Online monitoring system for early warning of discharge fault of air switch cabinet

Abstract

The invention relates to the technical field of on-line detection and fault diagnosis of power equipment, in particular to an on-line monitoring system for early warning of a discharge fault of an air switch cabinet. The system comprises a sensing module which is arranged at a preset position in the air switch cabinet and is used for detecting the temperature and humidity of air in the air switch cabinet, and components and component contents of gas decomposers generated by discharging; the preset position is determined according to a simulation result of an established gas diffusion simulation model of the air switch cabinet; the signal transmitting module is connected with the sensing module and is used for receiving the analog signal sent by the sensing module and converting the analog signal into a digital signal; and the fault diagnosis module is connected with the signal transmitting module and is used for receiving and analyzing the digital signal to obtain the fault type of the air switch cabinet. The problems that in the prior art, when gas decomposition products are detected, equipment is large in size and high in manufacturing cost, and online monitoring is difficult to achieve can be solved.

Description

Technical Field

[0001] This invention relates to the field of online detection and fault diagnosis technology for power equipment, and in particular to an online monitoring system for early warning of discharge faults in air switchgear. Background Technology

[0002] In recent years, with social development, power equipment has shown a trend towards miniaturization and integration. Air, as a natural insulating medium, is widely used in power equipment. Air-insulated switchgear has gradually become a hot topic, with a large number of units in the power system, wide coverage, and complex operating environment. Its reliable operation is of great significance to the safety of the power grid. However, during its manufacturing, installation, and operation, insulation defects such as impurities, bubbles, and burrs are inevitable. During long-term operation of the switchgear, these insulation defects can lead to partial discharge, causing the insulation medium to deteriorate and triggering accidents. Therefore, it is necessary to monitor partial discharge.

[0003] Partial discharge caused by insulation defects can lead to air decomposition, generating a series of gaseous decomposition products. The operating status of air switchgear can be assessed by detecting the component content of these products. Currently, commonly used methods for detecting gaseous decomposition products include chromatography, mass spectrometry, and spectroscopy. However, these methods suffer from drawbacks such as large equipment size, high cost, and difficulty in achieving online monitoring of air switchgear. Summary of the Invention

[0004] This invention provides an online monitoring system for early warning of discharge faults in air switchgear, which solves the problems of large equipment size, high cost, and difficulty in online monitoring when detecting gas decomposition products in the prior art.

[0005] In a first aspect, embodiments of the present invention provide an online monitoring system for early warning of discharge faults in air switchgear, comprising: a sensing module, a signal transmission module, and a fault diagnosis module; The sensing module is set at a preset position inside the air switch cabinet and is used to detect the temperature, humidity, composition and content of gas decomposition products generated by discharge in the air switch cabinet; the preset position is determined based on the simulation results of the established gas diffusion simulation model of the air switch cabinet. The signal transmission module is connected to the sensing module and is used to receive the analog signal sent by the sensing module and convert the analog signal into a digital signal. The fault diagnosis module is connected to the signal transmission module and is used to receive and analyze the digital signal to obtain the fault type of the air switch cabinet.

[0006] In one possible implementation, the sensing module includes: a temperature and humidity sensor, at least two gas sensors, and a temperature control device; The temperature and humidity sensor is used to detect the temperature and humidity of the air inside the air switch cabinet; The at least two gas sensors are used to detect the composition and content of gaseous decomposition products generated by partial discharge due to insulation faults in the air switch cabinet; The temperature control device is used to control the temperature inside the air switch cabinet within a preset range based on the detected temperature of the air inside the air switch cabinet.

[0007] In one possible implementation, the gas sensor includes an electrochemical ozone sensor and a chemiluminescence sensor; An SCR module is installed at the air inlet of the electrochemical ozone sensor. The SCR module is used to decompose the gaseous products at a preset temperature. A catalytic reduction reaction is carried out, enabling the electrochemical ozone sensor to detect the gas decomposition products inside the air switch cabinet. The content; The air inlet of the chemiluminescence sensor is set Module, the The module is used to absorb the gas decomposition products. This enables the chemiluminescence sensor to detect the gas decomposition products inside the air switch cabinet. The content of.

[0008] In one possible implementation, a protective coating is provided on the outer wall of the gas sensor.

[0009] In one possible implementation, the protective coating is a hydrophobic silicon dioxide film.

[0010] In one possible implementation, the fault diagnosis module is a host computer, which has a diagnostic model for training the diagnostic model to obtain a target diagnostic model, receives the digital signal, inputs the digital signal into the target diagnostic model, and outputs the analysis result, which is the fault type of the air switch cabinet.

[0011] In one possible implementation, training the diagnostic model includes: The data sent by the signal transmitter module is acquired. The data is collected by the sensor module during discharge tests of different voltage levels and different discharge fault types in the air switch cabinet. The data includes temperature and humidity data, composition and content of gas decomposition products. Detect whether the humidity data is greater than a preset humidity threshold; If the humidity data is greater than the preset humidity threshold, the component content of the gas decomposition products is multiplied by the preset humidity correction factor. If the humidity data is not greater than the preset humidity threshold, the training data is normalized, or the corrected component content and temperature and humidity data are normalized to obtain training samples. Using the training samples as input and the fault type as output, the initial diagnostic model is trained to obtain the trained diagnostic model.

[0012] In one possible implementation, the activation function in the pattern layer of the initial diagnostic model is calculated using the following formula: ; in, This represents the activation function value for each sample. This represents the input for training. Indicates training samples, Represents the smoothing coefficient. This indicates the sample number of the training sample.

[0013] In one possible implementation, The methods for determining include: During the discharge test, determine the corresponding discharge fault types. And set the smoothing coefficient-fault type table; Determine the corresponding fault type in the smoothing coefficient-fault type table. .

[0014] In one possible implementation, the smoothing coefficient in the smoothing coefficient-fault type table is a range of smoothing coefficient values. During the training of the initial diagnostic model, the diagnostic error of the validation samples is calculated in real time. If the diagnostic error exceeds the error threshold, the gradient descent method is used to dynamically adjust the value of the smoothing coefficient.

[0015] This invention provides an online monitoring system for early warning of discharge faults in air switchgear. A sensor module positioned at a predetermined location inside the air switchgear detects the temperature, humidity, and composition and content of gaseous decomposition products generated during discharge. The predetermined location is determined based on simulation results from a gas diffusion simulation model of the air switchgear. A signal transmission module connected to the sensor module receives analog signals and converts them into digital signals. A fault diagnosis module connected to the signal transmission module receives and analyzes the digital signals to determine the fault type of the air switchgear. By positioning the sensor module at the location determined by the gas diffusion simulation model, this invention enables the sensor module to quickly detect more accurate data. Furthermore, the sensor module used has advantages such as low power consumption, small size, and low cost, allowing for online monitoring of air switchgear and its large-scale application in power systems. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of an online monitoring system for early warning of discharge faults in air switchgear provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of the sensing module provided in an embodiment of the present invention; Figure 3 This is a flowchart illustrating the implementation of training a diagnostic model according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of the gas discharge test platform provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of the diagnostic model provided in an embodiment of the present invention. Detailed Implementation

[0018] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0019] To make the objectives, technical solutions, and advantages of the present invention clearer, specific embodiments will be described below in conjunction with the accompanying drawings.

[0020] Figure 1 A schematic diagram of an online monitoring system for early warning of discharge faults in air switchgear provided in an embodiment of the present invention is described in detail below: An online monitoring system for early warning of discharge faults in air switchgear may include: a sensing module 1, a signal transmission module 2, and a fault diagnosis module 3; Sensing module 1 is set at a preset position inside the air switch cabinet to detect the temperature, humidity, composition and content of gas decomposition products generated by discharge in the air inside the air switch cabinet; the preset position is determined based on the simulation results of the established gas diffusion simulation model of the air switch cabinet. The signal transmitter module 2 is connected to the sensor module 1 and is used to receive the analog signal sent by the sensor module 1 and convert the analog signal into a digital signal. The fault diagnosis module 3 is connected to the signal transmission module 2 and is used to receive and analyze digital signals to obtain the fault type of the air switch cabinet.

[0021] In one embodiment, in order to enable the online monitoring system for early warning of discharge faults in air switchgear to better detect fault gases, a gas diffusion simulation model of the air switchgear is established. Based on the simulation results, the arrangement position of the sensor module 1 in the switchgear is determined to ensure that the sensor module 1 can detect fault gases faster and more accurately.

[0022] In one embodiment, such as Figure 2 As shown, the sensing module 1 includes: a temperature and humidity sensor 11, at least two gas sensors 12, and a temperature control device 13; Temperature and humidity sensor 11 is used to detect the temperature and humidity of the air inside the air switch cabinet; At least two gas sensors 12 are used to detect the composition and content of gaseous decomposition products generated by partial discharge due to insulation faults in the air switchgear. Temperature control device 13 is used to control the temperature inside the air switch cabinet within a preset range based on the detected temperature of the air inside the air switch cabinet.

[0023] Optionally, the temperature and humidity sensor 11 can be a sensor that integrates a temperature sensor and a humidity sensor, and the temperature and humidity sensing chip of the temperature and humidity sensor 11 can be an SHT30. Alternatively, it can be two independent sensors. In this embodiment, the form of the temperature and humidity sensor 11 is not limited.

[0024] Optionally, since the temperature in the air switch cabinet can affect the detection accuracy of the gas sensor 12, it is necessary to control the temperature in the air switch cabinet to keep it stable within a certain range. This will ensure a stable working environment for the gas sensor 12 and extend its service life.

[0025] In one embodiment, the gas sensor 12 includes an electrochemical ozone sensor 121 and a chemiluminescence sensor 122; optionally, the various sensors in the sensing module 1 can form a sensing array.

[0026] The electrochemical ozone sensor 121 has a Selective Catalytic Reduction Module (SCR) at its air inlet. The SCR module is used to decompose nitrogen oxides in the gaseous products at a preset temperature. The catalytic reduction reaction is carried out, enabling the electrochemical ozone sensor to detect ozone in the gas decomposition products inside the air switch cabinet. The content of ).

[0027] The SCR module targets the gaseous decomposition products generated during the discharge of air switchgear. and The design addresses the mutual interference issues between components, with its core function being to achieve, through a specific catalyst, control interference at lower temperature conditions, such as 50℃-150℃. Selective reduction, thereby eliminating its effect on Interference in detection. Here, a specific catalyst can be... .

[0028] Since traditional SCRs require operation at temperatures between 200℃ and 400℃, while the interior of an air switchgear is typically at room temperature, the SCR module in this embodiment does not require additional high-temperature heating and can be directly integrated into the pre-amplifier unit of the electrochemical ozone sensor 121. This avoids the impact of high-temperature equipment on the internal environment of the air switchgear and also avoids the use of liquid absorption. The SCR module is suitable for long-term online monitoring in enclosed spaces, avoiding problems such as easy leakage and frequent maintenance.

[0029] In practical applications, when a fault discharge occurs, the gas inside the air switch cabinet first passes through the low-temperature SCR module. The ozone is efficiently reduced and removed, then enters the electrochemical ozone sensor for concentration detection, ensuring that the detection results only reflect... The true concentration improves the accuracy of discharge fault early warning.

[0030] The electrochemical ozone sensor 121 is small in size, has a response time of less than 10 seconds, and is highly resistant to dust and humidity interference.

[0031] The air inlet of the chemiluminescence sensor 122 is set Module, The module is used to absorb gas decomposition products. This enables the chemiluminescence sensor to detect gas decomposition products inside the air switch cabinet. The content of.

[0032] A module is a type of module used to eliminate The core function of the interference pre-processing module is to... The catalytic effect decomposes ozone in the air at room temperature, ensuring safety. The accuracy of the test.

[0033] The gaseous decomposition products generated by fault discharge in the air switchgear are and The mixture of gases, and This will interfere with the detection of the chemiluminescence sensor 122, such as affecting the response signal of the chemiluminescence sensor 122. At this time, The module serves as a pre-processing unit for the chemiluminescence sensor 122, utilizing... Catalytic activity at room temperature, A decomposition reaction occurs, which will interfere with... It is converted into harmless oxygen. ).

[0034] The module requires no additional heating or complex equipment; it can be achieved solely through a solid catalyst. The efficient decomposition avoids the leakage risk and high maintenance cost of liquid absorption methods (such as KI solution), and can be stably integrated into the sensing module 1 for a long time, providing a pure gas environment for the accurate detection of the subsequent chemiluminescence sensor 122, and ensuring the reliability of discharge fault early warning.

[0035] against The chemiluminescence sensor 122 can be a miniaturized chemiluminescence sensor with a sensitivity of up to ppb, capable of detecting low concentrations of NO generated by early weak discharge.

[0036] Optionally, for accuracy, multiple electrochemical ozone sensors 121 and multiple chemiluminescence sensors 122 can be used. The average value of the component content of gas decomposition products detected by the multiple electrochemical ozone sensors 121 is taken as the detection value of the gas sensor 12. Similarly, the average value of the component content of gas decomposition products detected by the multiple chemiluminescence sensors 122 can be taken. However, in this embodiment, the sensing module 1 needs to be miniaturized to adapt to the online monitoring requirements of the air switch cabinet; therefore, two electrochemical ozone sensors 121 and two chemiluminescence sensors 122 can be used.

[0037] In one embodiment, water vapor may be present in the air switch cabinet. When the ambient humidity increases, the humidity inside the air switch cabinet will also increase, which will affect the detection accuracy of the gas sensor 12. In fact, long-term operation in a high-humidity environment may damage the gas sensor 12. Therefore, a protective coating is provided on the outer wall of the gas sensor 12 to prevent water vapor from being directly adsorbed.

[0038] Optionally, the protective coating is a hydrophobic silica film.

[0039] Optionally, the gas sensing section of the sensing module may include a reference voltage, a non-inductive sampling resistor, and a filtering circuit. The reference voltage can use a linear voltage regulator chip TPS7A4700RGWT, which provides a stable output voltage.

[0040] The signal transmission module 2 may include a power supply unit, an analog-to-digital conversion unit, an electromagnetic compatibility (EMC) protection circuit, a microcontroller minimum system unit, and a wireless communication unit. The wireless communication unit is connected to the fault diagnosis module 3 and can convert analog signals into digital signals and send them to the fault diagnosis module 3.

[0041] The analog-to-digital conversion circuit in the analog-to-digital conversion unit uses the multi-channel analog-to-digital conversion chip ADS1256 to convert the analog signals collected by the sensing module into digital signals.

[0042] The EMC protection circuit is an important component of the signal transmission module 2. Its core function is to suppress electromagnetic interference (EMI) and prevent it from causing unacceptable electromagnetic interference to other equipment or systems in the environment, thus ensuring the stable operation of the device in complex electromagnetic environments.

[0043] The minimum system unit of a microcontroller can be the STM32F103C8T6 microcontroller.

[0044] The fault diagnosis module 3 can be a host computer that can identify the operating status of the air switch cabinet based on the relationship between the gas component content detected by the sensor module 1 and the status of the air switch cabinet, taking into account the influence of temperature and humidity on the discharge decomposition characteristics, through a diagnostic model.

[0045] The diagnostic model can be a probabilistic neural network model, which has strong adaptability and classification ability and has good results in the field of fault diagnosis.

[0046] In one embodiment, the fault diagnosis module is a host computer, which is equipped with a diagnostic model for training the diagnostic model to obtain a target diagnostic model. The host computer receives digital signals and inputs the digital signals into the target diagnostic model, and outputs analysis results, which are the fault types of the air switch cabinet.

[0047] In one embodiment, such as Figure 3 As shown, training a diagnostic model can include the following steps: Step 301: Obtain the data sent by the signal transmitter module. The data is collected by the sensor module during discharge tests of different voltage levels and different discharge fault types in the air switch cabinet. The data includes temperature and humidity data, composition and content of gas decomposition products.

[0048] Temperature and humidity in the air switch cabinet can affect the detection accuracy of the sensor. Therefore, in this embodiment, temperature and humidity data are collected. Firstly, temperature data can be used for temperature control to ensure that the sensing module 1 operates in a stable temperature environment. Secondly, temperature and humidity data can be used to train a diagnostic model. Based on the collected temperature and humidity data, a dynamic correction model of temperature and humidity-detection deviation table is established to learn the historical detection error patterns of the sensor under different temperatures and humidity, and to make real-time reverse corrections to the current detection value.

[0049] Optionally, to study the discharge decomposition characteristics of air-insulating media in switchgear, a gas discharge experimental platform can be built, such as... Figure 4 As shown, a discharge simulation platform is used to simulate discharge faults of different voltage levels and types. The gas in the sealed cavity 42 is evacuated by a vacuum pump 41, and then pure air is injected. A transformer 43 provides high voltage to break down the air-insulating medium. A current probe 44 and a voltage probe 45 are then used to measure the discharge waveform, which is displayed on an oscilloscope 46. By changing the discharge type and adjusting the electrode structure, simulations of different types of discharge faults are achieved, and the breakdown voltage and current waveforms under different conditions are recorded. The discharge sample gas generated by air discharge is analyzed using Fourier transform infrared spectroscopy 47 to determine the peak position, wavenumber, and peak shape characteristics of the waveform, identifying the components and their contents of the air discharge decomposition products.

[0050] Step 302: Detect whether the humidity data is greater than the preset humidity threshold.

[0051] In this embodiment, since humidity data can affect the detection accuracy of the sensor, a humidity correction factor is used to further correct the value detected by the gas sensor in addition to setting a protective coating on the gas sensor.

[0052] Optionally, the preset humidity threshold can be set according to requirements. In this embodiment, the value of the preset humidity threshold is not limited. For example, the preset humidity threshold can be 75%, 80%, or 85%.

[0053] Step 303: If the humidity data is greater than the preset humidity threshold, the component content of the gas decomposition products is multiplied by the preset humidity correction factor.

[0054] If the humidity data is greater than the preset humidity threshold, it means that the humidity inside the air switch cabinet is too high, which affects the component content of gas decomposition products detected by the gas sensor. Therefore, a preset humidity correction factor is used to multiply the component content of gas decomposition products to make the features input into the diagnostic model closer to the real gas composition and reduce the deviation of the diagnostic model caused by environmental interference.

[0055] Optionally, the preset humidity correction factor can be a fixed value, such as 0.8, or it can be set according to the humidity data and the type of gas sensor.

[0056] In one embodiment, ozone ( In high humidity environments, it easily reacts with water vapor, such as This results in a low detection value from the electrochemical ozone sensor 121. The preset humidity correction factor needs to be increased as humidity rises in order to compensate for signal attenuation.

[0057] The system detects whether the humidity data is less than or equal to a first humidity threshold. If the humidity data is less than or equal to the first humidity threshold, the electrochemical ozone sensor 121... The adsorption capacity is enhanced, and the response signal is prone to being too high. The preset humidity correction factor is set to the first value; for example, the first humidity threshold can be set to 30%, and the first value can be set to 0.9.

[0058] If the humidity data is greater than the first humidity threshold and less than or equal to the second humidity threshold, then the humidity... The interference of the detected value is small, and the preset humidity correction factor is the second value; for example, the second humidity threshold can be 60%, and the second value can be 1.0, that is, the detected value of the electrochemical ozone sensor 121 is not corrected.

[0059] If the humidity data is greater than the second humidity threshold and less than or equal to the third humidity threshold, the preset humidity correction factor is set to the third value. For example, the third humidity threshold can be set to 80%, and the third value can be set to 1.1, which is to slightly compensate and correct the detection value of the electrochemical ozone sensor 121.

[0060] If the humidity data is greater than the third humidity threshold Decomposition is accelerated, and the signal is significantly lower. The preset humidity correction factor is set to the fourth value; for example, the fourth value can be 1.2, which increases the compensation and correction of the detection value of the electrochemical ozone sensor 121.

[0061] In one embodiment, nitrogen dioxide ( It is readily soluble in water and undergoes the following reaction: High humidity will reduce its effective concentration on the surface of the chemiluminescence sensor 122. The correction logic is similar to that of the electrochemical ozone sensor 121, but the values ​​are different.

[0062] If the humidity data is less than or equal to 40%, it indicates that the humidity is low. It exhibits high stability, with the chemiluminescence sensor 122 showing a slightly stronger response; the preset humidity correction factor can be set to 0.95. If the humidity data is greater than 40% and less than or equal to 70%, the preset humidity correction factor can be 1.0; If the humidity data is greater than 70% and less than or equal to 90%, the humidity is high. Increased dissolution loss leads to lower detection values ​​from the chemiluminescence sensor 122; the preset humidity correction factor can be set to 1.15. If the humidity data is greater than 90%, it is considered extremely high humidity. The dissolution loss is severe, and the signal is significantly low. The preset humidity correction factor can be set to 1.3.

[0063] The value of the above-mentioned preset humidity correction factor can be obtained through... Figure 2 The gas discharge test platform shown was calibrated, and the true gas concentration was obtained by Fourier transform infrared spectrometer under different humidity conditions. The deviation between the sensor detection value and the true value was compared, and the optimal humidity correction coefficient for each humidity range was derived in reverse. Finally, a dynamic correction model adapted to the air switch cabinet environment was formed, which further improved the accuracy of fault diagnosis.

[0064] Step 304: If the humidity data is not greater than the preset humidity threshold, then the training data is normalized, or the corrected component content and temperature and humidity data are normalized to obtain training samples.

[0065] Step 305: Use the training samples as input and the fault type as output to train the initial diagnostic model and obtain the trained diagnostic model.

[0066] The initial diagnostic model is an untrained diagnostic model, which may include an input layer, a pattern layer, a summation layer, and an output layer.

[0067] Among them, such as Figure 5 As shown, the input layer consists of n neurons, the number of which is equal to the dimension of the input vector.

[0068] The pattern layer consists of m hidden neurons. The activation function in the pattern layer of the initial diagnostic model is calculated using the following formula: ; in, This represents the activation function value for each sample. This represents the input for training. Indicates training samples, Represents the smoothing coefficient. This indicates the sample number of the training sample.

[0069] In one embodiment, The determination method may include: During the discharge test, determine the corresponding discharge fault types. And set the smoothing coefficient-fault type table; Determine the corresponding fault type in the smoothness coefficient-fault type table. .

[0070] Optionally, the smoothing coefficient in the fault type table is a range of values; for example, for the tip discharge sample. The value can be between 0.3 and 0.5 for the air gap discharge sample. Values ​​can range from 0.5 to 0.7, etc.

[0071] During the training of the initial diagnostic model, the diagnostic error of the validation samples is calculated in real time. If the diagnostic error exceeds the error threshold, the gradient descent method is used to dynamically adjust the value of the smoothing coefficient.

[0072] It should be noted that if the diagnostic error exceeds the error threshold, and the diagnostic error increases, then the threshold should be increased. If the diagnostic error is greater than the error threshold, and the diagnostic error decreases, then decrease. .

[0073] like Figure 5 As shown, the number of neurons in the summation layer is the same as the total number of fault categories. The sum of the outputs of the pattern layer nodes corresponding to training samples of the same fault category is: ; in, For sample types, Let be the dimension of the input feature vector, and let be the fault types in the training samples. The samples are indivual.

[0074] The neuron with the highest posterior probability in the output summation layer: .

[0075] Using the prepared training samples as input, the probability relationship between the feature vector and the fault mode is calculated and continuously trained to obtain the optimal probabilistic neural network suitable for discharge fault diagnosis.

[0076] After training, discharge experiments with different voltage levels and discharge types were conducted to obtain test samples. The response signal of the sensor module during a discharge fault was input into the diagnostic model to determine whether the output fault type was consistent with the fault type corresponding to the discharge test, in order to verify the reliability of the diagnostic model.

[0077] This invention provides an online monitoring system for early warning of discharge faults in air switchgear. A sensor module, positioned at a predetermined location inside the air switchgear, detects the temperature, humidity, and composition and content of gaseous decomposition products generated during discharge. The predetermined location is determined based on simulation results from a gas diffusion simulation model of the air switchgear. A signal transmission module, connected to the sensor module, receives analog signals from the sensor module and converts them into digital signals. A fault diagnosis module, also connected to the signal transmission module, receives and analyzes the digital signals to determine the fault type of the air switchgear. This invention, by positioning the sensor module at a location determined by the gas diffusion simulation model, enables the sensor module to quickly detect more accurate data. Furthermore, the sensor module used has advantages such as low power consumption, small size, and low cost, allowing for online monitoring of air switchgear and its large-scale application in power systems.

[0078] In this embodiment, interference gas absorption modules are respectively set at the front end of the gas sensors. For example, an SCR module is set at the air inlet of the electrochemical ozone sensor, and a chemiluminescence sensor is set at the air inlet. The module reduces the content of interfering gases, enabling the gas sensor to detect specific gases and improve detection accuracy. In addition, a protective coating is applied to the outer wall of the gas sensor to prevent direct adsorption of water vapor, further improving the detection accuracy of the gas sensor.

[0079] Finally, by employing temperature control equipment to stabilize the temperature in the air switch cabinet, the detection accuracy of the gas sensor is further improved.

[0080] The lifespan of gas sensors can be extended by applying a protective coating and installing temperature control devices.

[0081] In this embodiment of the invention, a preset humidity correction factor is set during the diagnostic model training process to make the training data input into the diagnostic model closer to the actual gas composition, reducing model bias caused by environmental interference. Furthermore, the smoothing coefficient in the activation function is dynamically adjusted to improve the diagnostic accuracy, real-time performance, and environmental adaptability of the diagnostic model.

[0082] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0083] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. An online monitoring system for early warning of discharge faults in air switchgear, characterized in that, include: Sensing module, signal transmission module, and fault diagnosis module; The sensing module is set at a preset position inside the air switch cabinet and is used to detect the temperature, humidity, composition and content of gas decomposition products generated by discharge in the air switch cabinet; the preset position is determined based on the simulation results of the established gas diffusion simulation model of the air switch cabinet. The signal transmission module is connected to the sensing module and is used to receive the analog signal sent by the sensing module and convert the analog signal into a digital signal. The fault diagnosis module is connected to the signal transmission module and is used to receive and analyze the digital signal to obtain the fault type of the air switch cabinet.

2. The online monitoring system for early warning of discharge faults in air switchgear according to claim 1, characterized in that, The sensing module includes: a temperature and humidity sensor, at least two gas sensors, and a temperature control device; The temperature and humidity sensor is used to detect the temperature and humidity of the air inside the air switch cabinet; The at least two gas sensors are used to detect the composition and content of gaseous decomposition products generated by partial discharge due to insulation faults in the air switch cabinet; The temperature control device is used to control the temperature inside the air switch cabinet within a preset range based on the detected temperature of the air inside the air switch cabinet.

3. The online monitoring system for early warning of discharge faults in air switchgear according to claim 2, characterized in that, The gas sensor includes an electrochemical ozone sensor and a chemiluminescence sensor; An SCR module is installed at the air inlet of the electrochemical ozone sensor. The SCR module is used to decompose the gaseous products at a preset temperature. A catalytic reduction reaction is carried out, enabling the electrochemical ozone sensor to detect the gas decomposition products inside the air switch cabinet. The content; The air inlet of the chemiluminescence sensor is set Module, the The module is used to absorb the gas decomposition products. This enables the chemiluminescence sensor to detect the gas decomposition products inside the air switch cabinet. The content of.

4. The online monitoring system for early warning of discharge faults in air switchgear according to claim 3, characterized in that, A protective coating is provided on the outer wall of the gas sensor.

5. The online monitoring system for early warning of discharge faults in air switchgear according to claim 4, characterized in that, The protective coating is a hydrophobic silicon dioxide film.

6. The online monitoring system for early warning of discharge faults in air switchgear according to any one of claims 1-5, characterized in that, The fault diagnosis module is a host computer, which is equipped with a diagnostic model for training the diagnostic model to obtain a target diagnostic model. The host computer receives the digital signal and inputs the digital signal into the target diagnostic model, and outputs the analysis result, which is the fault type of the air switch cabinet.

7. The online monitoring system for early warning of discharge faults in air switchgear according to claim 6, characterized in that, Training the diagnostic model includes: The data sent by the signal transmitter module is acquired. The data is collected by the sensor module during discharge tests of different voltage levels and different discharge fault types in the air switch cabinet. The data includes temperature and humidity data, composition and content of gas decomposition products. Detect whether the humidity data is greater than a preset humidity threshold; If the humidity data is greater than the preset humidity threshold, the component content of the gas decomposition products is multiplied by the preset humidity correction factor. If the humidity data is not greater than the preset humidity threshold, the training data is normalized, or the corrected component content and temperature and humidity data are normalized to obtain training samples. Using the training samples as input and the fault type as output, the initial diagnostic model is trained to obtain the trained diagnostic model.

8. The online monitoring system for early warning of discharge faults in air switchgear according to claim 7, characterized in that, The formula for calculating the activation function in the pattern layer of the initial diagnostic model is as follows: ； in, This represents the activation function value for each sample. This represents the input for training. Indicates training samples, Represents the smoothing coefficient. This indicates the sample number of the training sample.

9. The online monitoring system for early warning of discharge faults in air switchgear according to claim 8, characterized in that, The methods for determining include: During the discharge test, determine the corresponding discharge fault types. And set the smoothing coefficient-fault type table; Determine the corresponding fault type in the smoothing coefficient-fault type table. .

10. The online monitoring system for early warning of discharge faults in air switchgear according to claim 9, characterized in that, The smoothing coefficient in the smoothing coefficient-fault type table is a range of smoothing coefficient values. During the training of the initial diagnostic model, the diagnostic error of the validation samples is calculated in real time. If the diagnostic error exceeds the error threshold, the gradient descent method is used to dynamically adjust the value of the smoothing coefficient.

Images