of the outer casing may not change significantly. Therefore, temperature monitoring methods are essentially impractical in GIS disconnector applications.

[0008] Therefore, it is necessary to propose a comprehensive technical solution for identifying the contact status of disconnector switch contacts and judging thermal risks using non-electrical parameters under non-visual conditions, so as to overcome the inherent limitations of existing technologies. Summary of the Invention

[0009] The technical problem to be solved by the present invention is to provide a method and system for judging the contact quality of GIS disconnect switches, which can realize early warning of potential thermal hazards at the contact level of GIS disconnect switches under non-visual conditions.

[0010] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for judging the contact quality of a GIS disconnector includes the following steps: S1. Select the measurement points that are rigidly coupled to the mechanical link of the GIS disconnect switch, and install fiber optic EFPI vibration sensors containing low-frequency and high-frequency channels at each measurement point. S2. When the GIS disconnect switch is closed, the vibration signals of all channels are collected in real time within a preset time window; S3. By extracting features from the vibration signals of each channel, the features of each channel are obtained; S4. Input the features of multiple channels into a pre-trained multi-channel model, and output the contact resistance and thermal risk index. Determine the contact quality of the GIS disconnector based on the contact resistance and the thermal risk index.

[0011] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows: A contact quality judgment system for a GIS disconnector switch includes: The sensing module is used to select the measuring points that are rigidly coupled to the mechanical link of the GIS disconnect switch, and to install fiber optic EFPI vibration sensors containing low-frequency and high-frequency channels at each measuring point. The data acquisition module is used to collect vibration signals of all channels in real time within a preset time window when the GIS disconnect switch is closed. The data processing module is used to extract features from the vibration signals of each channel to obtain the features of each channel; The diagnostic module is used to input the features of multiple channels into a pre-trained multi-channel model and output the contact resistance and thermal risk index. Based on the contact resistance and thermal risk index, the contact quality of the GIS disconnect switch is determined.

[0012] The beneficial effects of this invention are as follows: Utilizing fiber optic EFPI low-frequency / high-frequency dual-channel sensors distributed at the rigid coupling point of the contact finger's mechanical link, real-time acquisition of vibration during closing is performed, extracting time-domain, frequency-domain, and time-frequency characteristics. Then, a pre-trained multi-channel model is used to estimate the contact resistance and thermal risk index of the contact finger, achieving high-precision contact finger positioning and thermal safety early warning without electrical charge or contact. In this way, the vibration response characteristics of the contact finger at the moment of closing are perceived by the EFPI fiber optic sensor, enabling judgment of contact quality under non-visual conditions. When the contact resistance of a certain contact finger increases, the EFPI vibration signal can achieve: early identification of single contact finger anomalies, tracking of local contact degradation trends, and early warning before potential hotspot formation, significantly improving the operational safety and maintenance efficiency of GIS disconnect switches. Attached Figure Description

[0013] Figure 1 This is a flowchart of a method for judging the contact quality of a GIS disconnector according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a contact quality judgment system for a GIS disconnector according to an embodiment of the present invention. Detailed Implementation

[0014] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0015] Before detailing the embodiments of this application, some related concepts will first be explained: (1) GIS disconnector: A disconnector device in low-voltage air distribution (Gas-Insulated Switchgear, GIS). It uses high-voltage insulating gas (such as SF6) as the medium, has a closed design, small size, and low maintenance, and is widely used for isolation, branching and protection of high-voltage or medium-voltage power transmission systems.

[0016] (2) Contact finger: The metal contact inside the GIS switch used to contact the load or downstream cable, also known as a contact or contact group. When the switch is closed / opened, the contact finger will come into contact with adjacent contacts or conductors to form an electrical conduction path. Its mechanical and electrical condition (contact integrity, micro-arc, contact fatigue, etc.) directly determines the safety and reliability of the switch operation. The contact finger is usually made of high-temperature steel, metal elastomer or tempered insulating material, and is coupled to the contact finger steel tube through a support structure.

[0017] (3) Fiber Optic EFPI Vibration Sensor: This optical sensor utilizes the interference generated by light waves transmitted through an optical fiber in an external cavity (Fabry-Perot optical path), and converts the changes in the interference signal into displacement or tensile information of the measured medium (such as a diaphragm or piezoelectric material). Its core structure includes two beams of light (incident light and reflected light) that undergo multiple reflections within the cavity. The cavity length changes with the measured parameter, resulting in changes in the phase or intensity of the interference fringes. The signal is directly sent to an optical demodulator through an optical fiber interface, enabling high-precision vibration / displacement measurement with no electromagnetic interference, long-distance, and remote distributed reading.

[0018] (4) Thermal risk index: used to quantify the probability of thermal damage to the contact fingers of GIS disconnect switch at the moment of closing or during continuous operation.

[0019] In existing technologies, the determination of the closed-position status of GIS disconnectors mainly relies on mechanical indication, electrical quantity monitoring, and temperature rise monitoring. However, because the contact finger assembly of the GIS disconnector is completely enclosed within a metal casing and SF6 gas insulation cavity, traditional external measurement methods cannot obtain key status information such as contact quality, contact pressure, and arc development during the closing process. The main drawbacks of the monitoring methods used in existing technologies can be summarized as follows: (1) Existing technologies cannot identify the fine contact behavior during the closing process: During the closing process, the contact fingers of GIS disconnectors exhibit a series of micro-contact behaviors, including initial contact of the contacts, decrease in the transition resistance of the contact surface, establishment of contact pressure, change in insertion force, and contact lag of some contact fingers. Existing mechanical indication methods and electrical quantity methods cannot capture such millisecond-level and micro-displacement-level dynamic contact behaviors, thus failing to accurately identify whether "effective closing" has been truly achieved.

[0020] (2) Inability to detect the arc formation and arc extinguishing process online: A brief micro-arc or contact breakdown ionization phenomenon occurs at the moment of closing of the disconnecting switch. Although its duration is short, it has a significant impact on the surface condition of the contact fingers, the evolution of contact resistance, and subsequent thermal stability. In existing technologies: temperature monitoring methods cannot penetrate the GIS metal casing and cannot observe the arc thermal effect; mechanical indicating devices cannot reflect whether an arc has occurred; electrical quantity monitoring, due to its extremely low current, cannot detect the micro-arc behavior at the moment of closing. Therefore, current traditional technologies are completely unable to detect the arc development process.

[0021] (3) Unable to be used to calculate contact resistance or identify potential thermal risks: Contact resistance is a core parameter for judging contact quality and future overheating risks. However, since disconnecting switches operate under no-load conditions, existing technologies cannot invert changes in contact resistance through current or temperature signals. Temperature rise monitoring and current monitoring have long been unavailable for disconnecting switches, therefore: the early stages of contact resistance increase cannot be identified; early warnings cannot be given before local micro-hot spots form; and individual contact abnormalities cannot be reflected through overall parameters. This results in the disconnecting switch contact overheating hazard having the characteristics of "high concealment and external unobservable".

[0022] (4) Existing technologies lack the ability to identify each contact in a multi-contact structure within a contact group: GIS disconnect switches often use a multi-contact structure, with multiple contacts simultaneously undertaking the closing contact task. Existing technologies cannot achieve breakthroughs in the following aspects: they cannot determine whether the contact of each contact is synchronized; they cannot identify contact lag or insufficient insertion force of a certain contact; they cannot locate the arcing position within a single contact; and they cannot distinguish individual abnormal contacts among multiple contacts through external signals. Therefore, traditional methods cannot perform refined status assessment at the contact level.

[0023] In view of the above-mentioned deficiencies of the prior art, this invention proposes to use an optical fiber EFPI (Extrinsic Fabry-Perot Interferometer) high-sensitivity vibration sensor to collect micro-vibration signals in real time during the closing process of the disconnecting switch, extract key parameters such as vibration frequency, vibration amplitude, and arc excitation characteristics, and establish the correspondence between these parameters and the contact state of the contact finger, arc development, and contact overheating, thereby realizing non-electrical quantity diagnosis of the closing state of the GIS disconnecting switch.

[0024] The core technical problems that this invention aims to solve include: (1) Solve the problem of the invisibility of the actual contact state of the touch finger: The vibration response characteristics of the touch finger at the moment of closing are sensed by the EFPI fiber optic sensor, including: the high-frequency impact signal at the moment of contact; the vibration attenuation characteristics during the establishment of the contact pressure of the touch finger; and the frequency drift caused by the change of insertion force; thereby realizing the "judgment of the contact quality of the touch finger under non-visual conditions".

[0025] (2) Solve the problem of not being able to identify the generation and development process of electric arcs online: During the generation, diffusion and extinction of electric arcs, characteristic high-frequency vibration signals are generated, including: plasma breakdown high-frequency waves; continuous vibration caused by micro-arc jumps; wide-spectrum response of ionization process; EFPI high-sensitivity sensor can capture such signals and realize: micro-arc detection → arc duration calculation → arc energy inference → abnormal contact judgment.

[0026] (3) Solving the problem of not being able to provide early warning of potential thermal hazards at the finger level: When the contact resistance of a finger increases, its vibration energy, resonant mode, and impact response will all undergo characteristic changes. Through the EFPI vibration signal, it is possible to: identify early anomalies in a single finger; track the trend of local contact degradation; and provide early warning before the formation of potential hot spots. This is something that existing technologies cannot achieve at all.

[0027] Please refer to Figure 1 This invention provides a method for judging the contact quality of a GIS disconnector, including the following steps: S1. Select the measurement points that are rigidly coupled to the mechanical link of the GIS disconnect switch, and install fiber optic EFPI vibration sensors containing low-frequency and high-frequency channels at each measurement point. S2. When the GIS disconnect switch is closed, the vibration signals of all channels are collected in real time within a preset time window; S3. By extracting features from the vibration signals of each channel, the features of each channel are obtained; S4. Input the features of multiple channels into a pre-trained multi-channel model, and output the contact resistance and thermal risk index. Determine the contact quality of the GIS disconnector based on the contact resistance and the thermal risk index.

[0028] As described above, the beneficial effects of this invention are as follows: Utilizing fiber optic EFPI low-frequency / high-frequency dual-channel sensors distributed at the rigid coupling point of the contact finger's mechanical link, real-time acquisition of vibration during closing is performed, extracting time-domain, frequency-domain, and time-frequency characteristics. Then, a pre-trained multi-channel model is used to estimate the contact resistance and thermal risk index of the contact finger, achieving high-precision contact finger positioning and thermal safety early warning without electrical charge or contact. In this way, the vibration response characteristics of the contact finger at the moment of closing are perceived by the EFPI fiber optic sensor, enabling judgment of contact quality under non-visual conditions. When the contact resistance of a certain contact finger increases, the EFPI vibration signal can achieve: early identification of single-contact finger anomalies, tracking of local contact degradation trends, and early warning before potential hotspot formation, significantly improving the operational safety and maintenance efficiency of GIS disconnect switches.

[0029] Further, step S3 includes: Time-domain analysis was performed on the vibration signal of the low-frequency channel to extract the impact peak value and decay time. Frequency-domain analysis was performed on the vibration signal of the low-frequency channel to extract the main harmonic frequency. Time-frequency analysis is performed on the vibration signal of the high-frequency channel to calculate the high-frequency band energy within a preset high-frequency range.

[0030] As described above, by performing time-domain (impact peak value, decay time) and frequency-domain (dominant harmonic frequency) analysis on the low-frequency channel, and time-frequency analysis (calculating high-frequency energy within a preset frequency band) on the high-frequency channel, qualitative and quantitative information on the finger impact dynamics and micro-arc / partial discharge can be obtained from a single vibration signal. Low-frequency characteristics characterize the stiffness, contact pressure, and impact duration of the contact structure; high-frequency energy can promptly capture high-frequency vibrations generated by partial discharge or arcing. Synthesizing these two types of information into a multi-channel feature vector can significantly improve the accuracy of finger positioning determination, reduce false alarms, and provide thermal risk warnings upon the first occurrence of a micro-arc, thereby enhancing the safety and operational efficiency of the GIS disconnect switch.

[0031] In some embodiments, time-domain analysis is performed on the vibration signal of the low-frequency channel using a sampling rate of 100 kS / s, a Hann window, and an 8192-point FFT to extract the impact peak value and decay time. Frequency-domain analysis is performed on the vibration signal of the low-frequency channel using the same sampling rate, window, and FFT point count to extract the dominant harmonic frequency. Time-frequency analysis is performed on the vibration signal of the high-frequency channel using a sampling rate of 200 kS / s, a Hann window, and an 8192-point FFT to calculate the high-frequency band energy in the range of 1 kHz–20 kHz. The total energy is calculated and the energy ratio is obtained. If the energy ratio is greater than the ratio threshold (obtained through offline calibration), it is determined that the channel contains significant high-frequency energy.

[0032] Furthermore, after step S3, the method further includes determining the type of finger contact quality based on the characteristics of each channel, specifically: If the peak value of the impact is greater than the minimum value of the preset impact range, and the main harmonic frequency and the decay time are both within the corresponding normal range, then it is determined that the contact finger of the GIS disconnect switch is in mechanical position and the contact quality is normal. If the peak impact value is less than the preset impact threshold, or the rate of decrease of the decay time reaches the preset rate, it is determined that the contact pressure of the GIS disconnect switch is insufficient / the contact finger is too loose. If the impact peak is located within the preset impact range and the high-frequency band energy exceeds the energy threshold, it is determined that the contact finger of the GIS disconnect switch is mechanically in place, but there is poor contact / partial discharge or initial arc. If the characteristic differences of multiple channels are greater than the preset difference threshold, it is determined that the multiple contacts of the GIS disconnect switch are out of sync or the single contact is abnormal.

[0033] As described above, this judgment logic combines low-frequency impact peak, main harmonic frequency, decay time and high-frequency energy to form a multi-dimensional quality label for the contact finger closing state. In this way, it is possible to achieve panoramic online monitoring of contact finger positioning, contact pressure, arc risk and multi-contact finger synchronization status at the lowest additional cost in the absence of electricity and contact, which significantly improves the operational safety, fault early warning accuracy and maintenance decision efficiency of GIS disconnect switches.

[0034] Further, in step S4, the features from multiple channels are input into a pre-trained multi-channel model, and the output is the contact resistance, including: The multi-channel model includes a first fitting model and a second fitting model; the first fitting model is established based on the vibration characteristics under different historical contact pressures and contact areas, and the second fitting model is established based on the contact resistance under different historical contact pressures or contact areas. The impact peak value, the decay time, the main harmonic frequency, and the high-frequency band energy are input into the first fitting model, and the contact pressure and contact area are output by the first fitting model. The contact pressure and contact area are input into the second fitting model, and the contact resistance is obtained from the output of the second fitting model.

[0035] As described above, this step achieves a two-step mapping from vibration characteristics to contact pressure / contact area, and then to contact resistance. The first fitting model maps time-domain / frequency-domain characteristics such as impact peak value, decay time, main harmonic frequency, and high-frequency energy to the contact pressure and contact area of ​​the finger. The second fitting model predicts resistance based on known pressure / area data, fully reflecting the decisive influence of mechanical coupling on resistance. Through hierarchical regression or machine learning, this method can obtain a high-precision resistance estimate without directly measuring resistance, thereby further calculating the thermal risk index. Compared with traditional simple resistance measurement, vibration-based indirect estimation has advantages such as resistance to electromagnetic interference, non-contact operation, and online self-learning capability. It can also dynamically monitor the contact status at the moment of closing and during long-term operation, significantly improving the safety and operation and maintenance efficiency of GIS disconnect switches.

[0036] Furthermore, the output thermal risk index includes: The thermal risk index PI is calculated based on the contact resistance Rc and the high-frequency band energy ArcEnergy. PI=w1·Rc+w2·ArcEnergy+w3·ΔRc / Δt In the formula, w1 represents the weight of contact resistance, w2 represents the weight of high-frequency band energy, ΔRc / Δt represents the rate of change of contact resistance, and w3 represents the weight of the rate of change of contact resistance.

[0037] As described above, by implementing multi-dimensional thermal risk assessment using contact resistance, partial discharge energy, and the rate of change of contact resistance within the same index, the safety and reliability of GIS disconnect switches are significantly improved. Contact resistance directly reflects the contact area and friction state of the contact fingers; high-frequency energy measures the heat load generated by instantaneous partial discharge / arc; and the rate of change of contact resistance captures the dynamic deterioration of contact points, quickly reflecting sudden imbalances. These three indicators are linearly combined with adjustable weights to form a thermal risk index, enabling the system to quantify instantaneous thermal load within milliseconds and compare it with preset thresholds, triggering early warnings, shutdowns, or maintenance commands in advance. Compared with traditional single resistance or discharge monitoring, this index provides a more complete thermal state picture, taking into account both static and transient risks, and can achieve real-time online monitoring without electrical current or contact, significantly reducing the risk of sudden overheating or failure.

[0038] Please refer to Figure 2 Another embodiment of the present invention provides a contact quality judgment system for a GIS disconnector switch, comprising: The sensing module is used to select the measuring points that are rigidly coupled to the mechanical link of the GIS disconnect switch, and to install fiber optic EFPI vibration sensors containing low-frequency and high-frequency channels at each measuring point. The data acquisition module is used to collect vibration signals of all channels in real time within a preset time window when the GIS disconnect switch is closed. The data processing module is used to extract features from the vibration signals of each channel to obtain the features of each channel; The diagnostic module is used to input the features of multiple channels into a pre-trained multi-channel model and output the contact resistance and thermal risk index. Based on the contact resistance and thermal risk index, the contact quality of the GIS disconnect switch is determined.

[0039] As described above, the beneficial effects of this invention are as follows: Utilizing fiber optic EFPI low-frequency / high-frequency dual-channel sensors distributed at the rigid coupling point of the contact finger's mechanical link, real-time acquisition of vibration during closing is performed, extracting time-domain, frequency-domain, and time-frequency characteristics. Then, a pre-trained multi-channel model is used to estimate the contact resistance and thermal risk index of the contact finger, achieving high-precision contact finger positioning and thermal safety early warning without electrical charge or contact. In this way, the vibration response characteristics of the contact finger at the moment of closing are perceived by the EFPI fiber optic sensor, enabling judgment of contact quality under non-visual conditions. When the contact resistance of a certain contact finger increases, the EFPI vibration signal can achieve: early identification of single-contact finger anomalies, tracking of local contact degradation trends, and early warning before potential hotspot formation, significantly improving the operational safety and maintenance efficiency of GIS disconnect switches.

[0040] The method and system for judging the contact quality of GIS disconnect switches described above are applicable to early warning of potential thermal hazards at the contact finger level of GIS disconnect switches under non-visual conditions. The following is a detailed description of the implementation methods: This embodiment proposes a non-electrical quantity diagnostic system for GIS disconnect switches based on fiber optic EFPI high-sensitivity vibration sensing, indicating that the switch is in position when closed. Please refer to... Figure 2 The system includes a sensing module, a data acquisition module, a data processing module, and a vibration module.

[0041] Sensing module: Several fiber optic EFPI vibration sensing probes, for low-frequency vibration detection, with cavity length and sensitive unit size selected according to the measurement point requirements.

[0042] The sensor employs an extrinsic Fabry-Perot interferometer (EFPI), where the diaphragm serves as the sensing element. The cavity length changes with the deformation of the diaphragm, causing variations in the intensity of reflected light. The vibration time-domain signal is obtained through intensity demodulation or white light demodulation. This structure has been proven to respond to both low-frequency vibrations and high-frequency partial discharge signals, and it can operate stably in strong electromagnetic environments.

[0043] Furthermore, to adapt to the characteristics of closing impact and contact finger vibration (low-frequency main peak + high-frequency components accompanied by arc), a "low-frequency vibration detection EFPI" structure is adopted. Key design points include: (1) Sensitive unit material: borosilicate glass, to obtain a suitable first-order characteristic frequency; (2) Sensitive unit size: Based on the simulation / optimization results, the diaphragm radius and thickness are selected. A radius of 7 mm and a thickness of 20 μm are selected to obtain a first-order characteristic frequency of about 1.1 kHz. (3) Initial cavity length: The initial cavity length ranges from 170 to 200 μm, and under the experimental conditions, it has the best linearity and sensitivity for low-frequency response; (4) Diaphragm / shell structure: The design of coupling the transducer sheet with the base ensures both sensitivity and mechanical reliability; the shell can reduce the displacement amplitude of the sensitive unit to prevent overload.

[0044] An independent EFPI probe is placed at each key measurement point (near the finger support, insulating bracket, or thin wall of the housing). Two to six probes can cover a group of fingers, which is beneficial for finger-by-finger or subgroup positioning. A 1×N beam splitter can be connected to the optical link to achieve multiple probe reuse, but the amplitude attenuation after beam splitting must be considered (actual measurements show that connecting a 1×8 beam splitter will cause a slight decrease in amplitude of 2% to 11%). In situations with limited cabling, it is recommended to use a distributed demodulator or multiple demodulators in parallel.

[0045] Regarding sensor installation and fiber optic routing, details are as follows: (1) Measurement point selection: Prioritize rigid surfaces that are rigidly coupled to the mechanical link of the contact finger or directly in contact with the foundation support structure of the contact finger to ensure that the closing impact can be transmitted with high fidelity; each group of contact fingers should have at least 2 measurement points (near end and far end) to achieve differential discrimination.

[0046] (2) Coupling method: Vibration transmission pads are used, which are both insulating and vibration transmission. Attention should be paid to the selection of materials to avoid affecting the sealing and insulation safety of the SF6 cavity.

[0047] (3) Fiber optic cable out: a special sealing device is used to seal the pressure-resistant fiber optic cable through the GIS shell. The cabling avoids direct contact with high-voltage components. The fiber optic cable termination is equipped with a fiber optic protective sleeve and a stress relief structure.

[0048] (4) Environmental compatibility: The probe protection level meets the working requirements of the GIS cavity (pressure resistance, corrosion resistance, and temperature rise resistance), and the shell and connectors meet the airtightness and insulation specifications. The above installation, sealing and mechanical fixing methods refer to the description of the equipment interface and shell treatment in the example patent.

[0049] The system also includes optical links and demodulators, specifically including: a stable light source, a circulator / splitter, and a photoelectric detection and intensity / white light demodulation module (real-time sampling rate ≥50kHz, which can be higher depending on the requirements).

[0050] (1) Light source: stable narrow linewidth laser or broadband white light source (white light is used for initial cavity length measurement, and laser is used for dynamic intensity demodulation); the system also retains the white light cavity length calibration channel to ensure the accuracy of the initial cavity length.

[0051] (2) Circulator / splitter: The circulator is used to introduce / recover reflected light; the splitter is used for multi-channel distribution (note the splitting ratio and decoupling); actual measurements show that connecting the splitter will cause amplitude attenuation but improve the system linearity.

[0052] (3) Demodulation method: Intensity demodulation is used in the dynamic detection stage to meet the real-time requirements; white light interferometry demodulation is used for offline or initialization to improve the accuracy of cavity length measurement.

[0053] (4) Sampling rate and resolution: Closing impact and micro-arc contain high-frequency components. The demodulator sampling rate is ≥50kS / s to ensure the capture of high-frequency components. The ADC resolution is ≥16-bit, which helps to maintain the distinguishability of small vibration amplitudes.

[0054] (5) Time synchronization: If multiple channels are deployed simultaneously, precise time synchronization is required for spatial positioning of single and multiple touches.

[0055] Data Acquisition Module (DAQ): Analog-to-digital conversion, time synchronization, data preprocessing and caching.

[0056] Specifically, data preprocessing includes: (1) High frequency interference resistance: There are high frequency components inside the EFPI system. Low amplitude low frequency signals need to be low-pass or band-pass filtered first to remove the high frequency noise associated with partial discharge inside the demodulator. In the experiment, the detection accuracy can be improved by about 67% after low-pass filtering.

[0057] (2) Bandpass splitting: The signal is divided into low-frequency channels (closing impulse, structural resonance, typically 0-2kHz) and high-frequency channels (micro-arc, plasma breakdown, typically 1kHz-20kHz), and different feature extraction strategies are used for each.

[0058] (3) Time window and denoising: The closing is an instantaneous event. A short time window (1-10ms) is used to perform short-time Fourier transform or wavelet transform, combined with threshold denoising and adaptive baseline correction.

[0059] Data processing module: filtering, time-frequency analysis (short-time Fourier / wavelet), feature extraction, event detection (contact instant, micro-arc), contact resistance and thermal risk inversion model, multi-sensor fusion and decision logic.

[0060] For each closing event, the following key features need to be extracted from each channel (for subsequent model input): (1) Temporal characteristics: Peak impact amplitude (V) peak ): Characterizes contact impact strength and insertion force.

[0061] Rise time / impact duration: characterizes the dynamics of mechanical coupling at the moment of contact.

[0062] The decay time constant (Q) characterizes the establishment and damping of the contact pressure of the finger.

[0063] (2) Frequency domain characteristics: The main harmonic frequency f0 (the inherent frequency of the structure) and its amplitude: structural resonance coding touch finger stiffness / coupling situation.

[0064] High Frequency Energy Ratio (HF) energy / Total energy Increased high-frequency energy is often associated with electric arc / partial discharge.

[0065] (3) Time-frequency domain characteristics: Arc event detection: short-term energy surge + wide spectrum response + continuous high-frequency oscillation (micro-arc jumping characteristics); can be quantified by wavelet packet energy entropy or spectral packet ratio.

[0066] Contact establishment curve slope (displacement equivalent): the initial rising slope of the impact signal is mapped to the contact pressure establishment rate.

[0067] The definition and quantification methods for these characteristics refer to the statistical and spectral analysis methods used in EFPI low-frequency / high-frequency experiments.

[0068] Diagnostic module: The diagnostic model and decision logic specifically include: (1) Arc detection and characterization: If a channel experiences a short-term surge in high-frequency energy within the closing time window, accompanied by continuous high-frequency components (the high-frequency band energy in the wavelet packet exceeds the threshold and the duration is > τ), arc The system determines whether an electric arc or a micro-arc exists; it uses multiple channels to simultaneously determine the arc's near-source direction (channel with the largest amplitude or the channel with the smallest phase difference). Further calculations of the arc duration and energy integral (energy spectrum integral) are used to classify the arc severity (slight—moderate—severe).

[0069] (2) Contact quality judgment (closed position judgment) Based on the displacement-vibration coupling model, the following logic is established: If the peak impact value is ≥ V min If the main harmonic frequency f0 and the decay time Q are within the normal range, then it is determined that "the mechanical position is in place and the contact quality is normal"; If the peak impact value is abnormally low (V) peak < α If the contact pressure is insufficient or the pressure decreases rapidly (Q drops significantly), it is determined that the contact pressure is "insufficient / the finger is too loose". If the peak impact is normal but the high-frequency energy is abnormally high (and there are micro-arc characteristics), it is determined that "although the mechanical parts are in place, there is poor contact / partial discharge or initial arc". If multiple channels show significant inconsistencies (phase / amplitude / energy), it is determined that "multiple touch fingers are out of sync or a single touch finger is abnormal", and the specific channel location is given.

[0070] The above determination is achieved through threshold logic or supervised learning models (such as support vector machines or random forests); the threshold is automatically set / updated by a calibration test library or online self-learning.

[0071] (3) Inverse model of contact resistance and thermal risk Establish a two-level mapping from features extracted by the data processing module to contact parameters: Vibration → Pressure / Contact Area Mapping: Using experimentally calibrated (vibration response sets under different contact pressures / contact areas) fitting model g1: Pc = g1(V peak (,f0,Q,…); The mapping from pressure / contact area to contact resistance: Based on the contact resistance theory, g2 is established as: Rc = g2(Pc, material parameters).

[0072] The combined result is: Rc = g2(g1(characteristics of vibration signal), material parameters).

[0073] The mapping principle of vibration to contact pressure / contact area is as follows: At the moment of final contact closure, the disconnecting switch contacts generate specific impact vibrations, the amplitude, frequency components, and energy distribution of which have a stable correspondence with the contact pressure and effective contact area. For example, the greater the contact pressure, the more intense the impact process, and the higher the vibration peak value (V). peak The higher the contact area, the greater the contact stiffness, and the higher the main vibration frequency (f0) usually becomes; the better the contact tightness, the faster the vibration decay rate, which is reflected in a higher quality factor Q.

[0074] The mapping principle of contact pressure / area → contact resistance is as follows: when the contact pressure is insufficient, the effective contact area Aeff decreases, the number of micro-protrusion contact points decreases, and the contact resistance increases; conversely, when the contact pressure increases, the micro-protrusions are flattened, the contact area increases, and the contact resistance decreases.

[0075] Then, by combining the arc energy deduction with the contact resistance increase rate, a thermal risk warning index can be calculated: PI=w1·Rc+w2·ArcEnergy+w3·ΔRc / Δt A thermal risk alarm is triggered when the PI exceeds a specified threshold. This inversion method obtains the mapping parameters through calibration based on EFPI sensor experiments.

[0076] The system also includes a calibration and learning module, specifically used for: offline calibration test library and online self-learning (threshold adaptation / model fine-tuning).

[0077] (1) Offline calibration bench: A finger-to-finger docking bench (which can control contact pressure, contact surface state and insertion force) is set up in the laboratory, and EFPI probes and direct temperature / contact resistance measurement devices are set up (for comparison). Vibration and electrical data under different contact conditions are collected to establish training sets and regression / classification models.

[0078] (2) Model training: Multi-channel features are used as input, and the target variables are: contact pressure, contact resistance, whether there is an electric arc, and thermal risk level. Random forest or gradient boosting tree is preferred for the model to ensure robustness, and simplified threshold logic is derived after training for rapid on-site decision-making.

[0079] (3) Online adaptive calibration: After the system goes online, the results of running events and manual inspections are compared periodically, and the model weights and thresholds are updated using incremental learning (to ensure long-term drift correction). This process is based on the concept of digital twin / model calibration to improve field adaptability.

[0080] The system also includes a human-machine interface and alarms, specifically for: historical records, graphical waveform display, alarm thresholds, and maintenance suggestions.

[0081] The overall system deployment is as follows: The EFPI probe is fixed at an appropriate location on the GIS disconnector housing using a rigid or coupled base, enabling the sensor to detect the vibration transmitted to the housing or conductor support structure during the contact finger closing process; the optical signal is exported to the demodulator and processing unit via a pressure-resistant optical fiber.

[0082] The following provides typical application examples for illustration.

[0083] Application Example 1 (Single Finger Detection): Step A: Arrange three EFPI probes (near end / middle end / far end) at the contact finger support of a GIS disconnector, and connect them to an intensity demodulator using a 1×3 beam splitter; Step B: Synchronize the demodulator with the DAQ and set the sampling rate to 100 kS / s; use white light demodulation to calibrate the initial cavity length of each probe (170 μm target). Step C: Trigger closing, collect 10ms time window data in real time and perform real-time filtering and time-frequency analysis; Step D: Extract features such as peak impact, f0, and HF energy; input the features into the trained model, return the classification: normal closing / insufficient contact / existing arc / abnormal single contact finger, and output the contact resistance estimate and thermal risk index; Step E: If the result is abnormal, record the waveform and trigger the inspection alarm and maintenance order.

[0084] Application Example 2 (On-site Batch Deployment): The same GIS unit deploys EFPI probes on several disconnect switches and connects them to a centralized monitoring platform. The platform identifies common degradation trends based on the statistical trends of multiple switches and generates maintenance plans.

[0085] Combining EFPI low-frequency test data with sensor design: Sensitivity: In the range of 0–0.6g, the sensitivity of ordinary EFPI can reach about 100mV / 0.01g; low-frequency detection EFPI can maintain good linearity in the range of 0–1g.

[0086] Frequency band: Low-frequency response target approximately 0–2kHz (covering closing impact and structural natural frequency), high-frequency branch for arc detection can be extended to 10–15kHz.

[0087] Anti-high frequency interference: Even after adding a low-pass filter and a beam splitter, it can still maintain good linearity and usable signal-to-noise ratio. Actual measurements show that the amplitude drop is controllable (below 11% or 15-19% depending on the specific modifications) when a beam splitter is connected or the reflectivity is reduced.

[0088] By deploying low-frequency optimized EFPI sensors at key closing positions of GIS disconnectors, constructing a signal processing chain with time-frequency separation capabilities, and establishing a mapping relationship between vibration, contact, resistance, and thermal risk based on calibration data, this invention can: accurately determine whether the closing is truly in place and assess contact quality under non-visual, non-electrical conditions; detect micro-arcs and their energy / duration at the moment of closing online; and provide overheat risk warnings at the finger level before temperature rise occurs; thereby significantly improving the operational safety and maintenance efficiency of GIS enclosed switchgear.

[0089] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for judging the contact quality of a GIS disconnector switch, characterized in that, Including the following steps: S1. Select the measurement points that are rigidly coupled to the mechanical link of the GIS disconnect switch, and install fiber optic EFPI vibration sensors containing low-frequency and high-frequency channels at each measurement point. S2. When the GIS disconnect switch is closed, the vibration signals of all channels are collected in real time within a preset time window; S3. By extracting features from the vibration signals of each channel, the features of each channel are obtained; S4. Input the features of multiple channels into a pre-trained multi-channel model, and output the contact resistance and thermal risk index. Determine the contact quality of the GIS disconnector based on the contact resistance and the thermal risk index.

2. The method for judging the contact quality of a GIS disconnector according to claim 1, characterized in that, Step S3 includes: Time-domain analysis was performed on the vibration signal of the low-frequency channel to extract the impact peak value and decay time. Frequency-domain analysis was performed on the vibration signal of the low-frequency channel to extract the main harmonic frequency. Time-frequency analysis is performed on the vibration signal of the high-frequency channel to calculate the high-frequency band energy within a preset high-frequency range.

3. The method for judging the contact quality of a GIS disconnector according to claim 2, characterized in that, Step S3 is followed by determining the type of finger contact quality based on the characteristics of each channel, specifically: If the peak value of the impact is greater than the minimum value of the preset impact range, and the main harmonic frequency and the decay time are both within the corresponding normal range, then it is determined that the contact finger of the GIS disconnect switch is in mechanical position and the contact quality is normal. If the peak impact value is less than the preset impact threshold, or the rate of decrease of the decay time reaches the preset rate, it is determined that the contact pressure of the GIS disconnect switch is insufficient / the contact finger is too loose. If the impact peak is located within the preset impact range and the high-frequency band energy exceeds the energy threshold, it is determined that the contact finger of the GIS disconnect switch is mechanically in place, but there is poor contact / partial discharge or initial arc. If the characteristic differences of multiple channels are greater than the preset difference threshold, it is determined that the multiple contacts of the GIS disconnect switch are out of sync or the single contact is abnormal.

4. The method for judging the contact quality of a GIS disconnector according to claim 2, characterized in that, Step S4 involves inputting features from multiple channels into a pre-trained multi-channel model and outputting the contact resistance, including: The multi-channel model includes a first fitting model and a second fitting model; the first fitting model is established based on the vibration characteristics under different historical contact pressures and contact areas, and the second fitting model is established based on the contact resistance under different historical contact pressures or contact areas. The impact peak value, the decay time, the main harmonic frequency, and the high-frequency band energy are input into the first fitting model, and the contact pressure and contact area are output by the first fitting model. The contact pressure and contact area are input into the second fitting model, and the contact resistance is obtained from the output of the second fitting model.

5. The method for judging the contact quality of a GIS disconnector according to claim 4, characterized in that, Output thermal risk index, including: The thermal risk index PI is calculated based on the contact resistance Rc and the high-frequency band energy ArcEnergy. PI=w1·Rc+w2·ArcEnergy+w3·ΔRc / Δt In the formula, w1 represents the weight of contact resistance, w2 represents the weight of high-frequency band energy, ΔRc / Δt represents the rate of change of contact resistance, and w3 represents the weight of the rate of change of contact resistance.

6. A contact quality judgment system for a GIS disconnector switch, characterized in that, include: The sensing module is used to select the measuring points that are rigidly coupled to the mechanical link of the GIS disconnect switch, and to install fiber optic EFPI vibration sensors containing low-frequency and high-frequency channels at each measuring point. The data acquisition module is used to collect vibration signals of all channels in real time within a preset time window when the GIS disconnect switch is closed. The data processing module is used to extract features from the vibration signals of each channel to obtain the features of each channel; The diagnostic module is used to input the features of multiple channels into a pre-trained multi-channel model and output the contact resistance and thermal risk index. Based on the contact resistance and thermal risk index, the contact quality of the GIS disconnect switch is determined.

7. The contact quality judgment system for a GIS disconnector according to claim 6, characterized in that, The execution steps of the data acquisition module specifically include: Time-domain analysis was performed on the vibration signal of the low-frequency channel to extract the impact peak value and decay time. Frequency-domain analysis was performed on the vibration signal of the low-frequency channel to extract the main harmonic frequency. Time-frequency analysis is performed on the vibration signal of the high-frequency channel to calculate the high-frequency band energy within a preset high-frequency range.

8. The contact quality judgment system for a GIS disconnector according to claim 7, characterized in that, The diagnostic module is also used to determine the type of finger contact quality based on the characteristics of each channel, specifically: If the peak value of the impact is greater than the minimum value of the preset impact range, and the main harmonic frequency and the decay time are both within the corresponding normal range, then it is determined that the contact finger of the GIS disconnect switch is in mechanical position and the contact quality is normal. If the peak impact value is less than the preset impact threshold, or the rate of decrease of the decay time reaches the preset rate, it is determined that the contact pressure of the GIS disconnect switch is insufficient / the contact finger is too loose. If the impact peak is located within the preset impact range and the high-frequency band energy exceeds the energy threshold, it is determined that the contact finger of the GIS disconnect switch is mechanically in place, but there is poor contact / partial discharge or initial arc. If the characteristic differences of multiple channels are greater than the preset difference threshold, it is determined that the multiple contacts of the GIS disconnect switch are out of sync or the single contact is abnormal.

9. The contact quality judgment system for a GIS disconnector according to claim 7, characterized in that, The features from multiple channels are input into a pre-trained multi-channel model, and the output is the contact resistance, including: The multi-channel model includes a first fitting model and a second fitting model; the first fitting model is established based on the vibration characteristics under different historical contact pressures and contact areas, and the second fitting model is established based on the contact resistance under different historical contact pressures or contact areas. The impact peak value, the decay time, the main harmonic frequency, and the high-frequency band energy are input into the first fitting model, and the contact pressure and contact area are output by the first fitting model. The contact pressure and contact area are input into the second fitting model, and the contact resistance is obtained from the output of the second fitting model.

10. The contact quality judgment system for a GIS disconnector according to claim 9, characterized in that, Output thermal risk index, including: The thermal risk index PI is calculated based on the contact resistance Rc and the high-frequency band energy ArcEnergy. PI=w1·Rc+w2·ArcEnergy+w3·ΔRc / Δt In the formula, w1 represents the weight of contact resistance, w2 represents the weight of high-frequency band energy, ΔRc / Δt represents the rate of change of contact resistance, and w3 represents the weight of the rate of change of contact resistance.

Images