Method and system for identifying mixed gas in transformer oil, product and medium

Abstract

The invention discloses a method and a system for identifying mixed gas in transformer oil, a product and a medium, and relates to the technical field of testing or analyzing materials by means of measuring chemical or physical properties of the materials, and the method comprises the following steps: acquiring an internal structure diagram of a transformer, determining spatial position relations of parts and corresponding temperature distribution characteristics, and obtaining solubility differences of different spatial positions according to the spatial position relations and the corresponding temperature distribution characteristics. During operation, mixed gas concentration data of a plurality of detection points are collected, initial gas concentration data are recorded when an operation threshold range is deviated, and gas concentration change data are corrected according to solubility differences of different temperature zones. Calculating gas concentration space-time distribution characteristics, and determining a gas accumulation area according to the gas concentration space-time distribution characteristics and a space position relation. And separating the mixed gas components by using a gas chromatographic method, comparing preset gas types with the content of each component, and determining the corresponding gas type of each internal region. By implementing the method, the accuracy of identifying the types of the mixed gas in the transformer oil can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of testing or analyzing materials by measuring the chemical or physical properties of the materials, and in particular to a method, system, product and medium for identifying mixed gases in transformer oil. Background Art

[0002] As power grids continue to expand and electricity loads continue to grow, transformers, as key components in power systems, are crucial for ensuring safe and stable operation. Internal transformer faults often cause the insulating oil to decompose, producing various characteristic gases. The types and concentrations of these gases can reveal the transformer's operating status and potential fault types.

[0003] Currently, the identification of mixed gases in transformer oil is primarily accomplished by setting up sampling points on the top of the transformer tank and regularly collecting oil samples for offline analysis. After sampling, the mixed gases are separated using methods such as gas chromatography. The fault type is determined based on the proportions of the various gases and empirical rules.

[0004] During actual operation, transformers exhibit significant internal temperature variations, with oil temperatures in different areas varying by tens of degrees Celsius. This temperature variation significantly alters the solubility of gases in the oil, making it difficult for single-point sampling data to accurately reflect the actual gas distribution within the entire transformer, and thus complicating gas identification. Summary of the Invention

[0005] The present application provides a method, system, product and medium for identifying mixed gases in transformer oil, which are used to improve the accuracy of identifying the types of mixed gases in transformer oil.

[0006] In the first aspect, the present application provides a method for identifying mixed gases in transformer oil, which is applied to a mixed gas identification system. The method includes: obtaining an internal structure diagram of the transformer, determining the spatial position relationship of each internal component of the transformer and the temperature distribution characteristics corresponding to the spatial position relationship, and obtaining the solubility difference of the transformer at different spatial positions based on the temperature distribution characteristics; continuously collecting mixed gas concentration data of multiple detection points during the operation of the transformer, and when the mixed gas concentration data of multiple detection points deviate from the operating threshold range, recording the initial gas concentration data of the multiple detection points, and correcting the gas concentration change data according to the solubility difference in different temperature zones to obtain gas concentration change data; calculating the spatiotemporal distribution characteristics of the gas concentration based on the gas concentration change data, and determining the gas accumulation area of ​​the gas inside the transformer according to the spatiotemporal distribution characteristics of the gas concentration and the spatial position relationship; using gas chromatography to separate the mixed gas components to obtain the content of each component, and comparing the preset gas type in the gas accumulation area with the content of each component to determine the gas type corresponding to each area inside the transformer.

[0007] In the above example, a diagram of the transformer's internal structure is obtained, its temperature distribution characteristics are determined, and solubility differences at different spatial locations are calculated, enabling temperature correction of gas concentration data. Based on this corrected data, the temporal and spatial distribution characteristics are calculated to accurately locate gas accumulation areas. This is then matched with the preset gas types of the corresponding components in these areas for analysis. Ultimately, this allows for precise identification of mixed gases in transformer oil, effectively improving the accuracy of gas type identification.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the step of calculating the spatiotemporal distribution characteristics of gas concentration based on gas concentration change data, and determining the gas accumulation area inside the transformer according to the spatiotemporal distribution characteristics of gas concentration and the spatial position relationship, specifically includes: establishing a gas concentration distribution model including time and space dimensions according to the gas concentration change data, the gas concentration distribution model including concentration data of multiple detection points at different times and the spatial coordinate information of the detection points; calculating the gas concentration gradient and gas diffusion direction according to the gas concentration distribution model, and obtaining the spatiotemporal distribution characteristics of gas concentration according to the gas concentration gradient and gas diffusion direction; matching the spatiotemporal distribution characteristics of gas concentration with the spatial position relationship of each component inside the transformer, and determining the area with the largest gas concentration change as the gas accumulation area.

[0009] In the above embodiment, a gas concentration distribution model encompassing both temporal and spatial dimensions was established. Concentration data from detection points was combined with spatial coordinate information to calculate the gas concentration gradient and diffusion direction. By matching the temporal and spatial distribution characteristics with the transformer's internal structure, the areas with the greatest gas concentration variation were located. This provided accurate spatial positioning for subsequent gas type identification and improved the accuracy of determining gas accumulation areas.

[0010] In combination with some embodiments of the first aspect, in some embodiments, the steps of using gas chromatography to separate the mixed gas components to obtain the content of each component, and comparing the preset gas types in the gas accumulation area with the content of each component to determine the gas type corresponding to each area inside the transformer specifically include: according to the distribution position of the gas accumulation area, adopting an adaptive interval sampling method to obtain a mixed gas sample, separating the mixed gas sample by gas chromatography to obtain the response curve and component content of each gas component, and establishing the relationship between the change of gas component content and time; extracting the transformer component type where the gas accumulation area is located from the internal structure diagram of the transformer, and obtaining its corresponding preset gas type data set based on the component type, the preset gas type data set contains the typical content ratio relationship of various gas components under different fault types; performing pattern matching on the change relationship of the gas component content and the typical content ratio relationship in the preset gas type data set, calculating the similarity score based on the matching result, and selecting the preset gas type with the highest similarity score as the gas type corresponding to each area inside the transformer.

[0011] In the above embodiment, an adaptive sampling method was used to obtain mixed gas samples and establish a relationship between gas component content and time. The component types of the gas accumulation areas were extracted from the transformer's internal structure diagram, and a preset gas type dataset was obtained. Pattern matching was performed by calculating similarity scores, achieving accurate gas type identification and reducing the false positive rate during mixed gas identification.

[0012] In combination with some embodiments of the first aspect, in some embodiments, before the steps of obtaining the internal structure diagram of the transformer, determining the spatial position relationship of the internal components of the transformer and the temperature distribution characteristics corresponding to the spatial position relationship, and obtaining the solubility difference of the transformer at different spatial positions based on the temperature distribution characteristics, the method also includes: obtaining historical data on the mixed gas concentration of multiple detection points in the transformer oil tank, calculating the time difference when the maximum gas concentration is detected at each detection point; and obtaining the operating threshold range during normal operation based on the ratio of the maximum concentration values ​​of each detection point.

[0013] In the above embodiment, historical data on mixed gas concentrations at multiple detection points is collected, the time differences between the maximum gas concentrations at each detection point are calculated, and a threshold range for normal operation is established based on the ratio of the maximum concentration values ​​at each detection point. This operational benchmark, established based on the time differences and concentration ratios, forms a dynamic monitoring standard for gas concentration changes, improving the sensitivity of identifying abnormal gas conditions and enhancing the accuracy of mixed gas identification.

[0014] In combination with some embodiments of the first aspect, in some embodiments, after using gas chromatography to separate the components of the mixed gas to obtain the content of each component, and comparing the preset gas types in the gas accumulation area with the content of each component to determine the gas types corresponding to each area inside the transformer, the method also includes: establishing a gas diffusion path diagram according to the gas types corresponding to each area inside the transformer, the gas diffusion path diagram includes the diffusion path and diffusion time of the gas from the generation position to the detection point; based on the gas diffusion path diagram, combined with the spatiotemporal distribution characteristics of the gas concentration and the temperature distribution characteristics, the diffusion rate of the gas on different diffusion paths is calculated; when new gas concentration change data is detected, the gas generation position is reversely located according to the diffusion rate to determine the gas source position; the transformer is located according to the gas source position to obtain the transformer fault position.

[0015] In the above embodiment, a gas diffusion path map is created based on gas type. Combining the temporal and spatial distribution characteristics of gas concentration and temperature distribution, the diffusion rates along different diffusion paths are calculated. This diffusion rate is used to reversely locate the gas generation location, accurately identifying the gas source. This enhances the accuracy of transformer fault location and forms a complete gas tracking and location mechanism.

[0016] In combination with some embodiments of the first aspect, in some embodiments, after using gas chromatography to separate the components of the mixed gas to obtain the content of each component, and comparing the preset gas type in the gas accumulation area with the content of each component to determine the gas type corresponding to each area inside the transformer, the method also includes: establishing a fault feature library based on the gas source location and gas type, the fault feature library containing gas combination characteristics and gas generation rules under different fault types; constructing a fault prediction model based on the fault feature library, the fault prediction model taking the gas combination characteristics and gas generation rules as input parameters; using the fault prediction model to evaluate the operating status of the transformer, predict potential fault risks, and generate fault warning information based on the prediction results.

[0017] In the above example, a fault signature library was established based on gas source location and gas type, and a fault prediction model was constructed that incorporates gas combination characteristics and gas generation patterns. This fault prediction model was used to assess the transformer's operating status and generate fault warning information. This established a linkage mechanism between gas type identification and fault prediction, enhancing the accuracy and foresight of transformer operating status assessments.

[0018] In combination with some embodiments of the first aspect, in some embodiments, after using gas chromatography to separate the components of the mixed gas to obtain the content of each component, and comparing the preset gas types in the gas accumulation area with the content of each component to determine the gas type corresponding to each area inside the transformer, the method also includes: calculating the spatial distance difference value and the overlap of the gas type combination between the gas accumulation area of ​​the target transformer and the gas accumulation areas of each type of transformer, and determining the transformers whose spatial distance difference value is less than a first preset threshold and whose gas type combination overlap is greater than a second preset threshold as similar transformers; performing regional matching on the gas accumulation areas of similar transformers, and comparing the gas types of each gas accumulation area in the target transformer with the corresponding area of ​​the similar transformer; if the difference value between the gas type and the main gas type of the target area corresponding to the similar transformer exceeds the preset difference threshold, then increasing the sampling point density and sampling frequency of the target area.

[0019] In the above embodiment, the spatial distance difference and gas type combination overlap between the target transformer and each type of transformer are calculated to screen similar transformers for regional matching. When the difference between the gas type and the primary gas type in the corresponding area of ​​the similar transformer exceeds a preset difference threshold, the sampling point density and sampling frequency in the target area are increased, forming an adaptive sampling mechanism based on similarity analysis, improving the data collection quality for mixed gas identification.

[0020] In a second aspect, an embodiment of the present application provides a mixed gas identification system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the mixed gas identification system to execute the method described in the first aspect and any possible implementation method of the first aspect.

[0021] In a third aspect, an embodiment of the present application provides a computer program product comprising instructions. When the computer program product is run on a mixed gas identification system, the mixed gas identification system executes the method described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium comprising instructions. When the instructions are executed on a mixed gas identification system, the mixed gas identification system executes the method described in the first aspect and any possible implementation of the first aspect.

[0023] It is understood that the mixed gas identification system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the methods provided in the embodiments of this application. Therefore, the beneficial effects achievable by these methods can be referenced to the beneficial effects of the corresponding methods and will not be further elaborated here.

[0024] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. This application achieves temperature correction of gas concentration data by obtaining a diagram of the transformer's internal structure, determining temperature distribution characteristics, and calculating solubility differences at different spatial locations. Based on this corrected data, the temporal and spatial distribution characteristics are calculated to accurately locate gas accumulation areas. Matching analysis is then performed based on the preset gas types of the corresponding components in these areas. Ultimately, this allows for precise identification of mixed gases in transformer oil, effectively improving the accuracy of gas type identification.

[0025] 2. This application establishes a gas concentration distribution model that encompasses both temporal and spatial dimensions, combining concentration data from detection points with spatial coordinate information to calculate gas concentration gradients and diffusion directions. By matching the temporal and spatial distribution characteristics with the transformer's internal structure, the application locates the areas with the greatest gas concentration variation. This provides accurate spatial positioning for subsequent gas type identification and improves the accuracy of determining gas accumulation areas.

[0026] 3. This application uses adaptively spaced sampling to obtain mixed gas samples and establishes a relationship between gas component content and time. The component types of gas accumulation areas are extracted from the transformer's internal structure diagram, and a preset gas type dataset is obtained. Pattern matching is performed by calculating similarity scores, achieving accurate gas type identification and reducing the misjudgment rate during mixed gas identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a flow chart of a method for identifying mixed gas in transformer oil according to an embodiment of the present application; Figure 2 This is another flow chart of the method for identifying mixed gas in transformer oil according to an embodiment of the present application; Figure 3 This is a schematic diagram of the structure of a physical device of the mixed gas identification system in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The terms used in the following examples of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification of the present application, the singular expressions "a", "an", "above", "the", and "this" are intended to include plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in the present application refers to any or all possible combinations of one or more of the listed items.

[0029] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to imply or suggest relative importance or implicitly indicate the number of the technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of this application, unless otherwise specified, "plurality" means two or more.

[0030] For ease of understanding, the application scenarios of the embodiments of the present application are introduced below.

[0031] A transformer contains multiple functional components, such as windings, cores, and bushings. Each component generates specific characteristic gases when it fails. For example, a winding overheating fault primarily produces C2H4, while a core failure may produce CH4 and H2. However, due to the uneven temperature distribution within the transformer, the solubility of the same gas varies significantly in different temperature zones, resulting in the detected gas concentration data not directly reflecting the actual situation. Furthermore, due to the lack of systematic analysis of the preset gas types in different components, it is difficult to accurately associate the detected gases with specific transformer components.

[0032] While related art methods can obtain the concentrations of various gases through gas chromatography during gas analysis, they are unable to correlate these gases with specific transformer components. For example, when a gas mixture containing H2, CH4, and C2H4 is detected, the lack of a mapping between component type and gas species, and the lack of consideration of the effect of temperature on solubility, makes it impossible to determine whether these gases are caused by winding overheating, core failure, or other component issues. This analysis method ignores the characteristic gas combinations produced by different components during failures, reducing the accuracy of gas identification.

[0033] The system using the method of the present application first establishes a complete transformer structure diagram, accurately recording the spatial position of each component and the corresponding temperature distribution characteristics. After the system detects gas anomalies through multi-point monitoring, it determines the gas accumulation area through temperature correction and spatiotemporal distribution analysis. More importantly, the system can identify the specific transformer component type corresponding to the accumulation area (such as high-voltage windings, low-voltage leads, etc.) and call the preset gas type data set for this type of component. These data sets contain the characteristic gas combinations and their typical proportional relationships that each component may produce under different fault types. For example, when the system detects gas accumulation in the high-voltage winding area, it will compare the gas components obtained by gas chromatography analysis with the preset gas types for this type of winding, thereby accurately identifying the specific gas types in the accumulation area. This preset gas type matching method based on component type significantly improves the accuracy and pertinence of mixed gas identification.

[0034] For ease of understanding, the following describes the process of the method provided by this implementation in combination with the above scenario. Figure 1 , which is a flow chart of a method for identifying mixed gas in transformer oil in an embodiment of the present application.

[0035] S101. Obtain an internal structure diagram of a transformer, determine the spatial position relationship of internal components of the transformer and the temperature distribution characteristics corresponding to the spatial position relationship, and obtain the solubility difference of the transformer at different spatial positions based on the temperature distribution characteristics.

[0036] Among them, the transformer internal structure diagram represents an engineering drawing showing the layout, installation position and interconnection relationship of the functional components inside the transformer; the spatial position relationship refers to the relative position and distance relationship between the components inside the transformer; the temperature distribution characteristics are used to represent the temperature distribution law and change characteristics of various parts of the transformer during operation; the solubility difference represents the difference in the solubility of gas in transformer oil due to different temperatures.

[0037] Specifically, this step is performed before the transformer is put into operation. By obtaining the transformer's design drawings and technical documentation, the internal structure of the transformer is analyzed to determine the spatial distribution of various functional components, such as the core, windings, and bushings. Combining the transformer's operating parameters with thermal field simulation analysis, a temperature field distribution model is established to calculate the temperature distribution characteristics of different regions. Based on the relationship between gas solubility in oil and temperature, data on gas solubility differences in different temperature regions is obtained.

[0038] In some embodiments, temperature distribution characteristics and solubility differences can be obtained by the following methods: Optionally, by installing temperature sensors at key locations on the transformer, collecting actual operating data, and establishing an accurate three-dimensional temperature field distribution model in conjunction with thermal field calculation software; Optionally, based on the transformer's structural parameters and operating conditions, thermal field simulation can be performed using finite element analysis to obtain steady-state and transient temperature distributions; Optionally, by estimating the temperature rise and temperature gradient of each region through theoretical calculations and empirical formulas. It will be appreciated that other temperature field analysis methods can also be used to obtain temperature distribution characteristics, which are not limited here.

[0039] S102. Continuously collect mixed gas concentration data from multiple detection points during transformer operation. When the mixed gas concentration data from multiple detection points deviate from an operating threshold range, record the initial gas concentration data from the multiple detection points, and correct the gas concentration change data based on solubility differences in different temperature zones to obtain gas concentration change data.

[0040] Among them, the detection point represents the location of the sensor installed in the transformer oil tank for collecting gas concentration data; the mixed gas concentration data refers to the content data of various characteristic gases collected at the detection point; the operating threshold range represents the allowable fluctuation range of the gas concentration during normal operation of the transformer; and the gas concentration change data is used to represent the changing trend of the gas concentration over time.

[0041] Specifically, this step is performed continuously during transformer operation, synchronously collecting gas concentration data at multiple detection points and monitoring gas concentration changes in real time. When the gas concentration exceeds the normal operating threshold, the initial concentration value at each detection point is recorded as baseline data. Taking into account the impact of temperature on gas solubility, the gas concentration data for different temperature zones is corrected based on the solubility difference data obtained in step S101 to eliminate temperature interference.

[0042] In some embodiments, gas concentration data collection and correction can be achieved through the following methods: Optionally, a fiber optic sensor array can be deployed at multiple detection points within the transformer tank to achieve real-time online monitoring of gas concentration; Optionally, a concentration correction model can be established based on the gas solubility temperature coefficient to perform temperature compensation on the raw data; Optionally, a normal operation threshold can be determined through statistical data analysis to establish anomaly detection rules. It is understood that other gas detection and data processing methods can also be used to implement this step, and are not limited here.

[0043] S103. Calculate the spatiotemporal distribution characteristics of the gas concentration based on the gas concentration change data, and determine the gas accumulation area inside the transformer according to the spatiotemporal distribution characteristics of the gas concentration and the spatial position relationship.

[0044] Among them, the spatiotemporal distribution characteristics of gas concentration represent the changing patterns of gas concentration in time and space dimensions; the gas concentration gradient refers to the rate of change of gas concentration in space; the gas diffusion direction is used to represent the migration trend of gas in transformer oil; the gas accumulation area represents the main area where gas gathers inside the transformer; and the concentration distribution model refers to a mathematical model that describes the changes of gas concentration over time and space.

[0045] Specifically, this step is performed after obtaining the corrected gas concentration change data. First, the concentration data and spatial coordinate information of each detection point are integrated to establish a gas concentration distribution model that incorporates both temporal and spatial information. The gas concentration gradient is calculated by calculating the concentration difference and spatial distance between adjacent detection points. The gas diffusion direction is analyzed by combining the temperature field distribution and oil flow field characteristics. The temporal and spatial distribution characteristics of the gas concentration are then matched and analyzed with the internal structure of the transformer to identify areas with the most significant gas concentration changes as gas accumulation areas.

[0046] In some embodiments, the calculation of the spatiotemporal distribution characteristics of gas concentration and the determination of gas accumulation areas can be achieved in a variety of ways: optionally, a three-dimensional interpolation algorithm is used to spatially reconstruct the concentration data of discrete detection points to establish a continuous concentration distribution field, the concentration gradients in each direction are calculated by partial derivative calculations, the gas diffusion path is determined by combining with oil flow field simulation, and finally, high-concentration areas are identified by cluster analysis; optionally, a dynamic diffusion model is established based on the time series data of the detection points, the time evolution characteristics of the gas concentration are analyzed, the concentration change rate and accumulation amount are calculated, the potential accumulation areas are divided in combination with the structural characteristics of the transformer, and the final gas accumulation area is determined by setting a threshold. It is understandable that other data analysis and pattern recognition methods can also be used to determine the gas accumulation area, which is not limited here.

[0047] This step specifically includes: Based on the gas concentration change data, a gas concentration distribution model including time and space dimensions is established. The gas concentration distribution model includes concentration data of multiple detection points at different times and the spatial coordinate information of the detection points.

[0048] Among them, the gas concentration distribution model refers to a mathematical model that describes the distribution state of gas inside the transformer, including the change characteristics in two dimensions: time and space; the concentration data represents the gas content value measured at each detection point, in units of ppm or μL / L; the detection point spatial coordinate information includes the three-dimensional position information (x, y, z) of the detection point inside the transformer.

[0049] The process of establishing a gas concentration distribution model begins with data acquisition. First, the layout of the detection points is determined. Gas sensors are installed at key locations on the transformer, and the precise spatial coordinates of each sensor are recorded. Continuous sampling is performed at each detection point, with a sampling interval set to 1 hour, and data is recorded continuously for at least 168 hours (7 days). The collected data includes gas concentration values ​​and corresponding timestamps. The collected data is organized into a three-dimensional matrix, where the x, y, and z axes represent spatial coordinates, the fourth dimension represents time, and the matrix element values ​​represent the gas concentration at the corresponding time and space points. A three-dimensional interpolation algorithm is used to spatially interpolate the discrete detection point data to generate a continuous concentration distribution field. A spline interpolation method is used in the time dimension to obtain a continuous concentration change curve over time. Ultimately, a complete four-dimensional concentration distribution model is established to describe the dynamic changes in gas concentration in time and space.

[0050] The gas concentration gradient and the gas diffusion direction are calculated based on the gas concentration distribution model, and the spatiotemporal distribution characteristics of the gas concentration are obtained based on the gas concentration gradient and the gas diffusion direction.

[0051] Among them, the gas concentration gradient represents the rate of change of concentration per unit distance, reflecting the intensity of the change in the concentration field; the gas diffusion direction refers to the main direction of movement of gas molecules, which is determined by the normal vector of the isosurface of the concentration field; the spatiotemporal distribution characteristics of gas concentration include the spatial distribution characteristics and temporal evolution characteristics of the concentration field.

[0052] The concentration gradient and diffusion direction are calculated based on the established concentration distribution model. In the spatial dimension, the central difference method is used to calculate the concentration gradient vector of each point: ∇C=(∂C / ∂x, ∂C / ∂y, ∂C / ∂z). The modulus of the gradient vector represents the rate of change of concentration, and the direction points to the direction where the concentration increases fastest. At each spatial point, the isosurface of the local concentration field is calculated, and the normal vector of the isosurface is the diffusion direction of the point. The time series data is analyzed to calculate the time derivative of the concentration ∂C / ∂t to describe the rate of change of concentration over time. The spatial gradient and time derivative information are combined to construct a spatiotemporal distribution feature descriptor of the gas concentration. This descriptor includes: spatial distribution characteristics (gradient field distribution, diffusion direction field) and time evolution characteristics (concentration change rate, periodic characteristics).

[0053] The spatiotemporal distribution characteristics of gas concentration are matched with the spatial position relationship of each component inside the transformer, and the area with the largest change in gas concentration is determined as the gas accumulation area.

[0054] Among them, the spatial position relationship represents the relative position relationship between the gas concentration field and the transformer structural components; the gas accumulation area refers to the spatial area where the gas concentration continues to increase and remains at a high level, which is usually closely related to the fault location.

[0055] When performing spatial matching, first align the coordinates of the three-dimensional structural model of the transformer with the concentration distribution model. Calculate the average concentration value and concentration change rate around each structural component. Identify areas where the concentration change rate exceeds the set threshold (such as 0.5ppm / h) and mark these areas as potential gas accumulation areas. For each potential area, calculate the spatial second-order derivative of the concentration field to determine whether the area is a local maximum point. Select the area with a negative second-order derivative and the highest concentration value as the main gas accumulation area. Record the spatial range, center coordinates and corresponding structural component information of the accumulation area. Verify the stability of the area identification results by continuously monitoring the evolution of the accumulation area. The gas accumulation area is finally determined as the key focus area for fault diagnosis.

[0056] S104. Separate the mixed gas components using gas chromatography to obtain the content of each component, and compare the preset gas types in the gas accumulation area with the content of each component to determine the gas type corresponding to each area inside the transformer.

[0057] Among them, gas chromatography refers to an analytical method that uses the differences in the distribution coefficients of different gas components in the stationary phase and mobile phase to separate them; the response curve represents the curve of the change of the detector output signal over time during the chromatographic analysis process; the component content refers to the concentration or volume ratio of various gases in the mixed gas; the preset gas type data set is used to represent the characteristic gases that may be produced under different types of faults and their typical content ratios; the similarity score indicates the degree of match between the measured gas components and the preset data.

[0058] Specifically, this step is performed after the gas accumulation area is determined. First, an adaptive sampling strategy is used to obtain gas samples based on the location characteristics of the accumulation area. The samples are analyzed using a gas chromatograph, and the response curves of each component are recorded. The component content is calculated through integration. Component information corresponding to the accumulation area is extracted from the transformer structure diagram, and a preset gas type dataset is queried to obtain possible gas signatures. The measured gas component content is pattern-matched with the preset data, and a similarity score is calculated. The gas type with the best match is selected as the characteristic gas for the area.

[0059] In some embodiments, the analysis of mixed gas components and the determination of gas types can be achieved in a variety of ways: optionally, a programmed temperature gas chromatography method is used to optimize the carrier gas flow rate and column temperature program to improve the chromatographic separation, the content of each component is quantitatively analyzed by the external standard method, a curve of the change of component content over time is established, and a pattern recognition algorithm is used for feature matching; optionally, a multidimensional gas feature vector is established, including parameters such as component content, generation rate, ratio relationship, etc., a classification model based on a support vector machine is constructed, and the model is trained in combination with historical case data to achieve automatic identification of gas types. It is understandable that other analytical chemistry and machine learning methods can also be used to achieve gas type identification, which is not limited here.

[0060] This step specifically includes: According to the distribution of gas accumulation areas, an adaptive interval sampling method is used to obtain mixed gas samples. The mixed gas samples are separated by gas chromatography to obtain the response curves and component contents of each gas component, and the relationship between the change of gas component content and time is established.

[0061] Among them, the adaptive interval sampling method refers to a sampling strategy that dynamically adjusts the sampling time interval according to the rate of change of gas concentration; the mixed gas sample refers to a sample containing multiple gas components extracted from transformer oil; gas chromatography is an analytical method that uses the differences in the distribution coefficients of different gas components in the stationary phase and mobile phase to separate them; the response curve represents the curve of the change of the detector output signal over time during the chromatographic analysis process.

[0062] When performing gas composition analysis, sampling points are first arranged in the gas accumulation area. An adaptive control strategy is used for the sampling interval: when the concentration change rate is greater than 1 ppm / h, the sampling interval is set to 2 hours; when the change rate is between 0.1 and 1 ppm / h, the sampling interval is set to 6 hours; when the change rate is less than 0.1 ppm / h, the sampling interval is set to 12 hours. A gas extraction device is used to extract dissolved gases from the transformer oil, with each sample volume of 50 mL. The sample is injected into a gas chromatograph for analysis, with the chromatographic conditions set as: helium carrier gas, a flow rate of 30 mL / min, a column temperature of 60°C, and a detector temperature of 200°C. The complete chromatogram is recorded, and each gas component is identified by retention time. The chromatographic peaks are integrated, and the content of each component is calculated based on the peak area and the standard curve. A curve of the change in the content of each gas component over time over a 24-hour period is constructed, and the curve is processed continuously using piecewise linear interpolation.

[0063] The transformer component type where the gas accumulation area is located is extracted from the transformer internal structure diagram, and the corresponding preset gas type dataset is obtained based on the component type. The preset gas type dataset contains the typical content ratio relationship of various gas components under different fault types.

[0064] Among them, the transformer internal structure diagram represents an engineering drawing that shows the spatial layout and connection relationships of the transformer components; component types include functional units such as windings, cores, and bushings; the preset gas type data set is a gas characteristic database established based on historical fault cases and theoretical analysis; and the typical content ratio relationship refers to the relative content ratio of characteristic gases under different fault types.

[0065] When extracting component information in the gas accumulation area, first substitute the coordinates of the area center into the three-dimensional structural model of the transformer. Identify the component type where the coordinate point is located, and record the basic characteristics of the component, including material properties, operating temperature, and stress distribution information. Retrieve a preset gas type dataset based on the component type. This dataset contains characteristic gas information generated by each component under different fault modes. The structure of the dataset is: component type-fault type-gas component-content ratio. For example, for a winding overheating fault, the typical gas component ratio is: C2H4:C2H6:CH4=3:1:1. For a core grounding fault, the typical ratio is: H2:CH4:C2H6=4:2:1. Extract all gas ratio data related to the current component type to form a set of feature templates to be matched.

[0066] The changing relationship of gas component content is pattern matched with the typical content ratio relationship in the preset gas type data set. The similarity score is calculated based on the matching results, and the preset gas type with the highest similarity score is selected as the gas type corresponding to each area inside the transformer.

[0067] Among them, pattern matching refers to the process of comparing the measured gas component data with the preset feature template; the similarity score indicates the degree of matching between the measured data and the feature template; the preset gas type refers to the most likely gas combination type determined based on the matching results.

[0068] When performing pattern matching, the measured gas component data is first standardized. The content ratios of each component are calculated to construct a feature vector. For each feature template in the preset dataset, the cosine similarity with the measured feature vector is calculated. The similarity calculation formula is: cos(θ) = (A·B) / (|A|·|B|), where A is the measured feature vector and B is the template feature vector. The Euclidean distance is also calculated as an auxiliary evaluation metric. The similarity score and distance score are combined with a weight of 0.7:0.3 to obtain a comprehensive score. The gas type corresponding to the feature template with the highest comprehensive score is selected as the characteristic gas combination for that area. The matching results, including the gas type, matching score, and suboptimal matching result, are recorded to provide a basis for subsequent fault diagnosis.

[0069] In some embodiments, the following steps may be further included before step 101: The historical data of mixed gas concentration at multiple detection points in the transformer tank are obtained, and the time difference when the maximum gas concentration is detected at each detection point is calculated.

[0070] Among them, the historical data of mixed gas concentration represents the content record of dissolved gas in transformer oil over the past period of time, in units of μL / L or ppm; the detection point refers to the location of the gas sensor installed in the transformer oil tank; the time difference of maximum concentration refers to the time interval between different detection points reaching the gas concentration peak, reflecting the time characteristics of the gas diffusion process.

[0071] When acquiring historical data, first determine the time range for data collection, typically selecting the most recent 30 days of continuous monitoring data. For each monitoring point, a 5-minute sampling interval is used to record the concentrations of characteristic gases such as C₂H₂, CH₄, C₂H₄, C₂H₆, and H₂. Raw data is preprocessed, including outlier removal and data smoothing. Outlier removal uses the 3σ criterion, marking data points that deviate from the mean by more than three standard deviations as outliers. Data smoothing uses a 5-point moving average to reduce the impact of random fluctuations. For the processed data sequence, identify the peak concentration and its occurrence time at each monitoring point. Calculate the time difference between adjacent monitoring points to reach the peak, denoted as Δt. For example, if the closest monitoring point pair reaches a peak of 100 ppm at time t1 and monitoring point B reaches a peak of 80 ppm at time t2, the time difference Δt = t2 - t1. The time difference data for all monitoring point pairs is organized into a matrix for subsequent analysis.

[0072] The operating threshold range during normal operation is obtained based on the ratio of the maximum concentration values ​​of each detection point.

[0073] Among them, the ratio of maximum concentration values ​​represents the ratio of the maximum gas concentrations detected at adjacent detection points; the operating threshold range refers to the reasonable variation range of the gas concentration ratio under normal operating conditions of the transformer.

[0074] To calculate the operating threshold range, first calculate the maximum concentration ratio (r = Cmax1 / Cmax2) for each pair of adjacent monitoring points, where Cmax1 and Cmax2 are the maximum concentrations detected at the two monitoring points, respectively. The concentration ratio data for all pairs of monitoring points are then tallied, and their mean μ and standard deviation σ are calculated. Based on the characteristics of a normal distribution, the interval [μ - 2σ, μ + 2σ] is defined as the threshold range for normal operation. This interval encompasses approximately 95% of the normal operating data. For example, if the concentration ratio statistics for a pair of monitoring points show a mean μ = 1.2 and a standard deviation σ = 0.1, the operating threshold range is [1.0, 1.4]. Independent threshold ranges are established for each pair of monitoring points, forming a complete threshold matrix. When the measured concentration ratio exceeds the threshold range, it indicates an abnormal gas distribution, requiring further analysis. The calculated threshold range results are recorded in a database and serve as a reference for subsequent monitoring.

[0075] The following is a more detailed description of the process of the method provided by this implementation. Figure 2 , is another flow chart of the method for identifying mixed gas in transformer oil in an embodiment of the present application.

[0076] S201. Establish a gas diffusion path diagram according to the gas type corresponding to each area inside the transformer. The gas diffusion path diagram includes the diffusion path and diffusion time of the gas from the generation position to the detection point.

[0077] Among them, the gas diffusion path diagram represents a spatial network structure diagram describing the gas migration process in transformer oil, which contains information such as migration path, node location and transmission time; the diffusion path refers to the actual movement trajectory of gas molecules from the generation position to the detection point in transformer oil, which is jointly affected by the oil flow velocity field and temperature field; the diffusion time refers to the time interval required for the gas to be transmitted from the initial generation position to a specific detection point, which is related to factors such as path length, oil flow velocity and temperature distribution.

[0078] After determining the gas types corresponding to each region, the gas diffusion path map was constructed. First, the internal oil channel layout, including the main oil channel, auxiliary oil channel, and natural convection channel, was analyzed by analyzing the transformer structural diagram. A complete oil channel network model was then established. All detection points and possible gas generation locations were annotated as network nodes, and the connectivity between nodes was analyzed using graph theory. Three-dimensional modeling and simulation were performed for each oil channel using computational fluid dynamics software to calculate the oil velocity field distribution under steady-state conditions. Based on the physicochemical properties of different gases, such as molecular weight and diffusion coefficient, and combined with the oil flow field characteristics, the gas trajectories in each channel were calculated. Numerical integration methods were used to calculate the time required for the gas to reach the detection point along each possible path from the generation location. The resulting multi-layered path map includes primary, secondary, and emergency channels. Each channel is annotated with detailed spatial coordinate information and transmission time data. This path map provides important spatial topological information for subsequent gas source location and fault diagnosis.

[0079] S202. Based on the gas diffusion path diagram and in combination with the spatiotemporal distribution characteristics of the gas concentration and the temperature distribution characteristics, the diffusion rate of the gas on different diffusion paths is calculated.

[0080] Among them, the diffusion rate represents the migration speed of gas molecules in the transformer oil, which includes the combined effect of two mass transfer mechanisms: convection diffusion and molecular diffusion. The spatiotemporal distribution characteristics of concentration describe the variation law of gas concentration in time and space dimensions, reflecting the diffusion process of gas inside the transformer. The temperature distribution characteristics represent the temperature field distribution state of each region of the transformer, which affects the diffusion coefficient of gas and the flow characteristics of oil.

[0081] Based on the established gas diffusion path map, the diffusion rate is accurately calculated. First, the gas concentration data at each detection point is analyzed over time, and the rate of change of concentration over time is calculated using numerical difference methods. For each diffusion path, the spatial coordinates of the detection points at both ends of the path are extracted, and the actual length and geometric characteristics of the path (such as tortuosity and cross-sectional area change) are calculated. Combined with the concentration data at the detection points, the concentration gradient along the path is calculated. To account for the influence of temperature on the gas diffusion process, a modified Fick diffusion equation is used for modeling. The calculation begins by obtaining the average temperature and temperature gradient along the path. The actual diffusion coefficient is calculated based on the temperature dependence of the gas diffusion coefficient. The diffusion coefficient, concentration gradient, and path geometric parameters are substituted into the diffusion equation to solve for the gas diffusion flux. For areas with forced convection, the contribution of the oil flow velocity is added to obtain the combined mass transfer flux. The gas diffusion rate along that path is calculated by dividing the mass transfer flux by the effective cross-sectional area of ​​the path. This process is repeated to calculate the rates for all diffusion paths, establishing a complete diffusion rate database. This data provides a quantitative basis for subsequent fault source location.

[0082] S203: When new gas concentration change data is detected, the gas generation position is reversely located according to the diffusion rate to determine the gas source position.

[0083] Among them, the gas concentration change data represents the dynamic change information of gas concentration collected in real time at the detection point, including the concentration value and the corresponding timestamp; the reverse positioning process refers to the use of detection data and known diffusion characteristics to reversely infer the initial generation location of the gas; the gas source location represents the exact spatial location where the fault gas was initially generated, corresponding to the specific fault point inside the transformer.

[0084] Gas concentration data is continuously monitored during transformer operation. A reverse location algorithm is initiated when an abnormal change is detected. Data from all detection points is preprocessed, including noise filtering, data smoothing, and outlier detection. The moment the first significant change in gas concentration is detected at each detection point is extracted as the time stamp for the gas's arrival at that point. Combined with the calculated diffusion rate data, a system of equations based on spatiotemporal relationships is established. These equations include the spatial coordinates of each detection point, the detection time, and the diffusion rate along the corresponding path, with the three-dimensional coordinates of the gas source as the unknown. The least squares method is used to solve the system, and an iterative optimization process is performed to minimize the sum of the squared errors between the calculated theoretical transmission time and the actual detection time. During the optimization process, weighting coefficients are introduced to account for measurement and model errors. Upon convergence, the optimal gas source coordinates are obtained. These coordinates are then matched with the transformer structure diagram to determine the specific location of the fault, providing precise spatial location information for subsequent repair and resolution. The accuracy of source location is continuously improved through continuous updates and refinement of the location algorithm.

[0085] S204: Locate the transformer according to the gas source location to obtain the transformer fault location.

[0086] Among them, the gas source location represents the coordinate point where the fault gas is generated inside the transformer; the transformer fault location refers to the specific component or area location that causes the gas to be generated; and the positioning process represents the conversion process of mapping the gas source location to the actual structure of the transformer.

[0087] After obtaining the spatial coordinates of the gas source, they are converted to the fault location within the actual transformer structure. First, a 3D structural model of the transformer is created, including the precise location of all functional components, including the windings, core, bushing, and leads. The gas source coordinates are imported into the 3D model, and the structural location of the point is determined through spatial coordinate transformation. Based on the component's geometric dimensions and installation position, the shortest distance from the source point to each component surface is calculated. A distance threshold of 50 mm is set, and components with distances below the threshold are marked as potential fault locations. For each potential fault location, its operating characteristics and stress distribution are analyzed to assess the plausibility of the mechanism generating the fault gas. Factors such as component stress, temperature distribution, and electric field distribution are comprehensively considered to determine the final fault location. The fault location is annotated using the numbering scheme found in the equipment drawings, accurately identifying the specific component unit, such as "third lamination of the high-voltage winding, phase A" or "second elbow of the low-voltage lead, phase B." The surrounding environmental characteristics of the fault location are also recorded, providing complete spatial information for inspection and repair.

[0088] S205 . Establish a fault feature library based on the gas source location and gas type. The fault feature library contains gas combination features and gas generation rules under different fault types.

[0089] Among them, the fault feature library represents a data set that records the correspondence between transformer fault types and gas characteristics; the gas combination characteristics refer to the types of characteristic gases produced by different types of faults and their proportional relationships; the gas generation law represents a mathematical model of how gas concentration changes over time.

[0090] The process of establishing a fault feature library starts with data collection. The gas source location data in historical fault cases is collected, including information such as spatial coordinates, component type, and surrounding environment. For each fault case, the gas component data obtained by gas chromatography analysis is extracted, and the content of various gases and their ratio relationships are recorded. The rate of change of gas concentration is calculated through time series analysis, the concentration-time curve is fitted, and the characteristic parameters of the curve such as growth rate and saturation value are extracted. For each fault type, its typical gas combination pattern is statistically analyzed, including the main gas type, secondary gas type and its content range. A gas generation law model is established to describe the complete evolution process of gas concentration from fault occurrence to stability. All characteristic data are classified and stored according to the fault type to form a structured feature library. Each record contains complete information such as fault location, fault type, gas combination characteristics, concentration change curve, etc., which is convenient for subsequent pattern matching and fault prediction.

[0091] S206 , constructing a fault prediction model based on the fault feature library, wherein the fault prediction model takes gas combination characteristics and gas generation rules as input parameters.

[0092] The fault prediction model represents a mathematical model used to predict potential transformer faults; the input parameters include real-time monitored gas data and historical features extracted from a feature library; and the prediction results represent a quantitative assessment of the fault type and development trend.

[0093] A prediction model is constructed based on a complete fault feature library. First, the data in the feature library is preprocessed, including data standardization, feature dimensionality reduction, and outlier processing. Key feature parameters are selected as model inputs, including gas type combinations, concentration ratios, and change rates. A multi-layer prediction model is constructed using machine learning methods: the first layer uses a support vector machine to identify fault types, the second layer uses a neural network to predict fault development trends, and the third layer uses a decision tree to assess fault severity. During model training, the feature library data is divided into a training set and a validation set in a ratio of 7:3. Model parameters are optimized through cross-validation to improve prediction accuracy. For each new monitoring data point, it is analyzed in turn using the three-layer model to obtain the probability distribution of the fault type, the time curve of the development trend, and the quantitative score of the severity. Based on the prediction results, a standardized evaluation report is generated, which contains specific fault diagnosis conclusions and treatment recommendations.

[0094] S207. Evaluate the transformer operating status using the fault prediction model, predict potential fault risks, and generate fault warning information based on the prediction results.

[0095] Among them, the transformer operating status indicates the working conditions of the transformer at a certain moment, including parameters such as load level, temperature distribution and gas content; potential fault risk refers to the probability and severity of future faults obtained through prediction model evaluation; fault warning information includes fault type judgment, development trend prediction and processing suggestions.

[0096] Real-time transformer condition assessment begins by collecting operating parameter data, including gas concentration, temperature, and load current at each monitoring point. This data is fed into the first layer of the fault prediction model, where a support vector machine calculates the probability distribution of the current state belonging to various fault types. The three fault types with the highest probability are selected and fed into the second layer of the prediction model's neural network to calculate fault development trends. The neural network uses a time series prediction method to predict gas concentration changes over the next 24 hours, 72 hours, and seven days. For each prediction result at a specific time point, a decision tree model is used to assess the fault severity and categorize it into four levels: normal, caution, warning, and dangerous. If the prediction results indicate that the severity at a specific time point reaches the warning or dangerous level, the fault warning mechanism is triggered. The generated warning information includes the fault type and probability, a gas concentration trend chart, the estimated time to reach the critical value, recommended remedial measures, and key operating parameters requiring attention. Warning information is delivered using different delivery methods based on severity, ensuring that relevant personnel receive timely fault risk information.

[0097] S208. Calculate the spatial distance difference between the gas accumulation area of ​​the target transformer and the gas accumulation areas of each type of transformer and the overlap of the gas type combination, and determine the transformers whose spatial distance difference is less than a first preset threshold and whose gas type combination overlap is greater than a second preset threshold as similar transformers.

[0098] Among them, the spatial distance difference value represents the relative position deviation of the gas accumulation areas of the two transformers in three-dimensional space; the coincidence of gas type combinations refers to the similarity of the gas types and their ratios detected in the two transformers; similar transformers refer to other transformers that have a high degree of similarity with the target transformer in terms of structural characteristics and fault characteristics.

[0099] When performing similar transformer screening, the coordinates of the gas accumulation area of ​​the target transformer are first standardized to a unified spatial reference system. For each transformer in the database, the Euclidean distance between its gas accumulation area and the corresponding area of ​​the target transformer is calculated. The calculation of the spatial distance difference value takes into account the shape and size of the area, and uses two indicators: center of gravity distance and contour overlap. At the same time, the overlap of the gas type combination is calculated. The specific method is: after normalizing the various gases according to their content, a feature vector is constructed, and the cosine similarity between the vectors is calculated. The first preset threshold is set to 20% of the standardized spatial distance, and the second preset threshold is set to 0.8 of the gas overlap. Transformers that meet the dual threshold conditions are selected as a set of similar transformers. For the screened similar transformers, a one-to-one correspondence between the gas accumulation areas is established to provide a basis for subsequent detailed comparison.

[0100] S209: Perform regional matching on the gas accumulation regions of similar transformers, and compare the gas types in each gas accumulation region in the target transformer with the corresponding regions of the similar transformers.

[0101] Among them, regional matching means establishing a spatial correspondence between the gas accumulation areas of the target transformer and similar transformers; corresponding areas refer to the parts of the transformer that correspond to each other in spatial position and functional characteristics; gas type comparison refers to analyzing the similarities and differences in gas components at the same location in different transformers.

[0102] Perform regional matching analysis for each similar transformer. Divide the gas accumulation area into several sub-areas, each of which corresponds to a functional unit of the transformer. Establish a regional feature descriptor, which contains information such as spatial position, geometric shape, and component type. Use a feature matching algorithm to calculate the matching degree between the target transformer and each sub-area of ​​the similar transformer. The matching degree calculation comprehensively considers three aspects: spatial position similarity, structural feature similarity, and functional attribute similarity. After determining the regional correspondence, compare the gas types in each area in detail. Record the differences in gas types, including: component differences, content differences, and time series feature differences. For each difference point, analyze its cause and evaluate whether the monitoring strategy needs to be adjusted. The comparison results are used to optimize the sampling plan and update the fault feature library.

[0103] S210: If the difference between the gas type and the main gas type in the target area corresponding to the similar transformer exceeds a preset difference threshold, increase the sampling point density and sampling frequency of the target area.

[0104] Among them, the gas type difference value represents the quantitative difference index of the gas component content in the corresponding areas of the two transformers; the main gas type refers to the gas component whose content in a specific area accounts for more than 10%; the sampling point density refers to the number of gas detection points arranged per unit volume; the sampling frequency refers to the time interval for gas sampling and analysis at each detection point; the preset difference threshold represents the standard value for judging whether the difference in gas types is significant, which is usually set to 20%.

[0105] After completing the regional matching of similar transformers, gas species difference analysis is performed for each target area. First, gas composition information for each area is extracted from the gas chromatography data, and the percentage of each gas is calculated. Gas components with a concentration exceeding 10% are recorded as the primary gas species in that area. The difference in primary gas species between the target transformer and the corresponding area of ​​the similar transformer is calculated. The specific calculation method is: for each primary gas, the relative deviation of the gas content in the two transformers is calculated, and the root mean square value of all relative deviations for all primary gases is taken as the difference value. If the difference exceeds 20%, the monitoring plan for that area needs to be adjusted. The specific approach to increasing the sampling point density is to add new testing points based on a spatially uniform distribution principle to the existing testing points, reducing the spacing between testing points to half the original value. At the same time, the sampling frequency is increased, shortening the sampling interval from 6 hours to 2 hours. Calibration and data quality verification are performed on the newly added testing points to ensure the accuracy of the monitoring data. High-density monitoring data is continuously collected for at least 7 days to analyze the causes of the gas species differences. By increasing monitoring intensity, more detailed gas distribution information is obtained, providing more reliable data support for fault diagnosis.

[0106] The following describes the mixed gas identification system in the embodiment of the present invention from the perspective of hardware processing. Figure 3 , is a schematic diagram of the physical device structure of the mixed gas identification system in an embodiment of the present application.

[0107] It should be noted that Figure 3 The structure of the mixed gas identification system shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention.

[0108] like Figure 3 As shown, the mixed gas identification system includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes, such as the methods described in the above embodiments, based on programs stored in a read-only memory (ROM) 302 or programs loaded from a storage unit 308 into a random access memory (RAM) 303. RAM 303 also stores various programs and data required for system operation. CPU 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to bus 304.

[0109] The following components are connected to the I / O interface 305: an input section 306 including an audio input device, push button switches, and the like; an output section 307 including a liquid crystal display (LCD), an audio output device, indicator lights, and the like; a storage section 308 including a hard disk and the like; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card or a modem. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. Removable media 311, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 310 as needed, so that computer programs read from the removable media can be installed in the storage section 308 as needed.

[0110] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program including a computer program for executing the methods illustrated in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via the communication section 309 and / or installed from removable media 311. When executed by the central processing unit (CPU) 301, the computer program performs the various functions defined in the present invention.

[0111] It should be noted that specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more conductors, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0112] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram can represent a module, program segment, or part of the code, and the above-mentioned module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings.

[0113] Specifically, the mixed gas identification system of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, the mixed gas identification method in transformer oil provided by the above embodiment is implemented.

[0114] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the mixed gas identification system described in the above embodiments, or may exist independently and not incorporated into the mixed gas identification system. The storage medium carries one or more computer programs, which, when executed by a processor of the mixed gas identification system, enable the mixed gas identification system to implement the method for identifying mixed gases in transformer oil provided in the above embodiments.

[0115] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0116] As used in the above embodiments, the term “when” may be interpreted to mean “if” or “after” or “in response to determining that” or “in response to detecting that”, depending on the context. Similarly, the phrases “upon determining that” or “if (the stated condition or event) is detected” may be interpreted to mean “if it is determined that” or “in response to determining that” or “upon detecting (the stated condition or event)” or “in response to detecting (the stated condition or event)”, depending on the context.

[0117] Those skilled in the art will appreciate that all or part of the process steps in the above-described method embodiments can be implemented by a computer program instructing the relevant hardware. The program can be stored in a computer-readable storage medium, and when executed, the program can include the process steps in the above-described method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for identifying mixed gas in transformer oil, characterized in that: Applied to a mixed gas identification system, the method includes: Obtaining an internal structure diagram of the transformer, determining a spatial position relationship between internal components of the transformer and a temperature distribution characteristic corresponding to the spatial position relationship, and obtaining a solubility difference of the transformer at different spatial positions based on the temperature distribution characteristic; Continuously collecting mixed gas concentration data at multiple detection points during the operation of the transformer, recording initial gas concentration data at the multiple detection points when the mixed gas concentration data deviate from an operating threshold range, and correcting the gas concentration change data according to solubility differences in different temperature zones to obtain gas concentration change data; Calculating the spatiotemporal distribution characteristics of gas concentration based on the gas concentration change data, and determining a gas accumulation area inside the transformer according to the spatiotemporal distribution characteristics of gas concentration and the spatial position relationship; Gas chromatography is used to separate the components of the mixed gas to obtain the content of each component, and the preset gas types in the gas accumulation area are compared with the content of each component to determine the gas type corresponding to each area inside the transformer.

2. The method according to claim 1, characterized in that The step of calculating the spatiotemporal distribution characteristics of gas concentration based on the gas concentration change data, and determining the gas accumulation area of ​​gas inside the transformer according to the spatiotemporal distribution characteristics of gas concentration and the spatial position relationship specifically includes: Establishing a gas concentration distribution model including time and space dimensions based on the gas concentration change data, wherein the gas concentration distribution model includes concentration data of a plurality of detection points at different times and spatial coordinate information of the detection points; Calculating a gas concentration gradient and a gas diffusion direction according to the gas concentration distribution model, and obtaining a spatiotemporal distribution characteristic of the gas concentration according to the gas concentration gradient and the gas diffusion direction; The spatiotemporal distribution characteristics of the gas concentration are matched with the spatial position relationship of the components inside the transformer, and the area with the largest change in gas concentration is determined as the gas accumulation area.

3. The method according to claim 1, characterized in that The step of separating the mixed gas components by gas chromatography to obtain the content of each component, and comparing the preset gas types in the gas accumulation area with the content of each component to determine the gas type corresponding to each area inside the transformer specifically includes: According to the distribution location of the gas accumulation area, a mixed gas sample is obtained by adopting an adaptive interval sampling method, the mixed gas sample is separated by gas chromatography, the response curve and component content of each gas component are obtained, and the relationship between the change of the gas component content and time is established; Extracting the type of transformer component where the gas accumulation area is located from the transformer internal structure diagram, and obtaining a corresponding preset gas type data set based on the component type, wherein the preset gas type data set includes a typical content ratio relationship of various gas components under different fault types; The change relationship of the gas component content is pattern matched with the typical content ratio relationship in the preset gas type data set, and a similarity score is calculated based on the matching result. The preset gas type with the highest similarity score is selected as the gas type corresponding to each area inside the transformer.

4. The method according to claim 1, wherein Before the steps of obtaining the internal structure diagram of the transformer, determining the spatial position relationship of the internal components of the transformer and the temperature distribution characteristics corresponding to the spatial position relationship, and obtaining the solubility difference of the transformer at different spatial positions based on the temperature distribution characteristics, the method further includes: Obtain historical data on the mixed gas concentration at multiple detection points in the transformer tank, and calculate the time difference between each detection point detecting the maximum gas concentration; The operating threshold range during normal operation is obtained according to the ratio of the maximum concentration values ​​of each detection point.

5. The method according to claim 1, wherein After the steps of separating the mixed gas components by gas chromatography to obtain the contents of each component, and comparing the preset gas types in the gas accumulation area with the contents of each component to determine the gas type corresponding to each area inside the transformer, the method further includes: Establishing a gas diffusion path diagram according to the gas type corresponding to each area inside the transformer, wherein the gas diffusion path diagram includes the diffusion path and diffusion time of the gas from the generation location to the detection point; Based on the gas diffusion path diagram, combined with the spatiotemporal distribution characteristics of the gas concentration and the temperature distribution characteristics, the diffusion rate of the gas on different diffusion paths is calculated; When new gas concentration change data is detected, the gas generation position is reversely located according to the diffusion rate to determine the gas source position; The transformer is located according to the gas source position to obtain the transformer fault position.

6. The method according to claim 5, characterized in that After the steps of separating the mixed gas components by gas chromatography to obtain the contents of each component, and comparing the preset gas types in the gas accumulation area with the contents of each component to determine the gas type corresponding to each area inside the transformer, the method further includes: Establishing a fault feature library based on the gas source location and the gas type, wherein the fault feature library contains gas combination characteristics and gas generation rules under different fault types; Building a fault prediction model based on the fault feature library, wherein the fault prediction model uses the gas combination feature and the gas generation law as input parameters; The fault prediction model is used to evaluate the transformer operating status, predict potential fault risks, and generate fault warning information based on the prediction results.

7. The method according to claim 1, characterized in that After the steps of separating the mixed gas components by gas chromatography to obtain the contents of each component, and comparing the preset gas types in the gas accumulation area with the contents of each component to determine the gas type corresponding to each area inside the transformer, the method further includes: Calculate the spatial distance difference between the gas accumulation area of ​​the target transformer and the gas accumulation areas of each type of transformer and the overlap of the gas type combination, and determine the transformers with the spatial distance difference less than a first preset threshold and the gas type combination overlap greater than a second preset threshold as similar transformers; Performing region matching on the gas accumulation regions of the similar transformers, and comparing the gas types in each gas accumulation region of the target transformer with the gas types in the corresponding region of the similar transformer; If the difference between the gas type and the main gas type in the target area corresponding to the similar transformer exceeds a preset difference threshold, the sampling point density and sampling frequency of the target area are increased.

8. A mixed gas identification system, characterized in that: The mixed gas identification system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the mixed gas identification system to execute the method described in any one of claims 1 to 7.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instruction is executed on a mixed gas identification system, the mixed gas identification system is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on a mixed gas identification system, the mixed gas identification system is enabled to perform the method according to any one of claims 1 to 7.

Images