Gas insulation equipment digital model construction method and device, equipment and storage medium

Abstract

The embodiment of the invention discloses a gas insulation equipment digital model construction method and device, equipment and a storage medium. The method comprises the following steps: arranging multi-physics field sensors at a plurality of preset positions of the gas insulation equipment, and acquiring operation data acquired by the multi-physics field sensors; data fusion processing is carried out on the operation data to obtain a system sample data set, and the system sample data set comprises a training set and a test set; acquiring equipment information of the gas insulation equipment, and constructing a first model according to the equipment information; the training set and the first model are processed through a dynamic mode decomposition method to obtain a second model, and the model dimension of the second model is lower than that of the first model; testing the accuracy of the second model according to the test set; and correcting the second model according to the accuracy to obtain a third model, wherein the accuracy of the third model is greater than or equal to that of the second model. Therefore, the digital model of the gas insulation equipment can be constructed more accurately.

Description

Technical Field

[0001] The present application belongs to the technical field of power equipment, and in particular, relates to a method for constructing a digital model of gas-insulated equipment, a device for constructing a digital model of gas-insulated equipment, a computer device, and a computer-readable storage medium. Background Art

[0002] Digital twin is a dynamic virtual model that uses digital technology to build real-time changes. This model needs to be consistent with the changes of the corresponding entity. Digital twin technology maps physical entities into digital models, which can be simulated, analyzed and optimized in the model. This technology has been widely used in industrial production, medical and health fields, and the application of digital twin technology to the power system field has a wide range of application space and bright development prospects. In the power industry, digital twin models involve the coupling of multiple disciplines and multiple physical fields, including mechanics, materials science, electromagnetics, etc., which will result in a large number of design variables and state variables. A single type of data cannot well reflect the actual operating status of the equipment. It is necessary to integrate multiple types of data into a unified and rich system to enhance the accuracy of the model. It refers to the process of integrating data from different sources, different types or different sensors into a consistent, accurate and complete information set. Its purpose is to improve the quality, reliability and information content of data by combining the advantages of multi-source data to support more effective decision-making, analysis and application.

[0003] However, a large amount of data will cause the constructed model to be large in size and high in dimension, with high computational complexity and long time required, which cannot meet the real-time requirements. Therefore, the order reduction method came into being and became an important basis for the application of digital twin technology in the field of power equipment simulation. Order reduction refers to the process of reducing data dimensions while retaining the main key information of the data and ignoring the secondary information that has no obvious impact on the model state. Order reduction has gradually become an important component and one of the key steps in the application of digital twin technology. The use of order reduction technology can reduce computing costs, shorten optimization time, and achieve real-time reflection of the operating status of the physical system. Gas insulated switchgear (GIS) has a complex structure and complex multi-physical field coupling relationship. Traditional simulation modeling methods are difficult to establish a model that is both universal and accurate due to limitations such as mesh division and calculation speed. Therefore, how to build a more accurate digital simulation model of gas insulated equipment is a technical problem that needs to be solved urgently by technicians in this field.

[0004] The preceding description is intended to provide general background information and does not necessarily constitute prior art. Summary of the invention

[0005] Based on this, it is necessary to propose a method for constructing a digital model of gas-insulated equipment, a device for constructing a digital model of gas-insulated equipment, a computer device and a computer-readable storage medium to address the above problems, which can more accurately construct a digital model of gas-insulated equipment.

[0006] The present application solves the technical problem by adopting the following technical solutions:

[0007] The present application provides a method for constructing a digital model of a gas-insulated device, comprising the following steps: setting a multi-physical field sensor at multiple preset positions of the gas-insulated device to obtain operating data collected by the multi-physical field sensor; performing data fusion processing on the operating data to obtain a system sample data set, the system sample data set including a training set and a test set; obtaining device information of the gas-insulated device, and constructing a first model according to the device information; processing the training set and the first model by a dynamic mode decomposition method to obtain a second model, the model dimension of the second model being lower than the model dimension of the first model; testing the accuracy of the second model according to the test set; correcting the second model according to the accuracy to obtain a third model, the accuracy of the third model being greater than or equal to the accuracy of the second model.

[0008] In an optional embodiment of the present application, the support degree of any multi-physical field sensor in t by other multi-physical field sensors is calculated as follows:

[0009]

[0010] According to the fuzzy support and support, the sequence set is fused to obtain the system sample data set. The calculation method of the system sample data set is expressed as:

[0011]

[0012] is the variance of the running data during time period t.

[0013] In an optional embodiment of the present application, equipment information of a gas-insulated device is obtained, and a first model is constructed based on the equipment information, including: obtaining equipment information including geometric dimensions and material parameters, and constructing a 1:1 physical model of the gas-insulated device based on the geometric dimensions and material parameters, wherein the physical model is composed of multiple physical entities; simplifying the physical model at the physical and geometric level based on artificial experience; determining the influence of each physical entity, wherein the influence is used to indicate the degree of influence of the physical entity on the operating data; deleting physical entities having an influence below an influence threshold from the physical model, and marking the deleted physical model as the first model.

[0014] In an optional embodiment of the present application, the training set and the first model are processed by a dynamic mode decomposition method to obtain a second model, including: decomposing the running data into a first snapshot matrix and a second snapshot matrix, the number of snapshot matrices corresponds to the number of preset positions, and the first snapshot matrix and the second snapshot matrix are respectively expressed as:

[0015]

[0016] U in the first snapshot matrix and the second snapshot matrix q (x m ) is the operation data, q is the sampling number of the operation data, and m is the preset position of the operation data; the first model is represented according to the operation data, which is represented as:

[0017] U q+1 =BU q ;

[0018] Where B is the first matrix, which is a full-order matrix of order m; the second matrix is ​​obtained by performing singular value decomposition on the first snapshot matrix, which is expressed as:

[0019] X1=MΣV T ;

[0020] Wherein, M is the second matrix, which is the left singular matrix of the first snapshot matrix; V is the right singular matrix of the first snapshot matrix; Σ is the singular value matrix of the first snapshot matrix; the third matrix is ​​determined according to the second matrix, the second snapshot matrix, the right singular matrix and the singular value matrix, and the third matrix is ​​expressed as:

[0021]

[0022] In the formula, B k is the third matrix; V is the transposed matrix of the second matrix with predetermined high-order terms omitted and only the first k terms retained, k is a matrix that omits predetermined conditional high-order terms from the right singular matrix and retains only the first k terms. The predetermined conditional high-order terms are omitted in the singular value matrix, and only the inverse matrix of the first k terms is retained, where k is a preset integer and satisfies k<m; the third matrix is ​​used to replace the first matrix to obtain the second model, which is expressed as:

[0023] U′ q+1 =B k U′ q .

[0024] In an optional embodiment of the present application, the accuracy of the second model is tested according to the test set, including: obtaining the test input and the actual result in the test set, injecting the test input into the second model for processing to obtain the test result; calculating the relative mean square error according to the actual result and the test result, and characterizing the accuracy with the relative mean square error. The relative mean square error calculation method is expressed as:

[0025]

[0026] U i For the actual result, U i ' is the test result; E RMSE is the relative mean square error.

[0027] In an optional embodiment of the present application, the second model is corrected according to the accuracy to obtain the third model, including: determining whether the accuracy meets the preset verification conditions; if the accuracy does not meet the verification conditions, marking the second model as a discarded model and deleting it; re-acquiring the operating data, determining a new training set based on the operating data, processing again according to the new training set to obtain a new second model, and repeating the execution until the new second model meets the verification conditions; if the accuracy meets the verification conditions, marking the second model as the third model.

[0028] In an optional embodiment of the present application, multi-physical field sensors are arranged at multiple preset positions of gas-insulated equipment to obtain operating data collected by the multi-physical field sensors, including: the preset positions include at least one of insulators, busbars, flanges, circuit breakers, disconnectors, grounding switches, gas compartments, and joints; the operating data include at least one of temperature, humidity, air pressure, electromagnetism, pressure, and vibration.

[0029] The present application also provides a device for constructing a digital model of a gas-insulated device, comprising: an acquisition module, used to set a multi-physical field sensor at multiple preset positions of the gas-insulated device to obtain operating data collected by the multi-physical field sensor; perform data fusion processing on the operating data to obtain a system sample data set, the system sample data set includes a training set and a test set; a training module, used to obtain equipment information of the gas-insulated device, and construct a first model based on the equipment information; process the training set and the first model through a dynamic mode decomposition method to obtain a second model, the model dimension of the second model is lower than the model dimension of the first model; a verification module, used to test the accuracy of the second model based on the test set; correct the second model based on the accuracy to obtain a third model, the accuracy of the third model is greater than or equal to the accuracy of the second model.

[0030] The present application also provides a computer device, including a processor and a memory: the processor is used to execute a computer program stored in the memory to implement the aforementioned method.

[0031] The present application also provides a computer-readable storage medium storing a computer program, which implements the aforementioned method when the computer program is executed by a processor.

[0032] The embodiments of the present application have the following beneficial effects:

[0033] The present application can set up multi-physical field sensors at multiple preset positions of the gas-insulated equipment to obtain real operating data and construct a real first model. The first model can be understood as a mathematical model of 1:1 simulation of the gas-insulated equipment. The training set obtained by collecting the operating data is then used to reduce the order of the first model through the dynamic mode decomposition method, thereby reducing the complexity of the first model and obtaining a second model with a lower model dimension, reducing the dimension and complexity of the model, saving computing resources, and improving computing efficiency. Finally, the accuracy of the second model is verified and corrected using the test set obtained from the operating data, so that the final output third model can meet the requirements of a more simplified mathematical model while ensuring that the accuracy of the third model meets the corresponding requirements, thereby improving the real-time performance of the model.

[0034] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following preferred embodiments are specifically cited and described in detail with the accompanying drawings. It should be understood that the above general description and the detailed description below are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0036] in:

[0037] Figure 1 A schematic diagram of a flow chart of a method for constructing a digital model of a gas-insulated device provided by an embodiment;

[0038] Figure 2 A schematic diagram of a flow chart of a second model correction provided by an embodiment;

[0039] Figure 3 A diagram showing the relationship between internal functional modules of a device for building a digital model of a gas-insulated device provided by an embodiment;

[0040] Figure 4A schematic block diagram of the structure of a computer device provided by an embodiment. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0042] Gas-insulated equipment (GIS) has a complex structure and complex multi-physical field coupling relationships. Traditional simulation modeling methods are difficult to establish a model that is both universal and accurate due to limitations such as grid division and calculation speed. Therefore, a method for constructing a digital model of gas-insulated equipment based on digital twins is proposed, which can reduce the model dimension, reduce the complexity of data, and improve the calculation speed. It is of great significance to the application of digital twin technology in power systems. In order to clearly describe a method for constructing a digital model of gas-insulated equipment provided in this embodiment, please refer to Figure 1 , Figure 2 , including steps S110 to S130.

[0043] Step S110: multi-physical field sensors are set at multiple preset positions of the gas-insulated equipment to obtain operating data collected by the multi-physical field sensors; and the operating data is fused to obtain a system sample data set, which includes a training set and a test set.

[0044] In one embodiment, multi-physical field sensors are arranged at multiple preset positions of gas-insulated equipment to obtain operating data collected by the multi-physical field sensors, including: the preset positions include at least one of insulators, busbars, flanges, circuit breakers, disconnectors, grounding switches, gas compartments, and joints; the operating data include at least one of temperature, humidity, air pressure, electromagnetism, pressure, and vibration.

[0045] In one embodiment, in order to simulate the digital GIS, it is necessary to have a sufficient understanding of GIS. For this purpose, the corresponding physical field data can be collected at the key positions of the GIS, which are also preset positions, which can be selected by the user, including but not limited to insulators, busbars, flanges, circuit breakers, disconnectors, grounding switches, gas compartments, joints, etc. The actual selection includes at least one of the above positions, and in a preferred embodiment, multiple. The specific operation data is the physical field data generated by the GIS during operation, including but not limited to at least one of temperature, humidity, air pressure, electromagnetic, pressure, and vibration. In order to obtain data from different sensors by data fusion, standard optimization technology is used for integration, relatively more comprehensive equipment status information is obtained, and the performance of the GIS is summarized, analyzed and predicted in chronological order. In order to characterize the acquisition method and specific content of the operation data, the type and acquisition position can be separately indicated, such as recording the operation data in the form of (preset position, data type, specific value).

[0046] In one embodiment, it can be understood that although the operating data is divided into various types according to the physical field and the preset position, in fact, there may be influences between each specific operating data. For example, taking the preset position as an anchor, taking the cable joint as an example, the operating data that may be collected include temperature, partial discharge signals, and electromagnetic radiation; correspondingly, if the data type is taken as an example, taking temperature as an example, the collection location of the same temperature may include cable joints, busbars, conductor connection parts, circuit breaker contacts, etc., and may also specifically include conductor temperature, contact temperature, insulating material surface temperature, etc. It can be seen that there is a certain correlation between each specific operating data. If all the collected operating data are used for training and analysis, its complexity will be greatly increased. In order to reduce the complexity, data with certain correlations can be fused, and data fusion processing can be performed on the operating data to obtain a system sample data set, including: first, the operating data is divided into sequence sets according to multiple physical field sensors. The specific form of the sequence set can be expressed as:

[0047]

[0048] Y i Represents the sequence set of the nth multiphysics sensor.

[0049] Calculate the fuzzy support between two sequence sets in the time period t. The fuzzy support calculation method is expressed as:

[0050]

[0051] Z ij Y is the fuzzy support. i (t) and Y j(t) are two sequence sets, indicating that the data collected by the selected sensors a and b constitute a time series within the time period t. D is the dynamic bending distance, according to Y i (t) and Y j (t) is obtained. L is the preset fuzzy threshold. The fuzzy threshold is not a fixed value. It is set manually according to different usage scenarios (different number of sensors, different degree of association, etc.). If the currently set threshold causes the final result to deviate too much, it can be reset.

[0052] According to the fuzzy support, the support of any multi-physics field sensor by other multi-physics field sensors in the time period t can be calculated. The support calculation method is expressed as:

[0053]

[0054] The sequence set is fused according to the fuzzy support and support to obtain the system sample data set. Specifically, after comprehensively considering the support between sensors of the same type and the reliability of the sensors themselves, a weighted fusion algorithm is used to achieve data fusion between sensors of the same type. The weighted fusion algorithm integrates data based on the reliability and mutual support of sensors to obtain the system sample data set. The calculation method of the system sample data set is expressed as:

[0055]

[0056] is the variance of the running data during time period t. The above processing can realize the data fusion function. Further, the fused system sample data set can be used to divide the system sample data set representing different information into training set and test set according to a predetermined ratio. Specifically, the training set and the test set can be divided into 70% and 30% respectively. The actual ratio can be set arbitrarily according to actual needs. This is only an example of the method, not a limitation of the solution.

[0057] Step S120: Obtain equipment information of the gas-insulated equipment, and construct a first model according to the equipment information; process the training set and the first model by dynamic mode decomposition method to obtain a second model, wherein the model dimension of the second model is lower than the model dimension of the first model.

[0058] In one embodiment, device information of a gas-insulated device is obtained, and a first model is constructed based on the device information, including: obtaining device information including geometric dimensions and material parameters, and constructing a 1:1 physical model of the gas-insulated device based on the geometric dimensions and material parameters, wherein the physical model is composed of multiple physical entities; simplifying the physical model at a physical and geometric level based on artificial experience; determining an influence degree of each physical entity, wherein the influence degree is used to indicate the degree of influence of the physical entity on operating data; and deleting physical entities having an influence degree lower than an influence threshold from the physical model, and marking the deleted physical model as the first model.

[0059] In one embodiment, the training set and the first model are processed by dynamic mode decomposition to obtain the second model. In a preferred embodiment of the present application, the data in the training set are subjected to dynamic mode decomposition (DMD) to form a high-fidelity reduced-order model, i.e., the second model. Taking the flow field as an example, the m sampling points x1, x2, x3…x m After q samplings, we get the flow field data U(x1), U(x2), U(x3)…U(x m ). It is decomposed into a first snapshot matrix and a second snapshot matrix. The number of snapshot matrices corresponds to the number of preset positions. The first snapshot matrix and the second snapshot matrix are respectively expressed as:

[0060]

[0061] U in the first snapshot matrix and the second snapshot matrix q (x m ) is the running data, q is the sampling number of the running data, and m is the preset position of the running data. If the sampling number q is large enough, the flow field can be approximately considered as a linear system, that is, there is a linear matrix B with a higher dimension to represent the first model, which can be expressed as:

[0062] U q+1 =BU q (7)

[0063] In formula (7), B is the first matrix, which is a full-order matrix with an order of m. Formula (7) can also be used to represent the first model.

[0064] The second matrix is ​​obtained by performing singular value decomposition on the first snapshot matrix, which is expressed as:

[0065] X1=MΣV T (8)

[0066] In formula (8), M is the second matrix, which is the left singular matrix of the first snapshot matrix; V is the right singular matrix of the first snapshot matrix; Σ is the singular value matrix of the first snapshot matrix, and the elements on the diagonal are called singular values.

[0067] The third matrix is ​​determined according to the second matrix, the second snapshot matrix, the right singular matrix and the singular value matrix. Specifically, the third matrix can be obtained by omitting the high-order terms that are approximately 0 in the singular value matrix Σ and only taking the first k terms into consideration. The third matrix is ​​expressed as:

[0068]

[0069] In formula (9), B k is the third matrix. V is the transposed matrix of the second matrix with predetermined high-order terms omitted and only the first k terms retained, k is a matrix that omits predetermined conditional high-order terms from the right singular matrix and retains only the first k terms. The predetermined conditional high-order items are omitted from the singular value matrix, and only the inverse matrix of the first k items is retained, where k is a preset integer and satisfies k<m.

[0070] At this time, the third matrix B k For a k-order square matrix, the third matrix B k Replacing the first matrix B in equation (7) to obtain the reduced-order second model, the second model is expressed as:

[0071] U′ q+1 =B k U′ q (10)

[0072] Step S130: testing the accuracy of the second model according to the test set; modifying the second model according to the accuracy to obtain a third model, wherein the accuracy of the third model is greater than or equal to the accuracy of the second model.

[0073] In one embodiment, for the optimization and correction of the second model, the specific process can be referred to Figure 2 The flowchart of the second model correction shown in the figure expands and refines step S130, and specifically also includes steps S210 to S240.

[0074] Step S210: obtain the test input and actual results in the test set, inject the test input into the second model to process and obtain the test result; calculate the relative mean square error according to the actual result and the test result, and use the relative mean square error to characterize the accuracy.

[0075] In one embodiment, the GIS as a whole can be treated as a black box in the test set, so the collected data must include input and output, and the test set is obtained from the operating data, that is, the input and output are the input and output of the GIS under actual operation. For this purpose, the test input and actual result in the test set can be obtained, that is, the actual input and output of the GIS under specific circumstances. The test input is injected into the second model for processing, and the result of the simulated output of the simulation model, that is, the test result, can be obtained. By comparing the test results with the actual results, the accuracy of the second model can be determined, that is, whether the second model can accurately predict or simulate the actual GIS. In a preferred embodiment of the present application, the relative mean square error is calculated based on the actual results and the test results, and the accuracy is characterized by the relative mean square error to determine. The relative mean square error calculation method is expressed as:

[0076]

[0077] U i For the actual result, U i ' is the test result; E RMSE is the relative mean square error, which is also the corresponding accuracy.

[0078] Step S220: Determine whether the accuracy meets the preset verification conditions.

[0079] In one embodiment, the verification condition corresponds to the accuracy. Specifically, the smaller the relative mean square error value is, the smaller the error is, and the higher the accuracy of the reduced-order model is. Therefore, a threshold can be set as a definition. When the accuracy is higher than or equal to the threshold, it is determined that the verification condition is not met; otherwise, it is met. The specific verification condition can be set arbitrarily according to the actual needs of the user, and there is no restriction on this.

[0080] If the accuracy does not meet the verification condition, execute step S230: mark the second model as a discarded model and delete it; and return to step S110.

[0081] In one embodiment, if the accuracy of the second model after reduction fails to pass the verification, the above steps need to be performed again, returning to step 110, re-collecting the sensor data and reducing the order again until the accuracy verification passes. Re-acquire the operating data, determine a new training set based on the operating data, and process again to obtain a new second model based on the new training set, and repeat the process until the new second model meets the verification conditions. Furthermore, the regeneration process may also include but is not limited to adjusting the fuzziness threshold, thereby changing the system sample data set obtained in the data fusion process; it may also be that during the construction of the first model, the adjustment of the simplified entity may be to retain the originally deleted physical entity, or to further delete the physical entity. The actual operation process may be adjusted according to the reason for the failure of the verification. This is only for illustration and does not impose specific restrictions on the technology.

[0082] If the accuracy meets the verification condition, step S240 is executed: marking the second model as the third model.

[0083] In one embodiment, if the verification is passed, the second model is identified as the third model, that is, the mathematical model after price reduction. The GIS rule model is reduced in order through the DMD method, which reduces the dimension and complexity of the model, saves computing resources, and improves computing efficiency.

[0084] Therefore, the present application can set up multi-physical field sensors at multiple preset positions of the gas-insulated equipment to obtain real operating data and construct a real first model. The first model can be understood as a mathematical model of 1:1 simulation of the gas-insulated equipment. The training set obtained by collecting the operating data is then used to reduce the order of the first model through the dynamic mode decomposition method, thereby reducing the complexity of the first model and obtaining a second model with a lower model dimension, reducing the dimension and complexity of the model, saving computing resources, and improving computing efficiency. Finally, the accuracy of the second model is verified and corrected using the test set obtained from the operating data, so that the final output third model can meet the requirements of a more simplified mathematical model while ensuring that the accuracy of the third model meets the corresponding requirements, thereby improving the real-time performance of the model.

[0085] Figure 3 The internal functional module relationship diagram of a gas-insulated equipment digital model construction device in one embodiment is shown. The gas-insulated equipment digital model construction device 30 includes: an acquisition module 31, a training module 32 and a verification module 33. The acquisition module 31 is used to set a multi-physical field sensor at multiple preset positions of the gas-insulated equipment to obtain the operating data collected by the multi-physical field sensor; perform data fusion processing on the operating data to obtain a system sample data set, and the system sample data set includes a training set and a test set. The training module 32 is used to obtain the equipment information of the gas-insulated equipment and construct a first model according to the equipment information; the training set and the first model are processed by the dynamic mode decomposition method to obtain a second model, and the model dimension of the second model is lower than the model dimension of the first model. The verification module 33 is used to test the accuracy of the second model according to the test set; the second model is corrected according to the accuracy to obtain a third model, and the accuracy of the third model is greater than or equal to the accuracy of the second model.

[0086] Figure 4 FIG. 1 shows an internal structure diagram of a computer device in an embodiment. The computer device may be a terminal or a server. Figure 4As shown, the computer device includes a processor, a memory and a network interface connected via a system bus. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program. When the computer program is executed by the processor, the processor may implement a method for constructing a digital model of a gas-insulated device. The internal memory may also store a computer program. When the computer program is executed by the processor, the processor may execute a method for constructing a digital model of a gas-insulated device. Those skilled in the art may understand that Figure 4 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0087] In one embodiment, the present application further proposes a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the method described in any of the aforementioned embodiments.

[0088] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0089] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0090] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for constructing a digital model of a gas-insulated device, characterized in that: The steps include: Multi-physical field sensors are arranged at a plurality of preset positions of the gas-insulated equipment to obtain operation data collected by the multi-physical field sensors; data fusion processing is performed on the operation data to obtain a system sample data set, wherein the system sample data set includes a training set and a test set; Acquire device information of the gas-insulated device, and construct a first model according to the device information; Processing the training set and the first model by a dynamic mode decomposition method to obtain a second model, wherein the model dimension of the second model is lower than the model dimension of the first model; The accuracy of the second model is tested according to the test set; the second model is modified according to the accuracy to obtain a third model, and the accuracy of the third model is greater than or equal to the accuracy of the second model.

2. The method for constructing a digital model of a gas-insulated device according to claim 1, characterized in that: The performing data fusion processing on the operation data to obtain a system sample data set includes: The operation data is divided into sequence sets according to the multi-physics field sensors, which are expressed as: Y i represents the sequence set of the nth multi-physical field sensor; calculates the fuzzy support between the sequence sets in time period t, and the fuzzy support calculation method is expressed as: Z ij is the fuzzy support; i (t) and Y j (t) are the two sequence sets; L is the preset fuzzy threshold; D is the dynamic bending distance, according to Y i (t) and Y j (t) to obtain; According to the fuzzy support, the support of any of the multi-physics field sensors by the other multi-physics field sensors within the time period t is calculated. The support calculation method is expressed as: The sequence set is data-fused according to the fuzzy support and the support to obtain the system sample data set. The calculation method of the system sample data set is expressed as: is the variance of the operating data during time period t.

3. The method for constructing a digital model of a gas-insulated device according to claim 1, wherein: The step of acquiring device information of the gas-insulated device and constructing a first model according to the device information includes: Acquire the device information including geometric dimensions and material parameters, and construct a 1:1 physical model of the gas-insulated device according to the geometric dimensions and the material parameters, wherein the physical model is composed of a plurality of physical entities; Simplifying the physical model at the physical geometric level according to manual experience; determining the influence degree of each physical entity, wherein the influence degree is used to indicate the influence degree of the physical entity on the operating data; The physical entities whose influence is lower than an influence threshold are deleted from the physical model, and the deleted physical model is marked as the first model.

4. The method for constructing a digital model of a gas-insulated device according to claim 1, wherein: The step of processing the training set and the first model by a dynamic mode decomposition method to obtain a second model includes: The operation data is decomposed into a first snapshot matrix and a second snapshot matrix, the number of the snapshot matrices corresponds to the number of the preset positions, and the first snapshot matrix and the second snapshot matrix are respectively expressed as: The first snapshot matrix and the second snapshot matrix U q (x m ) is the operating data, q is the sampling number of the operating data, and m is the preset position where the operating data is located; The first model is represented according to the operating data, as: U w+1 =BU q ; Wherein, B is the first matrix, which is a full-order matrix of order m; the second matrix is ​​obtained by performing singular value decomposition on the first snapshot matrix, which is expressed as: X1=MΣV T ; Wherein, M is the second matrix, which is the left singular matrix of the first snapshot matrix; V is the right singular matrix of the first snapshot matrix; Σ is the singular value matrix of the first snapshot matrix; A third matrix is ​​determined according to the second matrix, the second snapshot matrix, the right singular matrix and the singular value matrix, and the third matrix is ​​expressed as: In the formula, B k is the third matrix; V is a transposed matrix in which predetermined conditional high-order terms are omitted from the second matrix and only the first k terms are retained, k is a matrix in which predetermined conditional high-order terms are omitted from the right singular matrix and only the first k terms are retained, The inverse matrix of the singular value matrix is ​​obtained by omitting predetermined conditional high-order terms and retaining only the first k terms, where k is a preset integer and satisfies k<m; The third matrix is ​​used to replace the first matrix to obtain the second model, and the second model is expressed as: U q ′ +1 =B k U q ′。 5. The method for constructing a digital model of a gas-insulated device according to claim 1, wherein: The testing the accuracy of the second model according to the test set includes: Acquire the test input and actual result in the test set, and inject the test input into the second model for processing to obtain the test result; The relative mean square error is calculated according to the actual result and the test result, and the accuracy is characterized by the relative mean square error. The relative mean square error calculation method is expressed as: U i For the actual result, U i ' is the test result; E RMSE is the relative mean square error.

6. The method for constructing a digital model of a gas-insulated device according to claim 1, wherein: The step of modifying the second model according to the accuracy to obtain a third model comprises: Determining whether the accuracy meets a preset verification condition; If the accuracy does not meet the verification condition, the second model is marked as a discarded model and deleted; the operation data is reacquired, a new training set is determined according to the operation data, and a new second model is obtained by processing the new training set again, and the execution is repeated until the new second model meets the verification condition; If the accuracy satisfies the verification condition, the second model is marked as the third model.

7. The method for constructing a digital model of a gas-insulated device according to claim 1, characterized in that: The step of arranging a multi-physical field sensor at a plurality of preset positions of the gas-insulated equipment and acquiring the operation data collected by the multi-physical field sensor comprises: The preset position includes at least one of an insulator, a busbar, a flange, a circuit breaker, a disconnector, a grounding switch, a gas compartment, and a joint; The operating data includes at least one of temperature, humidity, air pressure, electromagnetism, pressure, and vibration.

8. A gas-insulated equipment digital model building device, characterized in that: include: A collection module, used to set a multi-physical field sensor at a plurality of preset positions of the gas-insulated equipment and obtain operation data collected by the multi-physical field sensor; Performing data fusion processing on the operating data to obtain a system sample data set, wherein the system sample data set includes a training set and a test set; A training module, used for acquiring equipment information of the gas-insulated equipment and constructing a first model according to the equipment information; Processing the training set and the first model by a dynamic mode decomposition method to obtain a second model, wherein the model dimension of the second model is lower than the model dimension of the first model; A verification module is used to test the accuracy of the second model according to the test set; and to modify the second model according to the accuracy to obtain a third model, wherein the accuracy of the third model is greater than or equal to the accuracy of the second model.

9. A computer device, characterized in that: including a processor and a memory; The processor is configured to execute the computer program stored in the memory to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images