Method for measuring gas concentration in transformer oil

Abstract

The invention discloses a method for measuring the concentration of gas in transformer oil, and belongs to the technical field of transformer oil, and the method comprises the following steps: S1, collecting and processing the spectral data of the gas in the transformer oil, and determining the spectral characteristic data of the gas in the transformer oil; s2, training and optimizing the model for measuring and predicting the gas concentration in the transformer oil, and determining an optimal model for measuring and predicting the gas concentration in the transformer oil; and S3, analyzing the spectral characteristic data of the gas in the transformer oil, predicting the type and content of the gas in the transformer oil, and determining a measurement result of the concentration of the gas in the transformer oil. The problems that in the prior art, gas in transformer oil cannot be monitored in real time, the concentration cannot be measured, and transformer faults cannot be found in time are solved. According to the invention, real-time monitoring and concentration measurement can be carried out on the gas in the transformer oil, the internal state of the transformer can be effectively judged, the transformer fault can be found in time, and the use effect of the transformer can be improved.

Description

Technical Field

[0001] The invention relates to the technical field of transformer oil, in particular to a method for measuring gas concentration in transformer oil. Background Art

[0002] The transformer is one of the most important equipment in the power system, and its normal operation is crucial to ensure the stability and reliability of the power system. However, due to long-term use and aging, the insulation material inside the transformer will gradually damage, resulting in the generation of dissolved gas in the transformer oil. If these gases cannot be discharged in time, they will lead to further deterioration of the insulation material and equipment failure. Therefore, it is very important to regularly detect the gas concentration in the transformer oil.

[0003] The existing technology cannot monitor and measure the concentration of gas in transformer oil in real time, cannot effectively judge the internal state of the transformer, and cannot detect transformer faults in time, resulting in poor transformer use effect. Summary of the invention

[0004] The purpose of the present invention is to provide a method for measuring the gas concentration in transformer oil, which can monitor and measure the concentration of gas in transformer oil in real time, effectively judge the internal state of the transformer, and timely discover transformer faults, thereby improving the use effect of the transformer and solving the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions: A method for measuring gas concentration in transformer oil comprises the following steps: S1: Collect and process the spectrum data of gas in transformer oil to determine the spectrum characteristic data of gas in transformer oil, including wavelength range; S2: Train and optimize the prediction model for measuring gas concentration in transformer oil, use the prediction model for measuring gas concentration in transformer oil to analyze the spectral characteristic data of gas in transformer oil, predict the type and content of gas in transformer oil, determine the measurement result of gas concentration in transformer oil, and control the transformer; The process of determining the wavelength range includes: Extract the gas types included in the gas concentration detection requirements as target gases; Extract the temperature range of the transformer oil corresponding to the gas concentration detection required during the operation of the transformer, and obtain the vibration frequency data of the gas molecules corresponding to each target gas within the temperature range according to the temperature range of the transformer oil; the vibration frequency data includes the vibration frequency, vibration line width and vibration intensity of the gas molecules; Marking the temperature points of the temperature range corresponding to the transformer oil according to a preset unit temperature interval, and obtaining a plurality of corresponding temperature points within the temperature range of the transformer oil; wherein the value range of the preset unit temperature interval is 1 degree Celsius to 3 degrees Celsius; Extract the vibration frequency, vibration line width and vibration intensity of gas molecules corresponding to each target gas at each temperature point; The vibration evaluation coefficient corresponding to each target gas at each temperature point is obtained by using the vibration frequency, vibration line width and vibration intensity of the gas molecules corresponding to each target gas at each temperature point; The wavelength range is determined using the vibration evaluation coefficient corresponding to each temperature point corresponding to each target gas.

[0006] Preferably, in S1, collecting gas spectrum data in transformer oil includes: Prepare a transformer oil sample, collect transformer oil from the transformer, add the collected transformer oil to the spectrometer, and adjust the light source parameters of the spectrometer to put it in a working state; The transformer oil is heated so that the volatile gas in the transformer oil is gradually converted into thermal radiation to form a thermal light emission signal. The volatile gas in the transformer oil emits a spectral signal of a specific wavelength when heated. The spectral signal is collected to determine the gas spectral data in the transformer oil.

[0007] Preferably, in S1, processing the gas spectrum data in the transformer oil includes: Acquire gas spectrum data in transformer oil, and clean the gas spectrum data in transformer oil; Among them, based on the data processing tool, the gas spectrum data in the transformer oil is checked to determine whether the gas spectrum data in the transformer oil contains repeated values, missing values ​​and abnormal values, and the repeated values, missing values ​​and abnormal values ​​contained in the gas spectrum data in the transformer oil are processed to eliminate the noise in the gas spectrum data in the transformer oil; Acquire gas spectrum data in the cleaned transformer oil, and perform normalization processing on the gas spectrum data in the cleaned transformer oil; Among them, each spectral point in the gas spectral data in transformer oil is normalized based on the least square method, so that the color depth of all spectral points in the gas spectral data in transformer oil is the same, and the standardized gas spectral data in transformer oil is determined; Obtain standardized gas spectrum data in transformer oil, and perform feature extraction on the standardized gas spectrum data in transformer oil based on the correlation coefficient method; Among them, the features useful for measuring the gas concentration in transformer oil are extracted from the gas spectrum data in transformer oil, and the characteristic vector of the gas spectrum in transformer oil is determined, including wavelength range, intensity ratio, shape, relative intensity and peak shift; The extracted spectral feature vectors of gas in transformer oil are evaluated, and inappropriate spectral feature vectors of gas in transformer oil are screened based on information entropy to determine the optimal spectral feature data of gas in transformer oil.

[0008] Preferably, the vibration evaluation coefficient corresponding to each temperature point is obtained by the following formula: ; Where S represents the comprehensive vibration evaluation coefficient; w 01 、w 02 and w 03 Respectively represent the weight values ​​corresponding to the preset gas molecule vibration frequency, vibration line width and gas molecule vibration intensity; F t , D t and I t Indicates the vibration frequency, vibration line width and vibration intensity of gas molecules corresponding to each temperature point; D k and I k Indicates the vibration spectrum width span value and the vibration intensity span value of the gas molecules corresponding to the lower and upper temperature values ​​of the temperature range; F c and I c Indicates the preset vibration frequency reference value and vibration intensity test value; D s Indicates the vibration line width corresponding to the lower limit of the temperature range; T k Indicates the temperature span value corresponding to the temperature range; T indicates the temperature value corresponding to the temperature point; T s Indicates the temperature value corresponding to the lower limit of the temperature range.

[0009] Preferably, the vibration evaluation coefficient corresponding to each temperature point is used to obtain the vibration comprehensive evaluation coefficient corresponding to each target gas, including: Retrieve the vibration evaluation coefficient corresponding to each target gas at each temperature point; Obtain a comprehensive vibration evaluation coefficient corresponding to each target gas by using the vibration evaluation coefficient corresponding to each target gas at each temperature point; The comprehensive vibration evaluation coefficient is obtained by the following formula: ;

[0010] Among them, P s represents the comprehensive vibration evaluation coefficient; n represents the total number of temperature points corresponding to each target gas; S irepresents the vibration evaluation coefficient corresponding to the i-th temperature point; S p Represents the average value of the vibration evaluation coefficient corresponding to n temperature points; σ s Indicates the standard deviation of the vibration evaluation coefficient; σ t Indicates the temperature standard deviation corresponding to n temperature points; T i Indicates the temperature value corresponding to the i-th temperature point; T p W represents the average temperature corresponding to n temperature points; gi Indicates the weight coefficient corresponding to the preset i-th temperature point; The wavelength range corresponding to each target gas is obtained using the vibration comprehensive evaluation coefficient corresponding to each target gas.

[0011] Preferably, the wavelength range corresponding to each target gas is obtained by using the vibration comprehensive evaluation coefficient corresponding to each target gas, including: Retrieve emission spectrum data corresponding to the target gas from the spectrum database resource; Acquire a basic wavelength range from emission spectrum data corresponding to the target gas; Comparing the vibration comprehensive evaluation coefficient corresponding to each target gas with a preset vibration comprehensive evaluation coefficient threshold; When the vibration comprehensive evaluation coefficient does not exceed the preset vibration comprehensive evaluation coefficient threshold, the basic wavelength range is used as the wavelength range corresponding to the target gas; When the vibration comprehensive evaluation coefficient exceeds a preset vibration comprehensive evaluation coefficient threshold, the vibration comprehensive evaluation coefficient is retrieved to adjust the basic wavelength range corresponding to the target gas to obtain the adjusted wavelength range; The adjusted wavelength range is obtained by the following formula: ;

[0012] Among them, λ up Indicates the upper limit of the adjusted wavelength range corresponding to each target gas; λ up0 Indicates the upper limit value of the basic wavelength range corresponding to each target gas; P y Indicates the preset vibration comprehensive evaluation coefficient threshold; S max and S min Indicates the maximum and minimum values ​​of the vibration evaluation coefficient corresponding to each target gas; S z Indicates the middle value of the vibration evaluation coefficient corresponding to each target gas; ;

[0013] Among them, λ down Indicates the lower limit of the adjusted wavelength range; λ down0 Indicates the lower limit value corresponding to the basic wavelength range.

[0014] Preferably, in S2, training a prediction model for measuring gas concentration in transformer oil comprises: According to the demand for measuring gas concentration in transformer oil, collect historical data of gas spectrum in transformer oil; The collected historical data of gas spectra in transformer oil are divided into a training set, a test set and a validation set; Based on deep learning technology, the deep learning model is trained with a training set, so that the deep learning model can autonomously learn the gas concentration measurement process in transformer oil, predict the type and content of gas in transformer oil, and determine the prediction model for gas concentration measurement in transformer oil. Based on the test set, the performance of the gas concentration measurement prediction model in transformer oil is tested. Based on the precision, recall rate and F1 value, it is evaluated whether the performance of the gas concentration measurement prediction model in transformer oil can meet the expected requirements, and the performance test results of the gas concentration measurement prediction model in transformer oil are determined.

[0015] Preferably, in S2, optimizing the prediction model for measuring gas concentration in transformer oil includes: When the performance of the gas concentration measurement prediction model in transformer oil does not meet the expected requirements, Based on the validation set, the parameters of the transformer oil gas concentration measurement prediction model are adjusted, and the structure of the transformer oil gas concentration measurement prediction model is modified and adjusted. After continuous optimization of the transformer oil gas concentration measurement prediction model, the performance of the transformer oil gas concentration measurement prediction model can meet the expected requirements, and the optimal transformer oil gas concentration measurement prediction model is determined.

[0016] Preferably, in S3, predicting the type and content of gas in transformer oil includes: Obtain the optimal transformer oil gas concentration measurement prediction model, and deploy the optimal transformer oil gas concentration measurement prediction model in the actual transformer oil gas concentration measurement environment; The spectral characteristic data of gas in transformer oil are input into the optimal prediction model for measuring gas concentration in transformer oil. Based on the optimal prediction model for measuring gas concentration in transformer oil, the spectral characteristic data of gas in transformer oil are analyzed, and the type and content of gas in transformer oil are predicted to determine the measurement result of gas concentration in transformer oil.

[0017] Preferably, in S3, controlling the transformer includes: According to the measurement results of gas concentration in transformer oil, the internal state of the transformer is judged, and the transformer control method is formulated in time. The transformer is controlled based on the formulated transformer control method, providing a basis for timely detection of transformer faults and prevention of transformer accidents; Among them, the type of internal fault of the transformer is determined according to the type of gas in the transformer oil measured; When the gases measured in the transformer oil are hydrogen and acetylene, it indicates an arc fault in the oil; When the gases in the transformer oil are measured as hydrogen and carbon monoxide, it indicates a partial discharge fault; When the gases measured in the transformer oil are methane, ethylene and acetylene, it indicates that the oil-paper insulation is overheated; According to the type of internal fault of the transformer, timely warning is given to the transformer management personnel, so that the transformer management personnel can repair the transformer fault as soon as possible and quickly eliminate the transformer fault.

[0018] Compared with the prior art, the present invention has the following beneficial effects: The present invention collects spectral data of gas in transformer oil through a spectrometer, processes the spectral data of gas in transformer oil, determines spectral characteristic data of gas in transformer oil, trains a prediction model for measuring gas concentration in transformer oil, optimizes the prediction model for measuring gas concentration in transformer oil, determines an optimal prediction model for measuring gas concentration in transformer oil, analyzes spectral characteristic data of gas in transformer oil based on the optimal prediction model for measuring gas concentration in transformer oil, predicts the type and content of gas in transformer oil, determines the measurement result of gas concentration in transformer oil, realizes high-precision and high-sensitivity measurement of gas concentration in transformer oil, effectively judges the internal state of the transformer, and manages the transformer, provides a basis for timely discovering transformer faults and preventing transformer accidents, can monitor gas in transformer oil in real time and measure its concentration, can effectively judge the internal state of the transformer, can timely discover transformer faults, and can improve the use effect of the transformer. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 The present invention is a flow chart of the method for measuring gas concentration in transformer oil. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] In order to solve the existing problems of not being able to monitor and measure the concentration of gas in transformer oil in real time, not being able to effectively judge the internal status of the transformer, and not being able to detect transformer failure in time, resulting in poor transformer use effect, please refer to Figure 1 , this embodiment provides the following technical solutions: A method for measuring gas concentration in transformer oil comprises the following steps: S1. Data collection and processing: collecting gas spectrum data in transformer oil based on a spectrometer, and processing the gas spectrum data in transformer oil to determine the gas spectrum characteristic data in transformer oil; In this embodiment, collecting gas spectrum data in transformer oil includes: Prepare a transformer oil sample, collect transformer oil from the transformer, add the collected transformer oil to the spectrometer, and adjust the light source parameters of the spectrometer to put it in a working state; The transformer oil is heated so that the volatile gas in the transformer oil is gradually converted into thermal radiation to form a thermal light emission signal. The volatile gas in the transformer oil emits a spectral signal of a specific wavelength when heated. The spectral signal is collected to determine the gas spectral data in the transformer oil.

[0022] It should be noted that the spectrometer is a device that uses photomultiplier tubes and other light detectors to measure the intensity of spectral lines at different wavelengths. It consists of an incident slit, a dispersion system, an imaging system and an exit slit. The dispersion element is used to separate the electromagnetic radiation of the radiation source into the required wavelength or wavelength region, and the intensity is measured at the selected wavelength. By collecting the spectral signal of the gas in the transformer oil, the spectral data of the gas in the transformer oil is determined, providing data support for the measurement of the gas concentration in the transformer oil.

[0023] In this embodiment, the gas spectrum data in transformer oil is processed, including: Acquire gas spectrum data in transformer oil, and clean the gas spectrum data in transformer oil; Among them, based on the data processing tool, the gas spectrum data in the transformer oil is checked to determine whether the gas spectrum data in the transformer oil contains repeated values, missing values ​​and abnormal values, and the repeated values, missing values ​​and abnormal values ​​contained in the gas spectrum data in the transformer oil are processed to eliminate the noise in the gas spectrum data in the transformer oil; It should be noted that by cleaning the gas spectrum data in transformer oil, the noise in the gas spectrum data in transformer oil can be eliminated, and the subsequent processing accuracy and efficiency of the gas spectrum data in transformer oil can be improved.

[0024] Acquire gas spectrum data in the cleaned transformer oil, and perform normalization processing on the gas spectrum data in the cleaned transformer oil; Among them, each spectral point in the spectral data of gas in transformer oil is normalized based on the least square method, so that the color depth of all spectral points in the spectral data of gas in transformer oil is the same, and the standardized spectral data of gas in transformer oil is determined.

[0025] It should be noted that by normalizing the gas spectrum data in transformer oil, abnormal fluctuations in the gas spectrum data in transformer oil can be eliminated, thereby obtaining more stable and reliable gas spectrum data in transformer oil, and determining standardized gas spectrum data in transformer oil, which is convenient for subsequent feature extraction of the standardized gas spectrum data in transformer oil.

[0026] In this embodiment, the processing of gas spectrum data in transformer oil also includes: Obtain standardized gas spectrum data in transformer oil, and perform feature extraction on the standardized gas spectrum data in transformer oil based on the correlation coefficient method; Among them, the features useful for measuring the gas concentration in transformer oil are extracted from the gas spectrum data in transformer oil, and the characteristic vector of the gas spectrum in transformer oil is determined, including wavelength range, intensity ratio, shape, relative intensity and peak shift; The extracted spectral feature vectors of gas in transformer oil are evaluated, and inappropriate spectral feature vectors of gas in transformer oil are screened based on information entropy to determine the optimal spectral feature data of gas in transformer oil.

[0027] It should be noted that the wavelength range refers to the specific wavelength of light emitted by different gases in the process of thermal optical emission spectrum, among which the selection of wavelength range is usually related to the vibration frequency of gas molecules; the intensity ratio refers to the ratio of the intensity of absorbed light at a certain wavelength to the intensity at an adjacent wavelength, and the intensity ratio can be used to distinguish different gases; shape refers to the appearance of the spectrum, and the shape of the spectrum may vary due to the physical and chemical properties of the gas. For example, some gases may present a continuous platform or shoulder-to-shoulder absorption peaks on the spectrum; relative intensity refers to the proportional relationship between the intensity of absorbed light at a certain wavelength and its molar fraction, and relative intensity can be used to distinguish gases of different concentrations; peak shift means that when the gas concentration changes, the wavelength position of the spectral feature will also move accordingly. By observing the change in the peak position, the change in gas concentration inside the transformer can be inferred.

[0028] It should be noted that by extracting the characteristics of the gas spectrum data in transformer oil and determining the optimal spectral characteristic data of the gas in transformer oil, the type and content of the gas in transformer oil can be more accurately reflected, and high-precision measurement of the gas concentration in transformer oil can be achieved.

[0029] Specifically, the process of determining the wavelength range includes: Extract the gas types included in the gas concentration detection requirements as target gases; Extract the temperature range of transformer oil corresponding to the gas concentration detection required during transformer operation; Obtaining vibration frequency data of gas molecules corresponding to each target gas within the temperature range according to the temperature range of the transformer oil; wherein the vibration frequency data includes the vibration frequency, vibration line width and vibration intensity of the gas molecules; Marking the temperature points of the temperature range corresponding to the transformer oil according to a preset unit temperature interval, and obtaining a plurality of corresponding temperature points within the temperature range of the transformer oil; wherein the value range of the preset unit temperature interval is 1 degree Celsius to 3 degrees Celsius; Extract the vibration frequency, vibration line width and vibration intensity of gas molecules corresponding to each target gas at each temperature point; The vibration evaluation coefficient corresponding to each target gas at each temperature point is obtained by using the vibration frequency, vibration line width and vibration intensity of the gas molecules corresponding to each target gas at each temperature point; The vibration evaluation coefficient corresponding to each temperature point is obtained by the following formula: ;

[0030] Where S represents the comprehensive vibration evaluation coefficient; w 01 、w 02 and w 03 Respectively represent the weight values ​​corresponding to the preset gas molecule vibration frequency, vibration line width and gas molecule vibration intensity; F t , D t and I t Indicates the vibration frequency, vibration line width and vibration intensity of gas molecules corresponding to each temperature point; D k and I k Indicates the vibration spectrum width span value and the vibration intensity span value of the gas molecules corresponding to the lower and upper temperature values ​​of the temperature range; F c and I c Indicates the preset vibration frequency reference value and vibration intensity test value; D s Indicates the vibration line width corresponding to the lower limit of the temperature range; T k Indicates the temperature span value corresponding to the temperature range; T indicates the temperature value corresponding to the temperature point; T s Indicates the temperature value corresponding to the lower limit of the temperature range; The wavelength range is determined using the vibration evaluation coefficient corresponding to each temperature point corresponding to each target gas.

[0031] The technical effect of the above technical solution is: by extracting the vibration frequency data of the target gas in the temperature range of the transformer oil in detail (including the vibration frequency of the gas molecules, the vibration line width and the vibration intensity of the gas molecules), and combining the specific information of each temperature point, the vibration characteristics of the gas molecules can be more accurately evaluated. The vibration evaluation coefficient calculated using these detailed vibration data can more accurately reflect the actual state of the target gas at different temperatures, thereby improving the accuracy of gas concentration detection. The technical solution takes into account the temperature range changes of the transformer oil during operation, and marks the temperature range according to the preset unit temperature interval, so that the gas concentration under different temperature conditions can be detected. In this way, the technical solution can adapt to the gas concentration detection needs of different transformers and different operating conditions, and enhance the flexibility and applicability of the detection. The calculation of the vibration evaluation coefficient comprehensively considers multiple factors such as the vibration frequency of the gas molecules, the vibration line width and the vibration intensity of the gas molecules, and performs weighted processing through the preset weight value, so that the wavelength range finally determined is more scientific and reasonable. Using these vibration evaluation coefficients to determine the wavelength range can ensure that the selected wavelength range can accurately reflect the vibration characteristics of the target gas, thereby improving the sensitivity and reliability of gas concentration detection. This technical solution automatically calculates the vibration evaluation coefficient through mathematical formulas and algorithms, and determines the wavelength range accordingly, avoiding the tedious manual selection and adjustment process in traditional methods. This not only improves the detection efficiency, but also reduces the impact of human factors on the detection results, and improves the automation and intelligence level of gas concentration detection. This technical solution proposes a new method for determining the wavelength range based on the vibration evaluation coefficient, which provides new ideas for technological innovation in the field of gas concentration detection. At the same time, the application of this method is not limited to the detection of gas concentration in transformer oil, but can also be extended to other fields that require accurate detection of gas concentration, such as environmental monitoring, industrial production process control, etc.

[0032] In summary, the above technical scheme comprehensively considers the vibration characteristics of the target gas and the changes in the transformer oil temperature range, and uses detailed vibration data and scientific calculation methods to determine the wavelength range, thereby improving the accuracy, adaptability and efficiency of gas concentration detection, and providing strong support for technological innovation and application expansion in related fields.

[0033] Specifically, the vibration evaluation coefficient corresponding to each temperature point is used to obtain the vibration comprehensive evaluation coefficient corresponding to each target gas, including: Retrieve the vibration evaluation coefficient corresponding to each target gas at each temperature point; Obtain a comprehensive vibration evaluation coefficient corresponding to each target gas by using the vibration evaluation coefficient corresponding to each target gas at each temperature point; The comprehensive vibration evaluation coefficient is obtained by the following formula: ;

[0034] Among them, P s represents the comprehensive vibration evaluation coefficient; n represents the total number of temperature points corresponding to each target gas; S i represents the vibration evaluation coefficient corresponding to the i-th temperature point; S p Represents the average value of the vibration evaluation coefficient corresponding to n temperature points; σ s Indicates the standard deviation of the vibration evaluation coefficient; σ t Indicates the temperature standard deviation corresponding to n temperature points; T i Indicates the temperature value corresponding to the i-th temperature point; T p W represents the average temperature corresponding to n temperature points; gi Indicates the weight coefficient corresponding to the preset i-th temperature point; The wavelength range corresponding to each target gas is obtained using the vibration comprehensive evaluation coefficient corresponding to each target gas.

[0035] The technical effect of the above technical solution is: by calling the vibration evaluation coefficient corresponding to each target gas at each temperature point, and using these coefficients to calculate the comprehensive vibration evaluation coefficient, the technical solution can comprehensively and accurately evaluate the vibration characteristics of the target gas under different temperature conditions. The comprehensive vibration evaluation coefficient Ps takes into account multiple factors, including the average value, standard deviation, temperature standard deviation, temperature average value and weight coefficient of each temperature point of the vibration evaluation coefficient, so as to more comprehensively reflect the vibration state of the target gas. The process of determining the wavelength range corresponding to the target gas by using the calculated vibration comprehensive evaluation coefficient is more scientific and accurate. By comprehensively considering the vibration characteristics and temperature conditions of the target gas, the technical solution can select the wavelength range most suitable for gas concentration detection, thereby improving the sensitivity and accuracy of detection. The technical solution automatically calculates the comprehensive vibration evaluation coefficient and wavelength range through mathematical formulas and algorithms, avoiding the tedious manual selection and adjustment process. This not only improves the detection efficiency, but also reduces the influence of human factors on the detection results, and improves the automation level and intelligence level of gas concentration detection. Since the comprehensive vibration evaluation coefficient takes into account multiple factors and the weight coefficient can be adjusted according to actual conditions, the technical solution can more accurately reflect the actual vibration state of the target gas. It helps to improve the reliability and stability of the detection results and reduce the possibility of false alarms and missed alarms. This technical solution proposes a new method for determining the wavelength range based on the comprehensive vibration evaluation coefficient, which provides new ideas for technological innovation in the field of gas concentration detection. At the same time, the application of this method is not limited to gas concentration detection in specific scenarios, but can also be expanded to other fields that require accurate evaluation of gas vibration characteristics, such as environmental monitoring, industrial production process control, etc.

[0036] In summary, the above technical solution improves the accuracy, efficiency and reliability of gas concentration detection by comprehensively considering the vibration characteristics and temperature conditions of the target gas and calculating the comprehensive vibration evaluation coefficient and wavelength range using mathematical formulas and algorithms. At the same time, this technical solution also has broad application prospects and expansion potential, and can provide strong support for technological innovation and application in related fields.

[0037] Specifically, the wavelength range corresponding to each target gas is obtained by using the vibration comprehensive evaluation coefficient corresponding to each target gas, including: Retrieve emission spectrum data corresponding to the target gas from the spectrum database resource; Acquire a basic wavelength range from emission spectrum data corresponding to the target gas; Comparing the vibration comprehensive evaluation coefficient corresponding to each target gas with a preset vibration comprehensive evaluation coefficient threshold; When the vibration comprehensive evaluation coefficient does not exceed the preset vibration comprehensive evaluation coefficient threshold, the basic wavelength range is used as the wavelength range corresponding to the target gas; When the vibration comprehensive evaluation coefficient exceeds a preset vibration comprehensive evaluation coefficient threshold, the vibration comprehensive evaluation coefficient is called to adjust the basic wavelength range corresponding to the target gas to obtain the adjusted wavelength range; The adjusted wavelength range is obtained by the following formula: ;

[0038] Among them, λ up Indicates the upper limit of the adjusted wavelength range corresponding to each target gas; λ up0 Indicates the upper limit value of the basic wavelength range corresponding to each target gas; P y Indicates the preset vibration comprehensive evaluation coefficient threshold; S max and S min Indicates the maximum and minimum values ​​of the vibration evaluation coefficient corresponding to each target gas; S z Indicates the middle value of the vibration evaluation coefficient corresponding to each target gas; ;

[0039] Among them, λ down Indicates the lower limit of the adjusted wavelength range; λ down0 Indicates the lower limit value corresponding to the basic wavelength range.

[0040] The technical effect of the above technical solution is: the technical solution can more accurately determine the wavelength range of the target gas by combining the vibration comprehensive evaluation coefficient of the target gas with the emission spectrum data in the spectrum database. When the vibration comprehensive evaluation coefficient does not exceed the preset threshold, the basic wavelength range is directly used to ensure the accuracy of the wavelength range; when it exceeds the threshold, the basic wavelength range is adjusted according to the vibration comprehensive evaluation coefficient, further improving the accuracy of the wavelength range. The technical solution realizes flexible control of the determination of the wavelength range by introducing the vibration comprehensive evaluation coefficient threshold. According to different vibration comprehensive evaluation coefficients, the wavelength range can be automatically adjusted to meet the needs of different target gases and different detection conditions. The technical solution automatically calculates the adjusted wavelength range through mathematical formulas and algorithms, avoiding the cumbersome manual selection and adjustment process. This not only improves the detection efficiency, but also reduces the influence of human factors on the detection results, and improves the automation level and intelligence of gas concentration detection. By comprehensively considering the vibration characteristics and emission spectrum data of the target gas, the technical solution can more accurately reflect the actual state of the target gas. This helps to improve the reliability and stability of the detection results, reduce the possibility of false alarms and missed alarms, and improve the accuracy of gas concentration detection. This technical solution proposes a new method for determining the wavelength range based on the comprehensive vibration evaluation coefficient, which provides new ideas for technological innovation in the field of gas concentration detection. At the same time, the application of this method is not limited to gas concentration detection in specific scenarios, but can also be extended to other fields that require accurate determination of the wavelength range, such as environmental monitoring, industrial production process control, food safety testing, etc. By automatically adjusting the wavelength range, unnecessary waste of detection equipment and resources is avoided. This helps to reduce detection costs and improve economic benefits.

[0041] In summary, the above technical solution combines the vibration comprehensive evaluation coefficient of the target gas with the emission spectrum data in the spectrum database to achieve accurate and flexible determination of the wavelength range, improve detection efficiency and automation level, optimize the reliability and stability of the detection results, and promote technological innovation and application expansion. At the same time, the technical solution also helps save resources and costs, and has broad application prospects and economic benefits.

[0042] S2. Model training and optimization: training a prediction model for measuring gas concentration in transformer oil, optimizing the prediction model for measuring gas concentration in transformer oil, and determining the optimal prediction model for measuring gas concentration in transformer oil; In this embodiment, training a gas concentration measurement prediction model in transformer oil includes: According to the demand for measuring gas concentration in transformer oil, collect historical data of gas spectrum in transformer oil; The collected historical data of gas spectra in transformer oil are divided into a training set, a test set and a validation set; Based on deep learning technology, the deep learning model is trained with a training set, so that the deep learning model can autonomously learn the gas concentration measurement process in transformer oil, predict the type and content of gas in transformer oil, and determine the prediction model for gas concentration measurement in transformer oil. Based on the test set, the performance test of the gas concentration measurement prediction model in transformer oil is carried out. Based on the precision, recall rate and F1 value, the performance of the gas concentration measurement prediction model in transformer oil is evaluated to see whether it can meet the expected requirements, and the performance test results of the gas concentration measurement prediction model in transformer oil are determined; When the performance of the gas concentration measurement prediction model in transformer oil cannot meet the expected requirements, the parameters of the gas concentration measurement prediction model in transformer oil are adjusted based on the validation set, and the structure of the gas concentration measurement prediction model in transformer oil is modified and adjusted. After continuous optimization of the gas concentration measurement prediction model in transformer oil, the performance of the gas concentration measurement prediction model in transformer oil can meet the expected requirements, and the optimal gas concentration measurement prediction model in transformer oil is determined.

[0043] It should be noted that by training the gas concentration measurement prediction model in transformer oil and optimizing the gas concentration measurement prediction model in transformer oil, the optimal gas concentration measurement prediction model in transformer oil is determined, which is convenient for subsequent prediction of the type and content of gas in transformer oil, and then measuring the gas concentration in transformer oil.

[0044] S3. Concentration measurement and control: Based on the optimal transformer oil gas concentration measurement prediction model, the spectral characteristic data of the gas in the transformer oil is analyzed, and the type and content of the gas in the transformer oil are predicted, the measurement results of the gas concentration in the transformer oil are determined, and the transformer is controlled to provide a basis for timely detection of transformer faults and prevention of transformer accidents.

[0045] In this embodiment, the type and content of gas in transformer oil are predicted, including: Obtain the optimal transformer oil gas concentration measurement prediction model, and deploy the optimal transformer oil gas concentration measurement prediction model in the actual transformer oil gas concentration measurement environment; The spectral characteristic data of gas in transformer oil are input into the optimal prediction model for measuring gas concentration in transformer oil. Based on the optimal prediction model for measuring gas concentration in transformer oil, the spectral characteristic data of gas in transformer oil are analyzed, and the type and content of gas in transformer oil are predicted to determine the measurement result of gas concentration in transformer oil.

[0046] In this embodiment, the transformer is controlled, including: According to the measurement results of gas concentration in transformer oil, the internal state of the transformer is judged, and the transformer control method is formulated in time. The transformer is controlled based on the formulated transformer control method, providing a basis for timely detection of transformer faults and prevention of transformer accidents; Among them, the type of internal fault of the transformer is determined according to the type of gas in the transformer oil measured; When the gases measured in the transformer oil are hydrogen and acetylene, it indicates an arc fault in the oil; When the gases in the transformer oil are measured as hydrogen and carbon monoxide, it indicates a partial discharge fault; When the gases measured in the transformer oil are methane, ethylene and acetylene, it indicates that the oil-paper insulation is overheated; According to the type of internal fault of the transformer, timely warning is given to the transformer management personnel, so that the transformer management personnel can repair the transformer fault as soon as possible and quickly eliminate the transformer fault.

[0047] In summary, based on the optimal prediction model for measuring gas concentration in transformer oil, the spectral characteristic data of gas in transformer oil are analyzed, and the type and content of gas in transformer oil are predicted, the measurement result of gas concentration in transformer oil is determined, and high-precision and high-sensitivity measurement of gas concentration in transformer oil is achieved. By effectively judging the internal state of the transformer and controlling the transformer, a basis is provided for timely detection of transformer faults and prevention of transformer accidents. The gas in transformer oil can be monitored and the concentration measured in real time, which can effectively judge the internal state of the transformer, detect transformer faults in time, and improve the use effect of the transformer.

[0048] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0049] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for measuring gas concentration in transformer oil, characterized in that: include: The gas spectrum data in the transformer oil is collected and processed to determine the gas spectrum characteristic data in the transformer oil, including the wavelength range, wherein the wavelength range is determined by: determining the wavelength range based on the corresponding vibration evaluation coefficient of the target gas at the corresponding temperature point; Train and optimize the prediction model for measuring gas concentration in transformer oil, use the prediction model to analyze the spectral characteristic data of gas in transformer oil, predict the type and content of gas in transformer oil, determine the measurement results of gas concentration in transformer oil, and control the transformer.

2. The method for measuring gas concentration in transformer oil according to claim 1, characterized in that: Collect gas spectrum data in transformer oil, including: Prepare a transformer oil sample, collect transformer oil from the transformer, add the collected transformer oil into the spectrometer, adjust the light source parameters of the spectrometer, and put the spectrometer into a working state; The transformer oil is heated so that the volatile gas in the transformer oil is gradually converted into thermal radiation to form a thermal light emission signal. The volatile gas in the transformer oil emits a spectral signal of a specific wavelength when heated. The spectral signal is collected to determine the gas spectral data in the transformer oil.

3. A method for measuring gas concentration in transformer oil as claimed in claim 2, characterized in that: Processing of gas spectrum data in transformer oil, including: Acquire gas spectrum data in transformer oil, and clean the gas spectrum data in transformer oil; Acquire gas spectrum data in the cleaned transformer oil, and perform normalization processing on the gas spectrum data in the cleaned transformer oil; Obtain standardized gas spectrum data in transformer oil, and perform feature extraction on the standardized gas spectrum data in transformer oil based on the correlation coefficient method; The extracted spectral feature vectors of gas in transformer oil are evaluated, and inappropriate spectral feature vectors of gas in transformer oil are screened based on information entropy to determine the optimal spectral feature data of gas in transformer oil.

4. The method for measuring gas concentration in transformer oil according to claim 3, characterized in that: The process of determining the wavelength range includes: Extract the gas types included in the gas concentration detection requirements as target gases; Extract the temperature range of the transformer oil corresponding to the gas concentration detection required during the operation of the transformer, and obtain the vibration frequency data of the gas molecules corresponding to each target gas within the temperature range according to the temperature range of the transformer oil; the vibration frequency data includes the vibration frequency, vibration line width and vibration intensity of the gas molecules; Marking the temperature points of the temperature range corresponding to the transformer oil according to a preset unit temperature interval, and obtaining a plurality of corresponding temperature points within the temperature range of the transformer oil; wherein the value range of the preset unit temperature interval is 1 degree Celsius to 3 degrees Celsius; Extract the vibration frequency, vibration line width and vibration intensity of gas molecules corresponding to each target gas at each temperature point; The vibration evaluation coefficient corresponding to each target gas at each temperature point is obtained by using the vibration frequency, vibration line width and vibration intensity of the gas molecules corresponding to each target gas at each temperature point; The wavelength range is determined using the vibration evaluation coefficient corresponding to each temperature point corresponding to each target gas.

5. A method for measuring gas concentration in transformer oil as claimed in claim 4, characterized in that: The vibration evaluation coefficient corresponding to each temperature point is used to obtain the vibration comprehensive evaluation coefficient corresponding to each target gas, including: Retrieve the vibration evaluation coefficient corresponding to each target gas at each temperature point; Obtain a comprehensive vibration evaluation coefficient corresponding to each target gas by using the vibration evaluation coefficient corresponding to each target gas at each temperature point; The comprehensive vibration evaluation coefficient is obtained by the following formula: ； Among them, P s represents the comprehensive vibration evaluation coefficient; n represents the total number of temperature points corresponding to each target gas; S i represents the vibration evaluation coefficient corresponding to the i-th temperature point; S p Represents the average value of the vibration evaluation coefficient corresponding to n temperature points; σ s Indicates the standard deviation of the vibration evaluation coefficient; σ t Indicates the temperature standard deviation corresponding to n temperature points; T i Indicates the temperature value corresponding to the i-th temperature point; T p W represents the average temperature corresponding to n temperature points; gi Indicates the weight coefficient corresponding to the preset i-th temperature point; The wavelength range corresponding to each target gas is obtained using the vibration comprehensive evaluation coefficient corresponding to each target gas.

6. A method for measuring gas concentration in transformer oil as claimed in claim 5, characterized in that: The wavelength range corresponding to each target gas is obtained by using the vibration comprehensive evaluation coefficient corresponding to each target gas, including: Retrieve emission spectrum data corresponding to the target gas from the spectrum database resource; Acquire a basic wavelength range from emission spectrum data corresponding to the target gas; Comparing the vibration comprehensive evaluation coefficient corresponding to each target gas with a preset vibration comprehensive evaluation coefficient threshold; When the vibration comprehensive evaluation coefficient does not exceed the preset vibration comprehensive evaluation coefficient threshold, the basic wavelength range is used as the wavelength range corresponding to the target gas; When the vibration comprehensive evaluation coefficient exceeds a preset vibration comprehensive evaluation coefficient threshold, the vibration comprehensive evaluation coefficient is retrieved to adjust the basic wavelength range corresponding to the target gas to obtain the adjusted wavelength range.

7. A method for measuring gas concentration in transformer oil as claimed in claim 6, characterized in that: Training of prediction model for gas concentration measurement in transformer oil, including: According to the demand for measuring gas concentration in transformer oil, collect historical data of gas spectrum in transformer oil; The collected historical data of gas spectra in transformer oil are divided into a training set, a test set and a validation set; Based on deep learning technology, the deep learning model is trained with a training set, so that the deep learning model can autonomously learn the gas concentration measurement process in transformer oil, predict the type and content of gas in transformer oil, and determine the prediction model for gas concentration measurement in transformer oil. Based on the test set, the performance of the gas concentration measurement prediction model in transformer oil is tested. Based on the precision, recall rate and F1 value, it is evaluated whether the performance of the gas concentration measurement prediction model in transformer oil can meet the expected requirements, and the performance test results of the gas concentration measurement prediction model in transformer oil are determined.

8. A method for measuring gas concentration in transformer oil as claimed in claim 7, characterized in that: Optimize the prediction model for gas concentration measurement in transformer oil, including: When the performance of the gas concentration measurement prediction model in transformer oil does not meet the expected requirements, Based on the validation set, the parameters of the transformer oil gas concentration measurement prediction model are adjusted, and the structure of the transformer oil gas concentration measurement prediction model is modified and adjusted. After continuous optimization of the transformer oil gas concentration measurement prediction model, the performance of the transformer oil gas concentration measurement prediction model can meet the expected requirements, and the optimal transformer oil gas concentration measurement prediction model is determined.

9. A method for measuring gas concentration in transformer oil as claimed in claim 8, characterized in that: Predict the type and content of gas in transformer oil, including: Obtain the optimal transformer oil gas concentration measurement prediction model, and deploy the optimal transformer oil gas concentration measurement prediction model in the actual transformer oil gas concentration measurement environment; The spectral characteristic data of gas in transformer oil are input into the optimal prediction model for measuring gas concentration in transformer oil. Based on the optimal prediction model for measuring gas concentration in transformer oil, the spectral characteristic data of gas in transformer oil are analyzed, and the type and content of gas in transformer oil are predicted to determine the measurement result of gas concentration in transformer oil.

10. A method for measuring gas concentration in transformer oil as claimed in claim 9, characterized in that: Control and manage transformers, including: Determine the internal state of the transformer based on the measurement results of gas concentration in transformer oil; Among them, the type of internal fault of the transformer is determined according to the type of gas in the transformer oil measured; When the gases measured in the transformer oil are hydrogen and acetylene, it indicates an arc fault in the oil; When the gases in the transformer oil are measured as hydrogen and carbon monoxide, it indicates a partial discharge fault; When the gases measured in the transformer oil are methane, ethylene and acetylene, it indicates that the oil-paper insulation is overheating.

Images