Artificial Intelligence-Based Fault Monitoring System and Device for Dry-Type Transformers

Abstract

This invention provides an artificial intelligence-based dry-type transformer fault monitoring system and device, belonging to the field of dry-type transformer fault monitoring technology. It includes a real-time concentration measurement module, an airflow calculation module, a true concentration calculation module, a fault judgment module, and a fault analysis module. The real-time concentration measurement module acquires measured concentration data; the airflow calculation module calculates airflow data; the true concentration calculation module calculates true concentration data; the fault judgment module determines whether a fault exists in the dry-type transformer; and the fault analysis module acquires fault data. This invention improves the accuracy and reliability of fault monitoring by considering the impact of airflow caused by heat dissipation on the concentration of gas components inside the dry-type transformer and calculating the true concentration of internal gas components. This more accurately reflects the actual internal condition of the transformer and reduces the possibility of misjudgment and missed judgment.

Description

Technical Field

[0001] This invention relates to the field of dry-type transformer fault monitoring technology, and in particular to an artificial intelligence-based dry-type transformer fault monitoring system and an artificial intelligence-based dry-type transformer fault monitoring device. Background Technology

[0002] Dry-type transformers are important pieces of equipment widely used in power systems, mainly composed of a core, windings, insulation materials, and a casing. The core is typically made of high-quality cold-rolled silicon steel sheets to reduce hysteresis and eddy current losses, improving transformer efficiency. The windings are generally made of copper or aluminum wire, with the number of turns and wire diameter chosen based on the transformer's capacity and voltage level. Insulation materials are a key component of dry-type transformers, commonly including epoxy resin, which effectively isolates the windings from the core, preventing short circuits and leakage. The casing protects internal components, provides heat dissipation, and offers protection. The primary function of a dry-type transformer is voltage transformation. It can convert high voltage to low voltage for use in various electrical equipment, such as motors and lighting equipment in industrial production; it can also boost low voltage to meet the needs of long-distance power transmission, reducing line losses. Furthermore, dry-type transformers also provide isolation and filtering, offering a stable voltage output and protecting electrical equipment from voltage fluctuations and interference. Compared to traditional oil-immersed transformers, dry-type transformers offer many advantages. First, it eliminates the need for insulating oil, avoiding the risks of oil leaks and fires, making it more environmentally friendly and safer. Second, dry-type transformers are relatively easy to install and maintain, have lower environmental requirements, and can adapt to various harsh working environments. Furthermore, they operate with less noise, minimizing their impact on the surrounding environment.

[0003] When a dry-type transformer experiences a fault, the concentration of its internal gas components changes significantly. In the case of an overheating fault, the insulation material decomposes at high temperatures, producing gases. The concentration of hydrogen typically increases, as the thermal decomposition of the insulation material can generate hydrogen even at lower temperatures. As the temperature rises further, the concentrations of methane and ethylene also gradually increase. In severe overheating, acetylene may even be detected, although its concentration is generally relatively low. In the case of a discharge fault, the gas composition differs. During partial discharge, the concentrations of hydrogen and methane may increase. In spark discharge faults, the concentration of hydrogen increases significantly, and the concentrations of carbon monoxide and carbon dioxide may also rise. In arc discharge faults, the concentration of acetylene increases significantly, often becoming a key indicator of such severe faults. Furthermore, the concentrations of ethylene and hydrogen also increase substantially. It is worth noting that the changes in gas concentration caused by different types and degrees of faults are not absolutely isolated; they may overlap and superimpose. For example, when long-term overheating and discharge faults coexist, the concentrations of various gases will exhibit complex changes.

[0004] Monitoring and analyzing the concentration data of gas components inside dry-type transformers can promptly detect potential faults and assess their type and severity. This allows for appropriate maintenance and repair measures to be taken, ensuring the safe and stable operation of the transformer and reducing losses and impacts caused by faults. However, during operation, dry-type transformers require heat dissipation. The resulting airflow can dilute the concentration of gas components inside the transformer, leading to inaccurate fault monitoring and increasing the risk of misdiagnosis and missed diagnoses. Summary of the Invention

[0005] This invention provides an artificial intelligence-based dry-type transformer fault monitoring system and device to solve the defects of existing dry-type transformer fault monitoring that are not accurate enough and are prone to misjudgment and missed judgment.

[0006] On one hand, the present invention provides an artificial intelligence-based dry-type transformer fault monitoring system, comprising:

[0007] The data acquisition module is used to acquire temperature data and parameter data of the dry-type transformer, and to collect gas samples inside the dry-type transformer in real time.

[0008] The real-time concentration measurement module is used to perform component analysis on gas samples using gas chromatography to obtain gas component concentration data.

[0009] The air circulation calculation module is used to calculate the total heat dissipation of the dry-type transformer within a preset time period, and to calculate the air circulation data inside the dry-type transformer based on the total heat dissipation.

[0010] The true concentration calculation module is used to calculate the true concentration data of the gas components inside the dry-type transformer by combining air circulation data and measured concentration data.

[0011] The fault detection module compares the actual concentration data with the preset normal concentration threshold range. If the actual concentration data exceeds the normal concentration threshold range, it is marked as fault concentration data, indicating that there is a fault in the dry-type transformer.

[0012] The fault analysis module is used to build a fault analysis model based on a convolutional neural network. It takes fault concentration data as input and outputs fault data of dry-type transformers. The fault data includes fault type, fault location and fault severity.

[0013] The fault reporting module is used to send fault data to maintenance personnel.

[0014] According to the present invention, an artificial intelligence-based fault monitoring system for dry-type transformers includes parameter data such as the volume of gas inside the dry-type transformer and the surface area of ​​the dry-type transformer. Temperature data includes the surface temperature of the dry-type transformer and the ambient temperature.

[0015] According to the artificial intelligence-based dry-type transformer fault monitoring system provided by the present invention, the process of performing component analysis on a gas sample includes:

[0016] Inject the gas sample into the gas chromatograph and select the appropriate chromatographic column for the gas components of the gas sample.

[0017] A thermal conductivity detector is used to detect the gas composition state data of a gas sample at the column outlet. The gas composition state data includes the retention time and peak area of ​​the gas components at the column outlet.

[0018] Based on the gas composition state data, obtain the gas composition concentration data of the gas sample.

[0019] According to the present invention, an artificial intelligence-based dry-type transformer fault monitoring system includes an air circulation calculation module comprising a heat dissipation calculation unit. This unit calculates the total heat dissipation by calculating both radiative and convective heat dissipation. The formula for calculating radiative heat dissipation is as follows:

[0020]

[0021] The formula for calculating heat dissipation through heat convection is expressed as follows:

[0022]

[0023] The formula for calculating the total heat dissipation is as follows:

[0024]

[0025] In the formula, Indicates the surface emissivity of a dry-type transformer. denoted as Stefan-Boltzmann constant, A as heat dissipation area, T1 as surface temperature of dry-type transformer, T0 as ambient temperature, and h as convective heat transfer coefficient.

[0026] According to the artificial intelligence-based dry-type transformer fault monitoring system provided by the present invention, the air circulation calculation module further includes an air flow calculation unit. The air flow calculation unit is used to calculate the air flow inside the dry-type transformer based on the total heat dissipation, expressed by the formula:

[0027]

[0028] In the formula, This represents the specific heat capacity of air at constant pressure. This indicates air density.

[0029] According to the artificial intelligence-based dry-type transformer fault monitoring system provided by the present invention, the calculation formula for the actual concentration data is expressed as follows:

[0030]

[0031] In the formula, This represents the measured concentration data of the i-th gas component, V represents the volume of the internal space of the dry-type transformer, and t represents the airflow rate inside the dry-type transformer. The time elapsed.

[0032] According to the artificial intelligence-based dry-type transformer fault monitoring system provided by the present invention, the process of setting the normal concentration threshold range includes:

[0033] Historical data on gas composition concentrations during normal operation of dry-type transformers were collected as sample data.

[0034] Calculate the average concentration data for each gas component. and standard deviation .

[0035] Based on the mean and standard deviation, a normal concentration threshold range for the cost of each gas is set, expressed by the following formula:

[0036]

[0037] In the formula, A i B i denoted by , where represents the normal concentration threshold range of the i-th gas component, and k represents the confidence coefficient.

[0038] According to the artificial intelligence-based dry-type transformer fault monitoring system provided by the present invention, the process of constructing a fault analysis model based on a convolutional neural network includes:

[0039] Collect historical fault case data, including gas composition concentration data, temperature data, and operating parameters of dry-type transformers.

[0040] Historical failure case data is standardized and labeled with failure type, failure location and failure severity to obtain labeled data.

[0041] A basic convolutional neural network model is constructed, consisting of convolutional layers, pooling layers, and fully connected layers. Convolutional layers extract features from the labeled data, yielding feature data. Pooling layers reduce the feature dimensionality of the feature data, resulting in sharper features. Fully connected layers flatten these sharp features and output fault data.

[0042] The base model is trained using labeled data, and its parameters are iteratively updated.

[0043] The accuracy of the output fault data is calculated, and the basic model that achieves the preset accuracy is used as the fault analysis model.

[0044] According to the present invention, a fault monitoring system for dry-type transformers based on artificial intelligence includes a fault reporting module comprising a fault receiving unit, a report generation unit, and a report sending unit. The fault receiving unit receives fault data. The report generation unit performs structured processing on the fault data to obtain a fault report. The report sending unit sends the fault report to relevant maintenance personnel.

[0045] On the other hand, the present invention also provides an artificial intelligence-based dry-type transformer fault monitoring device. The fault monitoring device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The functional modules in the fault monitoring device are deployed in the manner described above for an artificial intelligence-based dry-type transformer fault monitoring system. When the processor executes the computer program, it achieves fault monitoring of the dry-type transformer, improving the accuracy and reliability of fault monitoring.

[0046] The artificial intelligence-based dry-type transformer fault monitoring system and device provided by this invention, by considering the impact of airflow caused by heat dissipation on the concentration of internal gas components in the dry-type transformer and calculating the true concentration of internal gas components, brings many significant benefits to the fault monitoring of dry-type transformers. In the early stages of a transformer fault, internal chemical reactions and physical changes lead to the generation and concentration changes of specific gas components. By accurately calculating the true concentration, these subtle changes can be captured before the fault develops to a severe stage, gaining valuable time for timely maintenance and protection measures and effectively avoiding power outages and equipment damage caused by sudden faults. Secondly, it improves the accuracy and reliability of fault monitoring. Traditional monitoring methods may be affected by environmental factors, measurement errors, etc., while by comprehensively considering the dynamic changes in airflow and gas component concentration, some external interference factors can be eliminated, more accurately reflecting the actual internal condition of the transformer and reducing the possibility of misjudgment and missed judgment. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in this 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 some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0048] Figure 1This is a schematic diagram of the structure of the artificial intelligence-based dry-type transformer fault monitoring system provided in an embodiment of the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0050] The following is combined Figure 1 This invention describes an artificial intelligence-based dry-type transformer fault monitoring system and device.

[0051] Figure 1 This is a schematic diagram of the structure of the artificial intelligence-based dry-type transformer fault monitoring system provided in an embodiment of the present invention.

[0052] like Figure 1 As shown in the embodiment of the present invention, the dry-type transformer fault monitoring system and device based on artificial intelligence can be executed by an artificial intelligence-based dry-type transformer fault monitoring system, including a data acquisition module, a real-time concentration measurement module, an air circulation calculation module, a true concentration calculation module, a fault judgment module, a fault analysis module, and a fault reporting module.

[0053] The acquisition module is used to obtain temperature data and parameter data of the dry-type transformer, and to collect gas samples inside the dry-type transformer in real time.

[0054] The parameter data includes the volume of gas inside the dry-type transformer and the surface area of ​​the dry-type transformer. The temperature data includes the surface temperature of the dry-type transformer and the ambient temperature.

[0055] In this embodiment, the acquisition module can comprehensively and in real-time acquire key data of the dry-type transformer. By acquiring temperature and parameter data, it provides basic information for subsequent analysis. Real-time acquisition of internal gas samples helps to promptly detect potential fault signs, enabling early warning and thus preventing further deterioration of the fault, reducing maintenance costs and potential power outage losses.

[0056] The real-time concentration measurement module is used to perform component analysis on gas samples using gas chromatography to obtain gas component concentration data.

[0057] The process of performing compositional analysis on a gas sample includes:

[0058] Inject the gas sample into the gas chromatograph and select the appropriate chromatographic column for the gas components of the gas sample.

[0059] A thermal conductivity detector is used to detect the gas composition state data of a gas sample at the column outlet. The gas composition state data includes the retention time and peak area of ​​the gas components at the column outlet.

[0060] Based on the gas composition state data, obtain the gas composition concentration data of the gas sample.

[0061] In this embodiment, the gas sample undergoes separation within the chromatographic column, with different components eluting at the column outlet at varying retention times. A thermal conductivity detector is used to detect the eluting gas components, obtaining peak shape data for each component, including retention time and peak area. Retention time reflects the characteristics of the gas component, and different gas components have different retention time values. Peak area is directly proportional to the concentration of the gas component and can be used to calculate concentration values.

[0062] Based on the known retention times and peak areas of standard gas samples, a calibration curve is established between gas component concentration and peak area, expressed by the formula:

[0063]

[0064] In the formula, C i Let k1 represent the measured concentration of the i-th gas component, and A represent the correction coefficient. i This represents the peak area of ​​the i-th gas component.

[0065] In this embodiment, the real-time concentration measurement module performs component analysis using gas chromatography to obtain gas component concentration data, enabling precise detection of the concentrations of various components in the gas sample. This helps to accurately determine the internal chemical changes of the transformer, providing key quantitative indicators for fault diagnosis and improving the accuracy and reliability of fault diagnosis.

[0066] The air circulation calculation module is used to calculate the total heat dissipation of the dry-type transformer within a preset time period, and to calculate the air circulation data inside the dry-type transformer based on the total heat dissipation.

[0067] The air circulation calculation module includes a heat dissipation calculation unit, which calculates the total heat dissipation by calculating both radiative and convective heat dissipation. The formula for calculating radiative heat dissipation is as follows:

[0068]

[0069] The formula for calculating heat dissipation through heat convection is expressed as follows:

[0070]

[0071] The formula for calculating the total heat dissipation is as follows:

[0072]

[0073] In the formula, Indicates the surface emissivity of a dry-type transformer. denoted as Stefan-Boltzmann constant, A as heat dissipation area, T1 as surface temperature of dry-type transformer, T0 as ambient temperature, and h as convective heat transfer coefficient.

[0074] The air circulation calculation module also includes an air flow calculation unit, which is used to calculate the air flow inside the dry-type transformer based on the total heat dissipation.

[0075] According to the law of conservation of energy, the total heat dissipation is equal to the heat absorbed by the air. The formula for calculating heat is as follows:

[0076]

[0077] In the formula, m represents the air mass.

[0078] Since mass equals density multiplied by volume, the formula for air mass is:

[0079]

[0080] In the formula, Q a This represents the airflow rate during time interval t. This indicates air density.

[0081] Substituting m into the formula for calculating heat, we get the formula for calculating heat as follows:

[0082]

[0083] The formula for airflow over time t can be derived as follows:

[0084]

[0085] In the formula, This represents the specific heat capacity of air at constant pressure. This indicates air density.

[0086] In this embodiment, the air circulation calculation module calculates the total heat dissipation and internal air circulation data over a preset time period, which helps to more accurately assess the transformer's operating status. Considering air circulation factors can eliminate interference from the external environment on gas concentration measurements, making the judgment of the actual internal conditions of the transformer more precise and providing a more reliable basis for subsequent fault analysis.

[0087] The true concentration calculation module is used to combine air circulation data and measured concentration data to calculate the true concentration data of the gas components inside the dry-type transformer.

[0088] In this embodiment, we consider the air volume flow rate due to heat dissipation over a time period t as: The internal volume of the transformer is V.

[0089] Assuming the gas component concentration is measured to be Without air circulation, after time t, the amount of gas components remains unchanged, and the concentration remains the same. .

[0090] However, due to air circulation, the volume of fresh air entering the transformer within time t is Q. a *t.

[0091] The introduction of this fresh air will dilute the existing gas composition.

[0092] Viewing this process as a continuous dilution process, for a small time interval The change in concentration can be expressed as:

[0093]

[0094] Simplifying the above equation and separating the variables, we get:

[0095]

[0096] Integrate both sides:

[0097]

[0098] Integrating on the left side yields:

[0099]

[0100] Right now:

[0101]

[0102] After sorting, we can obtain:

[0103]

[0104] The formula for calculating the actual concentration data is expressed as follows:

[0105]

[0106] In the formula, This represents the measured concentration data of the i-th gas component, V represents the volume of the internal space of the dry-type transformer, and t represents the airflow rate inside the dry-type transformer. The time elapsed.

[0107] In this embodiment, the true concentration calculation module calculates the true concentration data by combining air circulation data and measured concentration data. This eliminates the influence of environmental factors and yields a concentration value that more accurately reflects the actual internal condition of the transformer. This helps to more accurately determine whether a fault exists, reduces the possibility of false alarms, and improves the sensitivity and specificity of fault monitoring.

[0108] The fault detection module compares the actual concentration data with the preset normal concentration threshold range. If the actual concentration data exceeds the normal concentration threshold range, it is marked as fault concentration data, indicating that there is a fault in the dry-type transformer.

[0109] The process of setting the normal concentration threshold range includes:

[0110] Historical data on gas composition concentrations during normal operation of dry-type transformers were collected as sample data.

[0111] Calculate the mean and standard deviation of the concentration data for each gas component. The formula for calculating the mean is as follows:

[0112]

[0113] The formula for calculating the standard deviation is expressed as:

[0114]

[0115] In the formula, x j Let j represent the j-th sample value, and n represent the number of samples.

[0116] Based on the mean and standard deviation, a normal concentration threshold range for the cost of each gas is set, expressed by the following formula:

[0117]

[0118] In the formula, A i B i denoted by , where represents the normal concentration threshold range of the i-th gas component, and k represents the confidence coefficient.

[0119] In this embodiment, the fault diagnosis module determines faults by comparing the actual concentration data with the normal concentration threshold range, enabling a quick and intuitive identification of whether a transformer is faulty. This clear judgment standard helps improve the efficiency and accuracy of fault monitoring, allowing for timely problem detection and saving valuable time for subsequent maintenance and repair.

[0120] The fault analysis module is used to build a fault analysis model based on a convolutional neural network. It takes fault concentration data as input and outputs fault data of dry-type transformers. The fault data includes fault type, fault location and fault severity.

[0121] The process of constructing a fault analysis model based on a convolutional neural network includes:

[0122] Collect historical fault case data, including gas composition concentration data, temperature data, and operating parameters of dry-type transformers.

[0123] Historical failure case data is standardized and labeled with failure type, failure location and failure severity to obtain labeled data.

[0124] A basic convolutional neural network model is constructed, consisting of convolutional layers, pooling layers, and fully connected layers. Convolutional layers extract features from the labeled data, yielding feature data. Pooling layers reduce the feature dimensionality of the feature data, resulting in sharper features. Fully connected layers flatten these sharp features and output fault data.

[0125] The base model is trained using labeled data, and its parameters are iteratively updated.

[0126] The accuracy of the output fault data is calculated, and the basic model that achieves the preset accuracy is used as the fault analysis model.

[0127] In this embodiment, the fault analysis module constructs a fault analysis model based on a convolutional neural network, enabling it to deeply mine hidden information in real concentration data and accurately output the fault type, location, and severity. This provides maintenance personnel with detailed and accurate fault information, helping to improve the targeting and efficiency of maintenance and reduce maintenance time and costs.

[0128] The fault reporting module is used to send fault data to maintenance personnel.

[0129] The fault reporting module includes a fault receiving unit, a report generation unit, and a report sending unit. The fault receiving unit receives fault data. The report generation unit performs structured processing on the fault data to obtain a fault report. The report sending unit sends the fault report to relevant maintenance personnel.

[0130] In this embodiment, the fault reporting module sends fault data to maintenance personnel, ensuring that they receive fault information promptly and respond quickly. This helps shorten the time interval for fault handling, improves the stability and reliability of the power system, and reduces the impact of power outages caused by faults on production and daily life.

[0131] In summary, this embodiment provides an artificial intelligence-based dry-type transformer fault monitoring system. Through the collaborative work of multiple modules, it acquires data from multiple dimensions such as temperature, parameters, and gas samples, and performs in-depth analysis. This enables timely and accurate detection of potential faults, improving the accuracy and reliability of fault diagnosis. By comprehensively considering airflow factors and employing advanced analytical methods and models, the system reduces the risk of misjudgment and missed detection, significantly improving maintenance efficiency and quality. By considering the impact of airflow caused by heat dissipation on the concentration of internal gas components in the dry-type transformer and calculating the true concentration of internal gas components, it brings many significant benefits to the fault monitoring of dry-type transformers. In the early stages of a transformer fault, internal chemical reactions and physical changes lead to the generation and concentration changes of specific gas components. By accurately calculating the true concentration, these subtle changes can be captured before the fault develops to a severe stage, gaining valuable time for timely maintenance and protection measures and effectively avoiding power outages and equipment damage caused by sudden faults. Furthermore, it improves the accuracy and reliability of fault monitoring. Traditional monitoring methods can be affected by environmental factors and measurement errors. However, by comprehensively considering the dynamic changes in air circulation and gas concentration, some external interference factors can be eliminated, allowing for a more accurate reflection of the actual internal condition of the transformer and reducing the possibility of misjudgments and omissions. Furthermore, it helps optimize transformer maintenance strategies. Based on real-time calculated gas concentrations, maintenance plans can be formulated more scientifically, allowing for the rational allocation of maintenance time and resources, avoiding over-maintenance or under-maintenance, reducing maintenance costs, and extending the transformer's service life. In addition, it enhances the stability and reliability of the power system. Timely detection and handling of potential transformer faults ensures normal transformer operation, reduces power supply interruptions caused by transformer failures, thereby improving the overall power supply quality and stability, and meeting users' continuous and stable power demands. Finally, it promotes the development and innovation of dry-type transformer monitoring technology. This fault monitoring method based on gas concentration calculation provides new ideas and directions for related research, driving the continuous improvement and advancement of monitoring technology, and providing strong support for the intelligent and digital development of the power industry.

[0132] It also provides maintenance personnel with detailed fault data, making maintenance work more targeted, shortening maintenance time, reducing maintenance costs, and enhancing the stability and reliability of the power system. Timely detection and handling of faults reduces power outages caused by transformer failures, ensuring a continuous and stable power supply, which is of great significance for industrial production, residential life, and the normal operation of society. Furthermore, the system's high degree of intelligence and automation reduces human intervention, lowers the errors and risks that may arise from manual monitoring, and improves the efficiency and safety of monitoring work.

[0133] To realize the aforementioned AI-based dry-type transformer fault monitoring system and apply it to existing dry-type transformer fault monitoring to simplify the fault monitoring process and improve the accuracy and reliability of fault monitoring, this embodiment also provides an AI-based dry-type transformer fault monitoring device. The fault monitoring device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it performs fault monitoring on existing dry-type transformers, improving the accuracy and reliability of fault monitoring.

[0134] The computer device can be a smartphone, tablet, laptop, desktop computer, rack server, blade server, tower server, or cabinet server (including standalone servers or server clusters composed of multiple servers), etc., capable of executing programs. The computer device in this embodiment includes, but is not limited to, a memory and a processor that can communicate with each other via a system bus.

[0135] In this embodiment, the memory (i.e., the readable storage medium) includes flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory can be an internal storage unit of a computer device, such as the hard disk or RAM of the computer device. In other embodiments, the memory can also be an external storage device of the computer device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the computer device. Of course, the memory can also include both internal storage units and external storage devices of the computer device. In this embodiment, the memory is typically used to store the operating system and various application software installed on the computer device. In addition, the memory can also be used to temporarily store various types of data that have been output or will be output.

[0136] In some embodiments, the processor may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor is typically used to control the overall operation of a computer device. In this embodiment, the processor is used to run program code stored in memory or process data, thereby enabling fault monitoring of existing dry-type transformers and improving the accuracy and reliability of fault detection.

[0137] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0138] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0139] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A dry-type transformer fault monitoring system based on artificial intelligence, characterized in that, include: The acquisition module is used to acquire temperature data and parameter data of the dry-type transformer, and to collect gas samples inside the dry-type transformer in real time. The real-time concentration measurement module is used to perform component analysis on the gas sample by gas chromatography to obtain gas component concentration data. An air circulation calculation module is used to calculate the total heat dissipation of the dry-type transformer within a preset time period, and to calculate the air circulation data inside the dry-type transformer based on the total heat dissipation. The true concentration calculation module is used to calculate the true concentration data of the gas components inside the dry-type transformer by combining the air circulation data and the measured concentration data. The calculation formula for the true concentration data is expressed as follows: In the formula, C i Let V represent the measured concentration data of the i-th gas component, V represent the volume of the internal space of the dry-type transformer, and t represent the airflow rate Q inside the dry-type transformer. a Time elapsed; The fault detection module is used to compare the actual concentration data with a preset normal concentration threshold range. If the actual concentration data exceeds the normal concentration threshold range, it is marked as fault concentration data, indicating that the dry-type transformer has a fault. The fault analysis module is used to construct a fault analysis model based on a convolutional neural network. It takes the fault concentration data as input and outputs the fault data of the dry-type transformer. The fault data includes the fault type, fault location, and fault severity. The fault reporting module is used to send the fault data to maintenance personnel.

2. The artificial intelligence-based dry-type transformer fault monitoring system according to claim 1, characterized in that, The parameter data includes the volume of gas inside the dry-type transformer and the surface area of ​​the dry-type transformer; the temperature data includes the surface temperature of the dry-type transformer and the ambient temperature.

3. The artificial intelligence-based dry-type transformer fault monitoring system according to claim 1, characterized in that, The process of performing component analysis on the gas sample includes: The gas sample is injected into a gas chromatograph, and the appropriate chromatographic column is selected based on the gas composition of the gas sample; The gas composition state data of the gas sample at the outlet of the chromatographic column are detected using a thermal conductivity detector. The gas composition state data includes the retention time and peak area of ​​the gas components at the outlet of the chromatographic column. Based on the gas composition state data, the gas composition concentration data of the gas sample is obtained.

4. The artificial intelligence-based dry-type transformer fault monitoring system according to claim 1, characterized in that, The air circulation calculation module includes a heat dissipation calculation unit, which is used to obtain the total heat dissipation by calculating heat radiation and heat convection; the calculation formula for heat radiation is expressed as: P r =e*s*A*(T1 4 -T0 4 ) The formula for calculating heat dissipation via heat convection is expressed as follows: P c =h*A*(T1-T0) The formula for calculating the total heat dissipation is as follows: Q=P r +P c In the formula, ε represents the surface emissivity of the dry-type transformer, σ represents the Stefan-Boltzmann constant, A represents the heat dissipation area, T1 represents the surface temperature of the dry-type transformer, T0 represents the ambient temperature, and h represents the convective heat transfer coefficient.

5. The artificial intelligence-based dry-type transformer fault monitoring system according to claim 4, characterized in that, The air circulation calculation module further includes an airflow calculation unit, which is used to calculate the airflow inside the dry-type transformer based on the total heat dissipation, expressed by the formula: In the formula, c p ρ represents the specific heat capacity of air at constant pressure, and ρ represents the density of air.

6. The artificial intelligence-based dry-type transformer fault monitoring system according to claim 1, characterized in that, The process of setting the normal concentration threshold range includes: Historical data on gas composition concentration during normal operation of the dry-type transformer were collected as sample data; Calculate the average concentration data (μ) of each gas component. i and standard deviation β i ; Based on the average value and the standard deviation, a normal concentration threshold range for the cost of each gas is set, expressed by the following formula: A i B i =μ i ±k*β i In the formula, A i B i denoted by , where represents the normal concentration threshold range of the i-th gas component, and k represents the confidence coefficient.

7. The artificial intelligence-based dry-type transformer fault monitoring system according to claim 1, characterized in that, The process of constructing the fault analysis model based on the convolutional neural network includes: Collect historical fault case data, including gas composition concentration data, temperature data, and operating parameters of dry-type transformers; The historical fault case data is standardized and labeled with fault type, fault location and fault severity to obtain labeled data; A basic model of a convolutional neural network is constructed, which includes convolutional layers, pooling layers, and fully connected layers. The convolutional layers are used to extract features from the labeled data to obtain feature data; the pooling layers are used to reduce the feature dimension of the feature data to obtain distinctive features; and the fully connected layers are used to flatten the distinctive features and output fault data. The base model is trained using the labeled data, and the parameters of the base model are iteratively updated. The accuracy of the output fault data is calculated, and the basic model that achieves the preset accuracy is used as the fault analysis model.

8. The artificial intelligence-based dry-type transformer fault monitoring system according to claim 1, characterized in that, The fault reporting module includes a fault receiving unit, a report generating unit, and a report sending unit. The fault receiving unit is used to receive the fault data; the report generating unit is used to perform structured processing on the fault data to obtain a fault report; and the report sending unit is used to send the fault report to relevant maintenance personnel.

9. A fault monitoring device for dry-type transformers based on artificial intelligence, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, Each functional module in the AI-based dry-type transformer fault monitoring device is deployed in the manner described in any one of claims 1 to 8 as an AI-based dry-type transformer fault monitoring system.

Images