A method, device and computer program product for state evaluation of an environmentally friendly gas-insulated current transformer

Abstract

The application discloses a kind of state evaluation method, device and computer program product of environmental protection gas insulation current transformer, it is related to high-voltage power equipment state monitoring and low-carbon operation and maintenance technical field.The method obtains the gas chamber structure parameter of target equipment, insulation gas attribute parameter, and gas chamber state monitoring data and operating condition data;Gas quality is calculated based on gas state equation, and initial leakage is determined in combination with operation and maintenance event;Initial leakage is corrected using an artificial intelligence model with built-in physical constraints, decoupling environmental disturbance and sensor drift interference;Synchronously complete leakage carbon emission equivalent accounting, gas chamber insulation margin evaluation and output comprehensive risk early warning.The present application can improve the accuracy of micro-leakage identification, achieve traceable carbon emission accounting, balance equipment insulation safety and low-carbon operation and maintenance, and is suitable for lean operation and control of environmental protection gas current transformer in euhypertensive substation.

Description

Technical Field

[0001] This invention relates to the field of power equipment monitoring technology, specifically to a condition assessment method, device, and computer program product for an environmentally friendly gas-insulated current transformer. Background Technology

[0002] In ultra-high voltage substations with voltage levels of 500kV and above, current transformers (CTs) serve as the core primary equipment interface for relay protection, metering, fault recording, and condition monitoring. Their insulation and sensing performance directly affects system safety and measurement reliability. With the development of green and low-carbon power equipment technologies, clean air, nitrogen-oxygen mixtures, carbon dioxide mixtures, and low global warming potential (GWP) environmentally friendly insulating gases are gradually replacing traditional oil-immersed or fluorinated gas insulation solutions. Environmentally friendly gas-insulated current transformers have significant advantages in reducing greenhouse gas emissions and improving environmental friendliness. However, their gas chamber sealing characteristics, gas density maintenance requirements, insulation margin design, and gas replenishment maintenance methods differ significantly from traditional equipment, posing new technical challenges to condition monitoring and leakage management.

[0003] During on-site operation, the monitored parameters of environmental gas current transformers, such as gas pressure, gas density, and trace moisture, are affected by a combination of factors, including diurnal temperature fluctuations, load heating, altitude changes, gas compressibility factor deviations, sensor drift, gas replenishment operations, and valve status. Currently, commonly used leakage detection methods are mostly based on fixed thresholds or criteria with simple temperature compensation, making it difficult to effectively distinguish between environmental disturbances, sensor drift, and actual leaks. This easily leads to misjudging normal fluctuations as leaks, or mistaking slowly developing, minute leaks for normal fluctuations. Consequently, problems arise such as inaccurate leakage estimation, unreasonable timing of gas replenishment, and a lack of complete data for carbon emission calculations, hindering the refinement of equipment operation and maintenance and the reliability of carbon reduction management.

[0004] On the other hand, leaks of environmentally friendly gases not only involve greenhouse gas emissions but also directly affect the insulation strength and gas-solid interface safety margin of the CT gas chamber. A continuous decrease in gas density, abnormal moisture content, or changes in gas composition can all lead to a reduction in the partial discharge initiation voltage (PDIV), thereby inducing insulation risks. Therefore, it is urgent to unify leakage inversion, carbon equivalent calculation, insulation margin prediction, and risk warning to construct a traceable, verifiable, and closed-loop technical system, enabling the integration and collaborative decision-making of equipment operation safety and green, low-carbon management.

[0005] Traditional carbon emission ledger management relies primarily on manual entry of gas replenishment records, maintenance records, and gas procurement records, followed by post-event statistics based on gas type and corresponding global warming potential conversion factors. This method is essentially a bookkeeping-style aggregation, lacking online identification of actual leakage events and attribution analysis of gas changes, and cannot reliably distinguish whether changes in monitored quantities are caused by leakage, gas replenishment, temperature disturbances, or sensor drift. If a physical constraint artificial intelligence model can be introduced into the leakage inversion process, while simultaneously integrating gas mass balance relationships, equipment structural parameters, and digital twin status information, it is hoped that "bookkeeping-style carbon emission management" can be elevated to equipment-level green and safe operation diagnosis, providing a new technological foundation for the full life-cycle management of environmentally friendly gas current transformers. Summary of the Invention

[0006] The technical problem to be solved by this invention is to provide a condition assessment method, device and computer program product for environmentally friendly gas-insulated current transformers, so as to improve the accuracy of identifying minute leaks, realize traceable carbon emission accounting, and take into account both equipment insulation safety and low-carbon operation and maintenance.

[0007] To address the aforementioned technical problems, this invention provides a condition assessment method for environmentally friendly gas-insulated current transformers, comprising: Step S1: Obtain the gas chamber structure parameters, insulating gas property parameters, gas chamber status monitoring data, and operating condition data of the target environmentally friendly gas-insulated current transformer. Step S2: Based on the gas chamber structural parameters, insulating gas property parameters and gas chamber status monitoring data, calculate the gas mass in the gas chamber based on the gas state equation, and determine the initial leakage amount by combining the operation and maintenance event records. Step S3: Input the initial leakage amount, gas chamber temperature parameter and operating condition data into the preset physical constraint artificial intelligence model, correct the initial leakage amount under the preset physical constraint conditions, and output the leakage amount correction value. Step S4: Calculate the carbon emission equivalent based on the leakage correction value, the composition of the insulating gas, and the corresponding carbon emission conversion factor; Step S5: Based on the leakage correction value and the gas density change trend, determine the gas chamber insulation margin assessment result and output risk warning information.

[0008] Preferably, step S1 specifically includes: The effective volume of the gas chamber, sealing structure parameters, and rated operating parameters are obtained as the gas chamber structure parameters, and the gas component ratio, mixed gas molar mass, gas compressibility factor calculation parameters, and carbon emission conversion factor are obtained as the insulating gas property parameters. The gas chamber pressure, gas temperature, gas density, trace moisture content, and partial discharge characteristic quantities are collected as the gas chamber status monitoring data. The primary side load current, ambient temperature and humidity, gas circuit valve status, gas supply / venting records, maintenance records, and sensor calibration records are collected as the operating condition data. Timestamp alignment, outlier removal, and dimension normalization are performed on multi-source heterogeneous data to generate standardized feature sequences with uniform time resolution.

[0009] Preferably, step S2 specifically includes: Based on the real gas state equation, combined with the gas chamber volume, the molar mass of the mixed gas, the real-time gas pressure, the real-time gas temperature and the gas compressibility factor under the corresponding operating conditions, the total mass of the insulating gas in the current gas chamber is calculated. Select a sliding time window of a preset duration, calculate the change in gas mass within the window, deduct the increase in gas mass introduced by the gas replenishment operation and the decrease in gas mass discharged by the gas release and maintenance operation within the window, obtain the initial leakage rate, and perform non-negative truncation on the initial leakage rate, and integrate to obtain the initial cumulative leakage amount.

[0010] Preferably, the physical constraint artificial intelligence model in step S3 adopts a hierarchical modular architecture, which includes, from bottom to top, a mass balance layer, a disturbance identification layer, a drift identification layer and a leakage inversion layer. The mass balance layer is used to output an interpretable initial leakage baseline, the disturbance identification layer is used to distinguish temperature disturbances from actual leakage and generate a temperature correction term, the drift identification layer is used to identify sensor drift interference and generate a drift correction amount, and the leakage inversion layer is used to combine the output results of the aforementioned layers to output the corrected leakage amount and leakage identification result.

[0011] Preferably, during the training and inference process of the physical constraint artificial intelligence model, the preset physical constraint conditions include gas mass conservation constraint, leakage non-negativity constraint, operation and maintenance event consistency constraint, sensor drift smoothing constraint, and density safety boundary constraint. Each constraint condition is assigned an independent loss term, which is then weighted by preset weight coefficients to form the total loss function, thus forcing the model output to conform to physical laws.

[0012] Preferably, step S4 specifically includes: The leakage mass of each gas component is calculated based on the corrected cumulative leakage amount and the mass ratio of each component of the insulating gas. By combining the global warming potential coefficients of each component, the carbon dioxide equivalent emissions are obtained by weighted summation. It synchronously records the accounting formula, conversion factor version, original data source and model version information, and supports multi-dimensional carbon emission statistics for single leakage events, single devices, substations, and annually.

[0013] Preferably, step S5 specifically includes: Based on the current gas pressure, temperature, moisture content and gas composition parameters, the corrected breakdown field strength of the critical area of ​​the gas chamber is calculated, and the ratio of the corrected breakdown field strength to the local maximum electric field strength is used as the gas chamber insulation margin. By integrating the actual leakage probability, normalized leakage rate, normalized carbon emission equivalent, insulation margin deficit, and sensor drift probability, a comprehensive risk index in the range of 0 to 1 is obtained through weighted calculation and normalized mapping. Based on the current leakage rate and state change trend, predict the remaining time before the gas density or insulation margin first reaches the safety threshold, and output graded early warning information in combination with the comprehensive risk index.

[0014] Preferably, the method further includes: Step S6: Based on the identified leakage events, carbon emission equivalent calculation results, original monitoring data, operation and maintenance operation vouchers, model correction records and risk levels, generate a traceable digital ledger with a complete chain of evidence and output corresponding operation and maintenance handling suggestions; write back the leakage rate, sealing structure health, sensor drift parameters, carbon emission data and on-site handling feedback results to the current transformer digital twin model, and iteratively update the leakage risk threshold and artificial intelligence model parameters of similar equipment.

[0015] This invention also provides a condition assessment device for an environmentally friendly gas-insulated current transformer, comprising: One or more processors; Memory; One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, and the one or more applications are configured to perform the condition assessment method for the environmentally friendly gas-insulated current transformer.

[0016] The present invention also provides a computer program product, including computer instructions, which instruct a computer device to perform operations corresponding to the condition assessment method for the environmentally friendly gas-insulated current transformer.

[0017] The beneficial effects of this invention are as follows: Addressing the operational management needs of environmentally friendly gas-insulated current transformers, this invention effectively decouples multi-source interference such as temperature fluctuations, load changes, and sensor drift by combining gas mass balance calculations with a physical constraint artificial intelligence model, significantly improving the accuracy of identifying minute leaks and reducing false alarms and missed detections. Based on the corrected leakage amount, it automatically performs carbon equivalent calculations, linking original monitoring data, gas component parameters, and operation and maintenance vouchers to form a complete and traceable chain of accounting evidence, reducing the risks of errors, omissions, and subsequent re-entry in manual ledgers. This invention deeply links leakage status, carbon emission calculations, and insulation margin assessments, enabling early prediction of sealing anomalies and insulation performance degradation trends, achieving integrated management of green and low-carbon management and equipment safety operation and maintenance. It supports multi-dimensional carbon equivalent aggregation at the equipment level, bay level, station level, and annual level, providing a reliable data foundation for environmental equipment operation evaluation and carbon verification. Simultaneously, through digital twin closed-loop data accumulation and handling experience, it continuously iterates and optimizes the model, improving the consistency and accuracy of status assessments and operation and maintenance decisions. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating a condition assessment method for an environmentally friendly gas-insulated current transformer according to Embodiment 1 of the present invention.

[0020] Figure 2 This is a schematic diagram of the leakage inversion and disturbance removal process in an embodiment of the present invention.

[0021] Figure 3 This is a schematic diagram of the reasoning process of the physical constraint AI model in an embodiment of the present invention. Detailed Implementation

[0022] The following description of the embodiments is taken with reference to the accompanying drawings, which illustrate specific embodiments in which the invention can be implemented.

[0023] Please refer to Figure 1 As shown, Embodiment 1 of the present invention provides a condition assessment method for an environmentally friendly gas-insulated current transformer, comprising: Step S1: Obtain the gas chamber structure parameters, insulating gas property parameters, gas chamber status monitoring data, and operating condition data of the target environmentally friendly gas-insulated current transformer. In step S2, the gas mass in the gas chamber is calculated based on the gas state equation according to the gas chamber structure parameters, insulating gas property parameters and gas chamber state monitoring data, and the initial leakage amount is determined in combination with the operation and maintenance event record; In step S3, the initial leakage amount, gas chamber temperature parameter and operating condition data are input into a preset physical constraint artificial intelligence model, the initial leakage amount is corrected under preset physical constraint conditions, and a corrected leakage amount value is output; In step S4, the carbon emission equivalent is calculated according to the corrected leakage amount value, the components of the insulating gas and the corresponding carbon emission conversion coefficients; In step S5, the gas chamber insulation margin assessment result is determined according to the corrected leakage amount value and the gas density change trend, and risk warning information is output.

[0024] The state assessment method for environment-friendly gas-insulated current transformer described in this embodiment is applied to the ultra-high voltage substation scenario with a voltage level of 500kV and above, and is executed by a station-level edge computing device, a background monitoring system or a cloud data processing device, which can realize integrated management and control of accurate leakage identification, dynamic carbon equivalent accounting and linked insulation risk assessment.

[0025] The specific implementation of each step is described in detail below: Step S1 is used for parameter acquisition and multi-source data standardization processing, that is, acquiring basic physical parameters and full-dimensional operation data of the target device to provide an input benchmark for subsequent calculation, and the specific implementation process is as follows: (1) Gas chamber structure parameters: extracted from the device nameplate, factory design drawings and type test reports, specifically including the effective volume of the gas chamber, rated inflation pressure, sealing ring structure and layout parameters, positions of the gas replenishment valve and gas sampling port, installation layout points of pressure / temperature / density sensors, structural parameters of insulating supports, dimensions of key areas at the gas-solid interface, and factory insulation withstand test parameters. This type of parameters constitutes the structural benchmark of the gas chamber physical model.

[0026] (2) Insulating gas property parameters: specifically including the type of insulating gas, volume proportion and molar mass of each component of the mixed gas, calculation model parameters of gas compression factor, global warming potential (GWP) coefficient corresponding to each component, that is, the carbon emission conversion coefficient, and gas density parameters under standard conditions.

[0027] (3) Gas chamber state monitoring data: collected by the online monitoring device supporting the equipment, specifically including real-time state parameters such as internal gas pressure, gas temperature, gas density, micro-water content, gas component concentration, and partial discharge characteristic quantity of the gas chamber.

[0028] (4) Operating condition data, specifically including primary side load current of current transformer, ambient temperature and humidity, opening and closing status of gas valve, historical calibration records of sensors, equipment maintenance records, gas replenishment and venting operation records, and manual inspection records.

[0029] For data from different sources, with inconsistent sampling frequencies and formats, timestamp alignment, outlier removal, and dimension normalization are performed to convert them into standardized feature vectors with a unified time resolution, ensuring the consistency of input for subsequent calculations.

[0030] The following table illustrates the multi-source state data and standardized feature vectors:

[0031] Step S2 calculates the initial leakage rate based on the gas law equation. This involves calculating the gas mass within the chamber and deriving the initial leakage rate based on the real gas law equation and the principle of mass conservation. This initial leakage rate serves as the baseline for subsequent intelligent corrections. Please refer to... Figure 2 As shown, the specific explanation is as follows: Using the real gas law, combined with the gas chamber structural parameters, insulating gas property parameters, and real-time monitoring data, the total mass of insulating gas in the current gas chamber is calculated using the following formula: (1) in, for The total mass of insulating gas in the gas chamber at any given time; for The absolute pressure of the gas in the chamber at any given time; This is the equivalent volume of the air chamber; The average molar mass of the mixed insulating gas; for The gas compressibility factor under the corresponding operating conditions is dimensionless and is used to characterize the deviation between the compressibility characteristics of real gases and ideal gases. It changes dynamically with pressure, temperature, and gas composition. This is the universal gas constant; for The absolute thermodynamic temperature of the gas in the chamber at any given time.

[0032] Select a sliding time window of a preset duration, with a time step of 1. The difference in total gas mass between adjacent time points within the window is calculated. Simultaneously, maintenance event records within the time window are matched, and the gas mass increment introduced by gas replenishment operations and the gas mass reduction caused by venting operations or maintenance exhaust are deducted to obtain the net change in gas mass within the window. The initial leakage rate is then calculated. : (2) in, For the previous moment ( The total mass of the gas in the chamber at (time). For time step The mass of gas introduced during internal gas replenishment. For time step The quality of gas discharged during internal venting and maintenance exhaust operations; It is a non-negative truncation operator that ensures the initial leakage rate is not less than 0 and filters out negative leakage results that have no physical meaning.

[0033] The initial cumulative leakage amount for the corresponding time period can be calculated by further integration based on the initial leakage rate.

[0034] Step S3 corrects the leakage of the physical constraint artificial intelligence model.

[0035] This invention employs an artificial intelligence model with built-in physical constraints to decouple environmental disturbances, sensor drift, and actual leakage, thereby correcting the initial leakage amount. Specific details are as follows: The standardized data, including gas chamber state parameters, operating condition parameters, initial leakage obtained in step S2, maintenance event parameters, and sensor state parameters, are integrated into a single dataset. Input feature vector at time step Its composition is as follows: (3) in, The density of the gas in the gas chamber. The moisture content within the air chamber, This refers to the primary load current of the current transformer. The ambient temperature at the site. The humidity of the site environment, This refers to the open / closed state of the gas circuit valve. To calibrate the sensor residual, This is a partial discharge characteristic quantity; other optional operating condition characteristics can be flexibly expanded according to equipment type and monitoring configuration.

[0036] Standardize the feature vector With equipment structural parameter set Input a pre-defined physical constraint AI model. This physical constraint AI model adopts a layered and modular architecture, with a physical mechanism layer as the foundation, a data-driven layer for correction, and a risk fusion layer as the output. From bottom to top, it includes a mass balance layer, a disturbance identification layer, a drift identification layer, a leakage inversion layer, and a risk fusion layer. The input-output relationships and technical functions of each layer are shown in the table below:

[0037] Based on the aforementioned layered architecture, the model uses a set of trainable parameters. Construct an end-to-end inference network, denoted as After multi-level cascaded reasoning, the final leak identification and correction results are output. The overall reasoning process is as follows: Figure 3 As shown, it can be expressed as: (4) in, This represents the actual probability of a leak occurring, with a value ranging from 0 to 1. This represents the probability of sensor drift occurring, with a value ranging from 0 to 1. This is the actual leakage rate after model correction; Classification labels for suspected leak locations; The confidence level of the leak identification results.

[0038] During training and inference, the physical constraint AI model is forced to meet preset physical constraints. These physical constraints are implemented through a weighted loss function, and the model's total loss function is defined as follows: (5) in, This is the total loss function of the model; This is the data fitting loss term, which calculates the error between the predicted value and the true value based on the labeled samples. These are the weighting coefficients for each physical constraint term, used to balance the accuracy of data fitting with the strength of physical constraints; This is a gas mass conservation constraint loss term, constraining the mass balance relationship between gas mass changes and leakage, operation and maintenance; The non-negative constraint loss term for leakage amount penalizes abnormal outputs with negative leakage rate; This is a consistency constraint loss item for operation and maintenance events, which constrains the quality changes during gas replenishment and venting periods to match the operation and maintenance records. The loss term is a sensor drift smoothing constraint, which constrains the sensor drift to be a continuous and slow change process, and prohibits step-like sudden changes. The density safety boundary constraint loss term ensures that the calculated density of the constrained gas does not exceed the physically reasonable range.

[0039] By limiting the physical constraints mentioned above, the model can effectively decouple interference factors such as temperature fluctuations, load changes, and sensor drift from the actual leakage, and output the corrected leakage rate and cumulative leakage amount (i.e., the leakage amount correction value), while also outputting the suspected leakage location and the corresponding confidence level, thus achieving effective decoupling from interference factors such as temperature fluctuations, load changes, and sensor drift.

[0040] Step S4 involves dynamic carbon emission equivalent calculation, which is based on the corrected leakage amount to conduct traceable carbon equivalent calculation, replacing the traditional manual accounting method. Details are as follows: Based on the cumulative leakage correction value output in step S3, combined with the component ratio of the insulating gas and the carbon emission conversion factor corresponding to each component, the carbon dioxide equivalent emission of the corresponding leaked gas is calculated. The calculation formula is as follows: (6) in, Carbon dioxide equivalent emissions; The component numbers are the part numbers of the mixed insulating gas. For the first The leakage mass of the gas components; For the first Carbon emission conversion factor (i.e., global warming potential GWP value) for each gas component. This is the corrected total mass of leaked gas. For the first The mass percentage of each gaseous component.

[0041] In practice, carbon emissions from a single leakage event, cumulative carbon emissions from a single device, total carbon emissions from the entire substation, and annual carbon emissions can be calculated according to the required management granularity. At the same time, the calculation formulas, conversion factor versions, original data sources, and model versions used in the accounting are recorded to ensure that the accounting results have a complete chain of evidence and are traceable and verifiable.

[0042] Step S5 involves insulation margin assessment and risk warning, which links leakage status with insulation safety to achieve a quantitative assessment from leakage amount to insulation risk. The specific details are as follows: First, the insulation margin of the critical area of ​​the gas chamber is calculated based on the current gas state parameters of the gas chamber, which characterizes the safety reserve level of the current insulation withstand strength relative to the operating field strength. The insulation margin calculation formula is as follows: (7) in, for The insulation margin of the air cell at any given time is dimensionless; the larger the value, the higher the insulation safety reserve. The breakdown field strength of the gas-solid interface is corrected for gas pressure, temperature, trace moisture content, and gas composition. The gas composition-related correction parameters characterize the effect of different mixing ratios on the breakdown field strength. The maximum local electric field intensity in key areas within the gas chamber (such as the electrode edges, the surface of insulating support components, and other areas where the electric field is concentrated) is determined by the equipment structural parameters and operating voltage.

[0043] Secondly, by integrating multiple dimensions of indicators such as leakage probability, leakage rate, carbon emissions, insulation margin, and sensor drift probability, a comprehensive risk index is calculated. The calculation formula is as follows: (8) in, This is a comprehensive risk index, with a value ranging from 0 to 1. The higher the value, the higher the risk level. The Sigmoid activation function is used to normalize the linearly weighted result to the 0~1 range; These are the regression coefficients of the risk assessment model, obtained through sample training or engineering calibration; As a normalization function, it maps physical quantities with different dimensions to a unified numerical range, eliminating the influence of dimensional differences; This represents the insulation margin deficit; a larger value indicates a lower insulation safety reserve and a higher insulation risk.

[0044] Furthermore, based on the current leakage rate and the trend of gas state changes, the remaining time before the gas density or insulation margin first reaches the safety threshold is predicted, providing a quantitative basis in the time dimension for operation and maintenance decisions. The calculation expression is as follows: (9) in, The remaining time expected to reach the safety threshold; For the length of future time; For prediction Gas density in the chamber after a certain time; The minimum permissible gas density threshold for the gas chamber is determined by equipment standards and operating procedures; The minimum permissible insulation margin threshold; The OR operator indicates that the threshold is triggered if any one of the conditions is met. This indicates taking the minimum value that satisfies the constraints. The value is the time when the threshold is first reached.

[0045] Finally, by combining the severity of the leakage, the level of carbon emissions, and the insulation safety margin, a comprehensive risk warning result is output, which includes the leakage risk level, the carbon emission risk level, and the insulation risk level, providing a quantitative basis for operation and maintenance decisions.

[0046] As a further optimized implementation, the following steps may be included after step S5: Credible Digital Ledger Generation: Based on identified leak events, carbon equivalent calculation results, original sensor monitoring data, gas replenishment and maintenance operation vouchers, model correction records, and risk level information, a traceable digital ledger with a complete chain of evidence is generated, and operation and maintenance handling suggestions such as gas replenishment operation, leak point detection, sealing structure inspection, sensor calibration, trend tracking, power outage maintenance, or ledger review are output.

[0047] Digital twin closed-loop update: The calculated leakage rate, sealing structure health, sensor drift parameters, carbon equivalent emissions, on-site handling feedback and verification results are written back to the digital twin model of the current transformer, and the leakage risk threshold, artificial intelligence model parameters and sealing structure health evaluation standards of similar equipment are iteratively updated to achieve continuous self-optimization of the condition assessment model.

[0048] The following three typical examples illustrate the implementation effects of the present invention, corresponding to the identification of minute real leaks, the calculation of carbon equivalent of gas replenishment events, and the elimination of sensor drift false alarms, respectively.

[0049] Example 1: Identification of minute real leaks in a 500kV environmentally friendly gas-insulated current transformer.

[0050] A 500kV environmentally friendly gas-insulated current transformer uses clean air or a nitrogen-oxygen mixture as the insulating medium. The equivalent volume of the gas chamber is approximately 1.2 m³, and the rated absolute pressure is approximately 0.65 MPa. It is equipped with sensors for pressure, temperature, density, moisture content, valve status, and ambient temperature and humidity. Continuous monitoring over 30 days showed that the gas density decreased slowly after temperature compensation, with the daily average decrease rate increasing from the historical baseline of 0.02% / d to 0.11% / d. The pressure curve exhibited a periodic coupling with diurnal temperature variation, but a monotonically decreasing residual remained after removing temperature and load disturbances.

[0051] First, the equipment nameplate, chamber volume, rated pressure, sealing structure, and sensor placement are read to construct a digital model of the equipment chamber. Then, pressure, temperature, density, moisture content, load, and environmental data are input, and timestamp alignment is completed. The initial leakage rate is calculated according to equations (1) and (2). By further eliminating diurnal temperature fluctuations, load heating effects, and sensor zero-point offset through a physically constrained AI model, the true leakage probability is output. ≈0.87, sensor drift probability ≈0.09, corrected leakage rate The cumulative leakage is approximately 0.035 kg / d, with a cumulative leakage of about 1.05 kg. If the mass percentage of the low-GWP component in this gas mixture is... Then, calculate according to formula (6). The insulation margin model predicts that at the current leakage rate, the gas density will approach the lower warning limit in approximately 90 days. The system will ultimately generate a digital ledger of "suspected micro-leaks," recommending the scheduling of live-line leak detection, valve seal verification, and gas replenishment.

[0052]

[0053] Example 2: Automatic calculation and ledger generation of carbon equivalent under gas replenishment events.

[0054] During annual maintenance, a 500kV environmental gas CT underwent a gas replenishment operation. Before and after the replenishment, the pressure and density of the gas chamber experienced a step change, and on-site manual records showed a replenishment mass of approximately 3.0 kg. Traditional logs only record the replenishment mass and cannot determine whether the replenishment was to compensate for historical leaks, for maintenance venting, or for sensor deviation. In this embodiment, the system automatically identifies the replenishment event based on the opening and closing status of the replenishment valve, the weighing record of the replenishment cylinder, the pressure step change in the gas chamber, and the sensor curve, and separately calculates the continuous leakage before replenishment from the replenishment operation. The system identifies the opening time of the replenishment valve, the weight change of the replenishment cylinder, and the density step change. The replenishment mass is then included in the mass balance. Formula (2) is included to avoid misjudging the increase in mass caused by gas replenishment as a negative leak. The physical constraint AI model segments the gas replenishment event into "leakage section before gas replenishment, gas replenishment operation section, and stabilization section after gas replenishment". The system calculates the carbon equivalent based on the actual leakage mass rather than the gas replenishment mass; if the gas replenishment is used to compensate for historical leaks, the leakage event is associated with the gas replenishment event; if the gas replenishment is used for maintenance replacement, a separate maintenance event is generated. The ledger automatically records the cylinder number, gas replenishment mass, curve evidence, model version, conversion factor version, and manual confirmation status.

[0055]

[0056] Example 3: Removal of false leaks caused by sensor drift.

[0057] The pressure sensor in the gas chamber of an environmental protection gas current transformer showed a slow decline for 15 consecutive days, but the density relay, gas temperature, and moisture data did not show consistent changes, and adjacent similar equipment under the same conditions did not exhibit a similar trend. System access to the sensor calibration record revealed that the pressure sensor already showed a zero-point offset trend during the last calibration. A preliminary mass balance model was provided. A suspected leak result >0. The physical constraint AI model simultaneously inputs pressure sensor residuals, density sensor curves, temperature curves, and comparison curves from similar equipment to output the true leak probability. ≈0.18, sensor drift probability ≈0.76. The insulation margin did not show a decreasing trend, and the risk level is low. The system does not generate actual carbon emission leakage events, only "sensor drift verification events." It is recommended to conduct pressure sensor calibration and data quality verification. After calibration, the drift parameters should be written back to the digital twin model to correct the sensor's confidence weight.

[0058]

[0059] As can be seen from the above examples, this invention is based on the physical model of the environmentally friendly gas CT chamber and uses physically constrained AI to invert leakage, unifying leakage events, carbon equivalent calculation, insulation margin, and digital twin closed loop. For real micro-leak scenarios, this invention can detect leakage trends in advance; for gas replenishment scenarios, this invention can distinguish between leakage and operational events; for sensor drift scenarios, this invention can avoid false alarms and erroneous carbon emission records.

[0060] Corresponding to the condition assessment method for an environmentally friendly gas-insulated current transformer in Embodiment 1 of the present invention, Embodiment 2 of the present invention also provides a condition assessment device for an environmentally friendly gas-insulated current transformer, comprising: One or more processors; Memory; One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, and the one or more applications are configured to perform the condition assessment method for the environmentally friendly gas-insulated current transformer.

[0061] Corresponding to the condition assessment method for an environmentally friendly gas-insulated current transformer in Embodiment 1 of the present invention, Embodiment 3 of the present invention also provides a computer program product, including computer instructions, which instruct a computer device to perform the operation corresponding to the condition assessment method for the environmentally friendly gas-insulated current transformer.

[0062] Preferably, the processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor, or the processor can be any conventional processor. The processor is the control center of the device, connecting various parts of the device through various interfaces and lines.

[0063] The memory mainly includes a program storage area and a data storage area. The program storage area can store the operating system, applications required for at least one function, etc., while the data storage area can store related data, etc. Furthermore, the memory can be a high-speed random access memory, or a non-volatile memory, such as a plug-in hard drive, a SmartMedia Card (SMC), a Secure Digital (SD) card, and a Flash Card, or other volatile solid-state storage devices.

[0064] It should be noted that the above-mentioned devices may include, but are not limited to, processors and memory, as will be understood by those skilled in the art.

[0065] For the working principle and process of the above embodiments, please refer to the description of Embodiment 1 of the present invention, which will not be repeated here.

[0066] Compared with existing technologies, this invention has the following significant advantages: Addressing the operational management needs of environmentally friendly gas-insulated current transformers, this invention effectively decouples multi-source interference such as temperature fluctuations, load changes, and sensor drift by combining gas mass balance calculations with a physical constraint artificial intelligence model, significantly improving the accuracy of identifying minute leaks and reducing false alarms and missed detections. Based on the corrected leakage amount, it automatically performs carbon equivalent calculations, linking original monitoring data, gas component parameters, and operation and maintenance vouchers to form a complete and traceable chain of accounting evidence, reducing the risks of errors, omissions, and subsequent re-entry in manual ledgers. This invention deeply links leakage status, carbon emission accounting, and insulation margin assessment, enabling early prediction of sealing anomalies and insulation performance degradation trends, achieving integrated management of green and low-carbon management and equipment safety operation and maintenance. It supports multi-dimensional carbon equivalent aggregation at the equipment level, bay level, station level, and annual level, providing a reliable data foundation for environmental equipment operation evaluation and carbon verification. Simultaneously, through digital twin closed-loop data accumulation and handling experience, it continuously iterates and optimizes the model, improving the consistency and accuracy of status assessment and operation and maintenance decisions.

[0067] The above description is merely a preferred embodiment of the present invention and should not be construed as limiting the scope of the invention. Therefore, any equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A condition assessment method for an environmentally friendly gas-insulated current transformer, characterized in that, include: Step S1: Obtain the gas chamber structure parameters, insulating gas property parameters, gas chamber status monitoring data, and operating condition data of the target environmentally friendly gas-insulated current transformer. Step S2: Based on the gas chamber structural parameters, insulating gas property parameters and gas chamber status monitoring data, calculate the gas mass in the gas chamber based on the gas state equation, and determine the initial leakage amount by combining the operation and maintenance event records. Step S3: Input the initial leakage amount, gas chamber temperature parameter and operating condition data into the preset physical constraint artificial intelligence model, correct the initial leakage amount under the preset physical constraint conditions, and output the leakage amount correction value. Step S4: Calculate the carbon emission equivalent based on the leakage correction value, the composition of the insulating gas, and the corresponding carbon emission conversion factor; Step S5: Based on the leakage correction value and the gas density change trend, determine the gas chamber insulation margin assessment result and output risk warning information.

2. The method according to claim 1, characterized in that, Step S1 specifically includes: The effective volume of the gas chamber, sealing structure parameters, and rated operating parameters are obtained as the gas chamber structure parameters, and the gas component ratio, mixed gas molar mass, gas compressibility factor calculation parameters, and carbon emission conversion factor are obtained as the insulating gas property parameters. The gas chamber pressure, gas temperature, gas density, trace moisture content, and partial discharge characteristic quantities are collected as the gas chamber status monitoring data. The primary side load current, ambient temperature and humidity, gas circuit valve status, gas supply / venting records, maintenance records, and sensor calibration records are collected as the operating condition data. Timestamp alignment, outlier removal, and dimension normalization are performed on multi-source heterogeneous data to generate standardized feature sequences with uniform time resolution.

3. The method according to claim 1, characterized in that, Step S2 specifically includes: Based on the real gas law, combined with the gas chamber volume, the molar mass of the mixed gas, the real-time gas pressure, the real-time gas temperature and the gas compressibility factor under the corresponding operating conditions, the total mass of the insulating gas in the current gas chamber is calculated. Select a sliding time window of a preset duration, calculate the change in gas mass within the window, deduct the increase in gas mass introduced by the gas replenishment operation and the decrease in gas mass discharged by the gas release and maintenance operation within the window, obtain the initial leakage rate, and perform non-negative truncation on the initial leakage rate, and integrate to obtain the initial cumulative leakage amount.

4. The method according to claim 1, characterized in that, The physical constraint artificial intelligence model in step S3 adopts a hierarchical modular architecture, which includes, from bottom to top, a mass balance layer, a disturbance identification layer, a drift identification layer and a leakage inversion layer. The mass balance layer is used to output an interpretable initial leakage baseline, the disturbance identification layer is used to distinguish temperature disturbances from actual leakage and generate a temperature correction term, the drift identification layer is used to identify sensor drift interference and generate a drift correction amount, and the leakage inversion layer is used to combine the output results of the aforementioned layers to output the corrected leakage amount and leakage identification result.

5. The method according to claim 4, characterized in that, During the training and inference process of the physical constraint artificial intelligence model, the preset physical constraints include gas mass conservation constraints, non-negativity of leakage, consistency of operation and maintenance events, sensor drift smoothing constraints, and density safety boundary constraints. Each constraint condition is assigned an independent loss term, which is then weighted by preset weight coefficients to form the total loss function, thus forcing the model output to conform to physical laws.

6. The method according to claim 1, characterized in that, Step S4 specifically includes: The leakage mass of each gas component is calculated based on the corrected cumulative leakage amount and the mass ratio of each component of the insulating gas. By combining the global warming potential coefficients of each component, the carbon dioxide equivalent emissions are obtained by weighted summation. It synchronously records the accounting formula, conversion factor version, original data source and model version information, and supports multi-dimensional carbon emission statistics for single leakage events, single devices, substations, and annually.

7. The method according to claim 1, characterized in that, Step S5 specifically includes: Based on the current gas pressure, temperature, moisture content and gas composition parameters, the corrected breakdown field strength of the critical area of ​​the gas chamber is calculated, and the ratio of the corrected breakdown field strength to the local maximum electric field strength is used as the gas chamber insulation margin. By integrating the actual leakage probability, normalized leakage rate, normalized carbon emission equivalent, insulation margin deficit, and sensor drift probability, a comprehensive risk index in the range of 0 to 1 is obtained through weighted calculation and normalized mapping. Based on the current leakage rate and state change trend, predict the remaining time before the gas density or insulation margin first reaches the safety threshold, and output graded early warning information in combination with the comprehensive risk index.

8. The method according to claim 7, characterized in that, Also includes: Step S6: Based on the identified leakage events, carbon emission equivalent calculation results, original monitoring data, operation and maintenance operation vouchers, model correction records and risk levels, generate a traceable digital ledger with a complete chain of evidence and output corresponding operation and maintenance handling suggestions; write back the leakage rate, sealing structure health, sensor drift parameters, carbon emission data and on-site handling feedback results to the current transformer digital twin model, and iteratively update the leakage risk threshold and artificial intelligence model parameters of similar equipment.

9. A condition assessment device for an environmentally friendly gas-insulated current transformer, characterized in that, include: One or more processors; Memory; One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications being configured to perform the condition assessment method for an environmentally friendly gas-insulated current transformer as described in any one of claims 1 to 8.

10. A computer program product, characterized in that, Includes computer instructions that instruct computer equipment to perform operations corresponding to the condition assessment method for the environmentally friendly gas-insulated current transformer as described in any one of claims 1 to 8.

Images