High-voltage switch cabinet insulation medium dielectric spectrum rapid fitting method and device

Abstract

The present disclosure relates to a high-voltage switch cabinet insulation medium dielectric spectrum rapid fitting method and device, which can realize dielectric spectrum rapid fitting to meet the needs of on-site rapid diagnosis. The method comprises the following processing: S1, applying alternating current to the insulation material of the high-voltage switch cabinet; S2, obtaining the dielectric response data of the insulation material; S3, preprocessing the collected dielectric response data to obtain the preprocessed dielectric spectrum data; S4, selecting an initial dielectric response model according to the type of the insulation material; S5, initializing the parameters of the initial dielectric response model to obtain the predicted value of the complex dielectric constant; S6, calculating the adjustment decision coefficient according to the predicted value and the measured value; S7, adjusting the dielectric response model according to the coefficient; and S8, inputting the obtained dielectric response data into the adjusted dielectric response model for dielectric spectrum fitting processing.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power equipment insulation state evaluation, and particularly relates to a high-voltage switch cabinet insulation medium dielectric spectrum rapid fitting method and device. BACKGROUND

[0002] In power equipment, insulation parts are key components of high-voltage switch cabinets, and usually play the roles of electrical insulation, supporting conductors, or isolating air chambers. Operation experience shows that various types of insulation parts such as switch cabinet contact boxes and through-cabinet bushings often have various types of faults such as surface ablation and discharge, which become weak links that restrict the operation reliability of high-voltage switch cabinets.

[0003] As a core device of a power system, a high-voltage switch cabinet may have performance degradation of insulation parts, which may lead to serious faults such as partial discharge and insulation breakdown. Therefore, dielectric spectrum technology is usually used to measure the dielectric response (including relative permittivity and dielectric loss factor) of insulation part materials at different frequencies, and the dielectric response can reflect the polarization and loss characteristics of the insulation part materials, so as to evaluate the insulation state of the insulation part materials.

[0004] However, traditional dielectric spectrum data analysis relies on complex mathematical models and nonlinear fitting methods, and it is difficult to achieve accurate fitting. SUMMARY

[0005] The present application provides a high-voltage switch cabinet insulation medium dielectric spectrum rapid fitting method and device, which can realize rapid and accurate fitting of dielectric spectrum to meet the needs of on-site rapid diagnosis.

[0006] In one aspect, the high-voltage switch cabinet insulation medium dielectric spectrum rapid fitting method provided by the present application embodiment includes the following processing:

[0007] S1, applying an alternating current to the insulation part material of the high-voltage switch cabinet, the frequency of the alternating current being ;

[0008] S2, obtaining dielectric response data of the insulation part material of the high-voltage switch cabinet, the dielectric response data including a measured value of a complex permittivity, wherein the complex permittivity includes a real part and an imaginary part, the real part representing a relative permittivity, and the imaginary part representing a dielectric loss factor;

[0009] S3, preprocessing the collected dielectric response data to obtain preprocessed dielectric spectrum data;

[0010] S4, selecting one of a first model and a second model as an initial dielectric response model according to the type of the insulation part material, the first model being different from the second model;

[0011] S5, initializing parameters of the initial dielectric response model based on the preprocessed dielectric spectrum data to obtain a predicted value of complex dielectric constant;

[0012] S6, calculating an adjustment decision coefficient according to the predicted value and the measured value according to the following formula:

[0013]

[0014] wherein, the adjustment decision coefficient, p represents the number of model parameters, M represents the number of frequency points in dielectric spectrum measurement, and R represents a residual vector whose elements are the difference between the measured value and the predicted value;

[0015] S7, adjusting the dielectric response model according to the calculated adjustment decision coefficient;

[0016] S8, inputting the obtained dielectric response data into the adjusted dielectric response model for dielectric spectrum fitting processing.

[0017] The embodiment of the present application introduces the adjustment decision coefficient , which can adaptively adjust the dielectric response model to achieve accurate fitting.

[0018] In some embodiments, in the processing S3, the pre-processing of the collected dielectric response data includes:

[0019] The wavelet threshold denoising method is adopted, Daubechies wavelet basis is selected to perform multi-scale decomposition on the original signal corresponding to the collected dielectric response data, the high-frequency noise component is suppressed through a soft threshold function, and the effective signal is retained to obtain the processed dielectric spectrum data.

[0020] According to the embodiment of the present application, by performing the above processing on the dielectric response data, the problem that the dielectric response model is not unique and the dielectric response models adapted for different material types are inconsistent, but it is difficult to directly find a suitable matching dielectric response model according to the dielectric spectrum test data can be solved. Moreover, by introducing the adjustment decision coefficient, the model can be changed when the model is not suitable, and the fitting efficiency is improved.

[0021] In some embodiments, the first model is a single relaxation Debye model, and its expression is as follows:

[0022]

[0023] The second model is a Cole-Cole model, and its expression is as follows:

[0024]

[0025] wherein, ε∞ represents the high-frequency limit dielectric constant, εs represents the static dielectric constant, τ represents the relaxation time, j represents the imaginary unit in complex number, α represents the distribution parameter, 0 < α < 1, ≤1.

[0026] In some embodiments, in the processing S5, the initializing the parameters of the initial dielectric response model by the frequency domain feature method comprises:

[0027] extracting the average value of the real part of the complex dielectric constant in a low frequency band from the preprocessed dielectric spectrum data as , the low frequency band including Hz to 1 Hz;

[0028] extracting the average value of the real part of the complex dielectric constant in a high frequency band from the preprocessed dielectric spectrum data as , the high frequency band including Hz to Hz;

[0029] estimating the relaxation time by the loss peak frequency ;

[0030] for the Cole-Cole model, the distribution parameter is initialized as 0.8.

[0031] According to the model initialization parameter setting of the embodiments of the present application, the efficiency of the later optimization iteration is high.

[0032] In some embodiments, the method further comprises: before the processing S7, the optimizing the model parameters by the Levenberg-Marquardt algorithm comprises:

[0033] setting the objective function as minimizing the weighted residual sum of squares:

[0034]

[0035] wherein, is the objective function expressed by the weighted residual sum of squares, is the test value of the real part of the complex dielectric constant at the frequency ω i , is the test value of the imaginary part of the complex dielectric constant at the frequency ω i , is the test value of the imaginary part of the complex dielectric constant at the frequency ω iRe(ω) is the real part of the complex permittivity at frequency ω, and the calculated value is obtained according to the fitting model, ω is the frequency, i Im(ω) is the imaginary part of the complex permittivity at frequency ω, and the calculated value is obtained according to the fitting model, and m is the number of frequencies tested, is the parameter to be optimized, and w1 and w2 represent weights, which can be dynamically adjusted according to measurement errors;

[0036] The Levenberg-Marquardt algorithm is used for iterative calculation to obtain the optimized parameters, wherein the iterative formula is as follows:

[0037]

[0038] In the formula: J is the Jacobian matrix, each row of which corresponds to the partial derivative of the real part of the complex permittivity at frequency ω i r is the residual vector, is the damping factor, T represents matrix transposition, k represents the iteration step, and diag represents the operation of extracting the diagonal elements of the matrix and constructing a diagonal matrix.

[0039] In some embodiments, in the above processing, the Jacobian matrix calculation and matrix inversion operation are distributed to GPU parallel processing.

[0040] Through GPU acceleration, the traditional fitting process of several hours is shortened to minutes, meeting the real-time diagnosis demand on site.

[0041] In some embodiments, the method further comprises verifying the model parameters by the following formula:

[0042]

[0043] In the formula: N represents the length of the residual sequence, e i represents the ith residual, which is specifically defined as follows:

[0044] For the real part residual:

[0045] For the imaginary part residual: ;

[0046] When , it is determined that there is a system error, and the parameters are reinitialized or the model is adjusted.

[0047] In some embodiments, in the processing S7:

[0048] When is greater than or equal to a set value, the selected dielectric response model is accepted;

[0049] When ​When the dielectric response is less than the set value and the selected dielectric response model is the Cole-Cole model, an asymmetry coefficient β is introduced to extend the model to the following HN function model:

[0050]

[0051] During parameter initialization, β=0.5, and the remaining parameters inherit the Cole-Cole fitting results;

[0052] when When the dielectric response is less than the set value and the selected dielectric response model is the Debye model, the multi-scale polarization is characterized by superimposing multiple relaxation terms according to the following formula:

[0053]

[0054] During parameter initialization, the number of relaxation terms K is determined based on the number of loss peaks in the spectrum curve of the imaginary part of the complex permittivity, and the relaxation time is deduced from the loss peak frequency: .

[0055] In some implementations, the real part (relative permittivity) and imaginary part (dielectric loss factor) of the complex permittivity obtained by dielectric spectrum fitting in S8 are compared with the corresponding parameters in the dielectric model library to determine the insulation state of the insulating material of the high-voltage switchgear; wherein, the dielectric model library stores the model parameters after historical fitting.

[0056] In some embodiments, determining the insulation state of the insulating material of the high-voltage switchgear includes:

[0057] for 10Hz to 10 2 If the real part of the complex permittivity (relative permittivity) of Hz increases beyond the first threshold after the comparison, it is determined that the insulating material is aging.

[0058] for 10Hz to 10 2 If the imaginary part of the complex permittivity constant (dielectric loss factor) at Hz increases beyond the second threshold after the comparison, it is determined that the insulating material is damp.

[0059] On the other hand, embodiments of the present invention also provide a computer device including a memory and a processor, the memory storing computer-readable instructions, characterized in that the processor executes the computer-readable instructions to implement the method or any processing of any of the embodiments or examples herein.

[0060] In summary, the application introduces parameter initialization optimization, parallel computing and adaptive algorithm, significantly improves the fitting speed and accuracy, and realizes real-time state evaluation of insulating materials. Among them, through GPU acceleration, the traditional fitting process of several hours is shortened to minutes, meeting the real-time diagnosis demand on site; adaptive model switching and residual analysis mechanism is adopted to avoid overfitting and underfitting problems; and multiple dielectric models are supported, which can be extended to the insulation evaluation of power equipment such as transformers and cables. BRIEF DESCRIPTION OF DRAWINGS

[0061] Various aspects, features, advantages, etc. of embodiments of the present application will be described in greater detail below with reference to the accompanying drawings, in which:

[0062] Figure 1 is a flowchart of a high-voltage switch cabinet insulating medium dielectric spectrum fast fitting method according to an exemplary embodiment of the present application;

[0063] Figure 2 is a schematic diagram showing high-voltage switch cabinet insulating medium complex dielectric constant imaginary part dielectric spectrum fitting according to an exemplary embodiment of the present application;

[0064] Figure 3 is a computer device for performing a high-voltage switch cabinet insulating medium dielectric spectrum fast fitting method according to an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0065] Exemplary embodiments of the present application will be described in detail below with reference to the accompanying drawings. It should be noted that the present application can be implemented in various forms, and is not limited to the specific embodiments described herein or shown in the accompanying drawings.

[0066] The terms "substantially", "approximately", "about", and similar terms are used herein as descriptive terms, not as exact qualifiers. These terms are intended to cover a reasonable range of error or variation.

[0067] The terms "comprise", "include", "have", and similar terms are used herein to indicate the presence of certain features, processes, steps, operations, elements, and / or components, but do not exclude the presence or addition of other features, processes, steps, operations, elements, components, or combinations thereof.

[0068] It should be understood that the terms "first", "second", and the like similar terms do not indicate any priority or order, but are only used to distinguish different objects in the description.

[0069] Figure 1 Exemplary embodiments of the high-voltage switch cabinet insulating medium dielectric spectrum fast fitting method of the present application are shown. As shown, the method includes but is not limited to processes S1 to S8, which are described in detail below.

[0070] In the processing S1, an alternating current is applied to the insulating material of the high-voltage switch cabinet, and the frequency of the alternating current is In the processing S2, the dielectric response data of the insulating material of the high-voltage switch cabinet is obtained, and the dielectric response data includes the measured value of the real part of the complex dielectric constant. The complex dielectric constant includes the real part and the imaginary part . The real part represents the relative dielectric constant, and the imaginary part represents the dielectric loss factor.

[0071] In some embodiments, the dielectric response data of the insulating material of the high-voltage switch cabinet is collected by using a broadband dielectric spectrometer. Specifically, the broadband dielectric spectrometer applies an alternating current to the insulating material, and the frequency of the alternating current is , and the dielectric response data of the insulating material is obtained, which includes and .

[0072] In the processing S3, the collected dielectric response data is preprocessed to obtain the preprocessed dielectric spectrum data. In the preprocessed dielectric spectrum data, the real part of the complex dielectric constant is represented as and the imaginary part is represented as .

[0073] In some embodiments, the preprocessing of the collected dielectric response data includes: using a wavelet threshold denoising method, selecting a Daubechies wavelet basis to perform multi-scale decomposition on the original signal corresponding to the collected dielectric response data, suppressing high-frequency noise components by using a soft threshold function, and retaining effective signals, thereby obtaining dielectric spectrum data including and .

[0074] In the processing S4, one of the first model and the second model is selected as an initial dielectric response model according to the type of the insulating material, and the first model is different from the second model.

[0075] In some embodiments, the first model is a single relaxation Debye model, and the expression is as follows:

[0076]

[0077] The second model is a Cole-Cole model, and the expression is as follows:

[0078]

[0079] wherein represents the high-frequency limit dielectric constant, represents the static dielectric constant, denotes the relaxation time, j denotes the imaginary unit in complex numbers, denotes the distribution parameter, 0 < p < 1

[0080] In some embodiments, when the insulating medium of the insulating member material is a single homogeneous medium, a single relaxation Debye model is initially selected; when the insulating medium of the insulating member material is a non-single homogeneous medium, a Cole-Cole model is initially selected.

[0081] In the processing S5, based on the pre-processed dielectric spectrum data, the parameters of the initial dielectric response model are initialized to obtain a predicted value of the complex dielectric constant. The model parameters include the high-frequency limit dielectric constant, the static dielectric constant, the relaxation time, and the distribution parameter.

[0082] In some embodiments, the initialization of the parameters of the initial dielectric response model by the frequency domain feature method includes:

[0083] From the pre-processed dielectric spectrum data, the average value of the low-frequency band (0.1 Hz to 1 Hz) of the imaginary part of the complex dielectric constant is extracted as the static dielectric constant.

[0084] From the pre-processed dielectric spectrum data, the average value of the high-frequency band (100 kHz to 1 MHz) of the imaginary part of the complex dielectric constant is extracted as the high-frequency limit dielectric constant.

[0085] The relaxation time τ is estimated by the loss peak frequency f p:

[0086] For the Cole-Cole model, the distribution parameter p is initialized to 0.8.

[0087] In the processing S6, the adjustment decision coefficient R is calculated according to the predicted value and the measured value of the complex dielectric constant:

[0088] wherein p represents the number of model parameters; M represents the number of frequency points in the dielectric spectrum measurement, for example, if 48 frequency points are uniformly selected in the range of 0.1 Hz to 1 MHz, then M = 48; R represents the residual vector, and the elements of the residual vector are the differences between the measured value and the predicted value of the complex dielectric constant.

[0089]

[0090] ​​​​​​​​​​​​​​

[0091] In the processing S7, the dielectric response model is adjusted according to the calculated adjustment decision coefficient. In some embodiments, when the adjustment decision coefficient is greater than or equal to a set value, the selected dielectric response model is accepted or used.

[0092] When the adjustment decision coefficient is less than the set value and the initially selected dielectric response model is the Cole-Cole model, an asymmetry coefficient β is introduced to extend the model to the HN function model as follows:

[0093] In the parameter initialization, β = 0.5, and the remaining parameters are inherited from the Cole-Cole fitting result.

[0094] When the adjustment decision coefficient is less than the set value and the initially selected dielectric response model is the Debye model, a multi-relaxation term is superimposed to represent the multi-scale polarization according to the following formula:

[0095]

[0096] In the parameter initialization, the number of relaxation terms K is determined according to the number of loss peaks in the complex dielectric constant imaginary part frequency spectrum curve (for example, the curve shown in FIG. 6), and the relaxation time is inversely deduced from the loss peak frequency: .

[0097] . Figure 2 In the exemplary embodiments, the set value for comparison is 0.9.

[0098] In the exemplary embodiments, the set value for comparison is 0.9.

[0099] In the processing S8, the obtained dielectric response data is input into the adjusted dielectric response model for dielectric spectrum fitting processing.

[0100] In some embodiments, the method comprises, before the processing S7, optimizing the model parameters using the Levenberg-Marquardt algorithm according to the predicted value and the measured value of the complex dielectric constant.

[0101] In some embodiments, the optimization of the model parameters using the Levenberg-Marquardt algorithm comprises:

[0102] Setting the objective function as minimizing the weighted residual sum of squares of the measured value and the predicted value of the complex dielectric constant:

[0103]

[0104] wherein, is the objective function expressed by the weighted residual sum of squares,​​​ is the real part of the complex permittivity test value at frequency ω i is the imaginary part of the complex permittivity test value at frequency ω i is the real part of the complex permittivity, the calculated value according to the fitting model at frequency ω i is the imaginary part of the complex permittivity, the calculated value according to the fitting model at frequency ω i are to be optimized parameters, w1 and w2 represent weights, which can be dynamically adjusted according to measurement errors;

[0105] The Levenberg-Marquardt algorithm is used for iterative calculation to obtain the optimized parameters, wherein the iterative formula is as follows:

[0106]

[0107] In the formula: represents the Jacobian matrix, each row of which corresponds to the partial derivative of frequency ω i represents the residual vector, represents the damping factor, T represents the matrix transpose, k represents the iteration step, and diag represents the operation of extracting the diagonal elements of the matrix and constructing a diagonal matrix.

[0108] In some embodiments, the Jacobian matrix is as follows, each row of which corresponds to the partial derivative of frequency ω i .

[0109]

[0110] In some embodiments, the Jacobian matrix calculation and matrix inversion operation are distributed to GPU parallel processing. Through GPU acceleration, the traditional fitting process of several hours is shortened to minutes, meeting the needs of real-time diagnosis on site.

[0111] As shown in Figure 2 , a schematic diagram of fitting of the imaginary part of the complex permittivity dielectric spectrum of the high-voltage switch cabinet insulation medium according to an exemplary embodiment of the application is shown.

[0112] In some embodiments, the method further comprises verifying the model parameters by the following formula:

[0113]

[0114] In the formula: N represents the length of the residual sequence, e i represents the ith residual, which is specifically defined as follows:

[0115] ​​​​​For the real residual:

[0116] For the imaginary residual: ;

[0117] when If a systematic error is detected, the parameters should be reinitialized or the model adjusted.

[0118] In some implementations, the relative permittivity (real part) and dielectric loss factor (imaginary part) obtained by dielectric spectrum fitting in S8 are compared with the corresponding parameters in the dielectric model library to determine the insulation state of the insulating material of the high-voltage switchgear; wherein, the dielectric model library stores the model parameters after historical fitting.

[0119] In some embodiments, determining the insulation state of the insulating material of the high-voltage switchgear includes:

[0120] for 10Hz to 10 2 The relative permittivity of Hz, after the comparison, is determined to be aging of the insulating material when its increase exceeds the first threshold.

[0121] for 10Hz to 10 2 If the dielectric loss factor of Hz increases beyond the second threshold after the comparison, it is determined that the insulating material is damp.

[0122] In some implementations, the first threshold is 20% and the second threshold is 30%.

[0123] On the other hand, embodiments of the present invention also provide a computer device including a memory and a processor, the memory storing computer-readable instructions, wherein the processor executes the computer-readable instructions to implement the method or any processing of any of the embodiments or examples herein.

[0124] The above describes in detail the method for rapid fitting of dielectric spectrum of high voltage switchgear insulation medium according to the embodiments of the present invention. It should be understood that some or all of the operations of the method can be programmed into a program product, the program of which can be stored on a storage medium and executed by a computer or similar device. Figure 3 The functional architecture of a computer device for performing a rapid fitting method for the dielectric spectrum of high-voltage switchgear insulation media according to an exemplary embodiment of the present invention is shown.

[0125] Reference Figure 3The computer device 100 comprises at least a processor 101 and a memory 102, wherein the memory 102 can be an example of a storage medium, and computer programs (or computer readable instructions) are stored on the memory 102, and the processor 101 executes the computer programs to perform some or all of the operations in the method according to any one of the embodiments of the present application.

[0126] In some embodiments, the computer device 100 can further comprise a data storage device 103, a display 104, a speaker 105, and a communication module 106. The processor 101, the memory 102, and the data storage device 103 communicate with each other through a bus, and communicate with peripheral devices such as the display 104 and the communication module 106 through the bus and an I / O module 107.

[0127] Specifically, the data storage device 103 can store application programs and various configuration files and data thereof. The memory 102 stores computer programs. The processor 101 executes the computer programs to perform various processes, operations, or steps of the method for rapid fitting of dielectric spectrum of high-voltage switch cabinet insulation medium, for example, including but not limited to: calling the communication module 106 to communicate with a broadband dielectric spectrometer to obtain measured values of dielectric response data; pre-processing the collected dielectric response data to obtain pre-processed dielectric spectrum data; selecting one of a first model and a second model as an initial dielectric response model according to the type of the insulation material, wherein the first model and the second model are different; initializing parameters of the initial dielectric response model based on the pre-processed dielectric spectrum data by a frequency domain feature method to obtain predicted values of complex dielectric constant; calculating an adjustment decision coefficient according to the predicted values and the measured values of the complex dielectric constant; adjusting the dielectric response model according to the calculated adjustment decision coefficient; inputting the obtained dielectric response data into the adjusted dielectric response model for dielectric spectrum fitting processing; or further displaying the fitting image on the display 104 or sending the fitting image to other systems, devices, or equipment for display through communication connection.

[0128] In optional embodiments, the communication module 106 can be omitted in the case where communication with the other systems, devices, or equipment is not required. In some embodiments, the computer device 100 can further comprise a speaker 105 for voice output.

[0129] In some embodiments, the computer device can be integrated with a broadband dielectric spectrometer. Alternatively, in some embodiments, the broadband dielectric spectrometer can be an example of the computer device of the present application, which has the function of performing the method for rapid fitting of dielectric spectrum of high-voltage switch cabinet insulation medium and related configurations.

[0130] In some embodiments, the processor 101 can include any suitable semiconductor-based electronic processing unit, chip, microchip, or integrated circuit (IC). The memory 102 is a programmable memory that can include any suitable electronic storage device configured to store instructions and can be reprogrammed. For example, the programmable memory can include an erasable programmable read-only memory (EPROM) device. The data storage 103 can be a persistent storage that can include any suitable electronic memory configured to retain stored information when power is cycled. For example, the data storage 103 can include a hard disk drive, a solid state drive (SSD), a flash drive, a hybrid drive, or the like, or any combination of these.

[0131] Accordingly, the processor 101 can control the output of information on the I / O module 107 by storing information to the memory and / or executing programs / instructions stored in the memory. For example, aspects of the methods described herein can be performed by the processor 101 according to programs / instructions stored in the memory (e.g., the memory 102 and / or the data storage 103).

[0132] Also, the processor 101 can be in electronic communication with the I / O module 107 and / or the communication module 106 to receive or transmit relevant instructions and information. The I / O module 107 can include any suitable mechanical or virtual user interface configured to enable a user to interact information with the computer device 100 or to perform one or more functions of the computer device 100 itself, such as a graphical user interface (GUI) on a screen or other display. In some instances, the user interface can include a voice interface capable of voice recognition by which an operator can provide voice commands to the processor.

[0133] The communication module 106 can include any suitable device and / or structure configured to facilitate the exchange of information between the computer device 100 and external electronic devices. The communication module 106 can include a device configured to transmit and / or receive wireless or wired information with other devices. For example, the communication module 106 can include an antenna, a transceiver, a connector for wired reception and / or transmission of data, a data exchange device, or the like, or any combination of these. The communication module 106 can also include ancillary elements, such as a filter circuit, an encryption / decryption circuit, and / or an integrated circuit (IC) chip for processing signals (e.g., a Bluetooth® chip). In some embodiments, the communication module 106 can include a WiFi device configured to connect to a local wireless network.

[0134] In some embodiments, the computer device 100 can include a smartphone, a wearable computer, a portable / mobile electronic device, a tablet computer, a smartwatch, a personal digital assistant (PDA), a personal computer (PC), a desktop computer, a notebook computer, a server, etc. The computer device 100 can include or be installed with one or more application programs (APPs), where one APP is configured to perform the high voltage switchgear insulation medium dielectric spectrum fast fitting method described herein.

[0135] Although not shown, it should be understood that the computer device 100 also includes a power supply component, which can include any suitable devices and / or structures configured to provide an electrical interface between the computer device 100 and a power source. The power source can include any suitable source of electrical energy, such as a battery, a plug, a capacitor, a fuel cell, etc., or any combination of these. Additionally or alternatively, the power source can be included in the power supply component. For example, a battery or battery pack can be included within the computer device 100. In some embodiments, the battery can be rechargeable, such as by an interface provided by a cable or the power supply component. In some embodiments, the power supply component can share features or duplicate features of the communication module 106. For example, a USB or micro-USB cable connector can be included in the computer device 100, such that power supply or data communication is performed by the same component.

[0136] Those skilled in the art will understand that the above-disclosed embodiments are merely exemplary of the application and do not limit the scope of the application. Various modifications, changes, substitutions, and alternatives can be made or adapted without departing from the spirit and scope of the application as described in the application and its accompanying claims.

Claims

1. A high voltage switchgear insulation medium dielectric spectrum rapid fitting method, characterized in that, The method comprises the following processes: S1, applying an alternating current to the insulating material of the high-voltage switch cabinet, the frequency of the alternating current being ; S2, obtaining dielectric response data of the insulating material of the high-voltage switch cabinet, the dielectric response data comprising a measured value of a complex dielectric constant, the complex dielectric constant comprising a real part and an imaginary part; S3, preprocessing the collected dielectric response data to obtain preprocessed dielectric spectrum data; S4, selecting one of a first model and a second model as an initial dielectric response model according to the type of the insulating material, the first model being different from the second model; S5, initializing parameters of the initial dielectric response model based on the preprocessed dielectric spectrum data to obtain a predicted value of the complex dielectric constant; S6, calculating an adjustment decision coefficient according to the predicted value and the measured value of the complex dielectric constant according to the following formula: wherein, represents the adjusted determination coefficient, p represents the number of model parameters, M represents the number of frequency points in the dielectric spectroscopy measurement, and R represents the residual vector whose elements are the differences between the measured and predicted values; S7, adjusting the dielectric response model according to the calculated adjustment decision coefficient; S8, inputting the obtained dielectric response data into the adjusted dielectric response model for dielectric spectrum fitting processing.

2. The method of claim 1, wherein, In process S3, the preprocessing of the collected dielectric response data comprises: A wavelet threshold denoising method is adopted, a Morlet wavelet basis is selected to perform multi-scale decomposition on the original signal corresponding to the collected dielectric response data, high-frequency noise components are suppressed by a soft threshold function, and effective signals are retained to obtain the preprocessed dielectric spectrum data.

3. The method of claim 1, wherein: the first model is a single relaxation Debye model, and an expression of the first model is as follows: the second model is a Cole-Cole model, and an expression of the second model is as follows: wherein denotes the high-frequency limit dielectric constant, denotes the static dielectric constant, denotes the relaxation time, j denotes the imaginary unit in complex numbers, denotes the distribution parameter, 0 < a < 1, ≤ 1.

4. The method of claim 3, wherein, In process S5, the parameters of the initial dielectric response model are initialized by a frequency domain feature method, which comprises: From the pre-processed dielectric spectrum data, the average value of the real part of the complex dielectric constant in a low frequency band is extracted as , the low frequency band including Hz to 1 Hz; From the pre-processed dielectric spectrum data, an average value of the real part of the complex dielectric constant in a high frequency band is extracted as , the high frequency band including Hz to Hz; By loss peak frequency Estimate relaxation time ; For the Cole-Cole model, the distribution parameters are initialized to 0.

8.

5. The method of claim 4, wherein, The method further comprises, before process S7, optimizing the model parameters by using a Levenberg-Marquardt algorithm; wherein the optimization comprises: a target function is set as a weighted residual sum of squares of the measured value and the predicted value: wherein, is the objective function, which is expressed in terms of a weighted sum of squared residuals, is the real part of the complex permittivity at frequency ω i under test is the imaginary part of the complex permittivity at frequency ω i under test is the real part of the complex permittivity at frequency ω i calculated from the fitted model, is the imaginary part of the complex permittivity at frequency ω i calculated from the fitted model, m represents the number of test frequencies, For the parameters to be optimized, w1 and w2 represent weights, which are dynamically adjusted according to the measurement error; the Levenberg-Marquardt algorithm is used for iterative calculation to obtain the optimized parameters, and an iterative formula is as follows: wherein: denotes the Jacobian matrix, each row of which corresponds to a frequency point ω i of the partial derivative with respect to Represents the residual vector. Indicates the damping factor. T represents matrix transposition, k represents an iteration step, diag represents an operation of extracting diagonal elements of a matrix and constructing a diagonal matrix.

6. The method of claim 5, wherein, The Jacobian matrix calculation and matrix inversion operation are distributed to GPU parallel processing.

7. The method of claim 5, wherein, The method further comprises: the model parameters are verified by the following formula: wherein: N represents the length of a residual sequence, e i represents the i-th residual, which is defined as follows: For the real part residual: For the imaginary part residual: ; When a system error is determined to exist, the parameters are reinitialized or the model is adjusted.

8. The method of claim 3, wherein, in process S7: When Accept the selected dielectric response model if greater than or equal to a set value. When When the selected dielectric response model is the Cole-Cole model, an asymmetry coefficient β is introduced, extending the model to the HN function model as follows: when the parameters are initialized, β = 0.5, and the remaining parameters inherit the Cole-Cole fitting results; When When the selected dielectric response model is the Debye model, the multi-scale polarization is characterized by superimposing a multi-relaxation term according to the following equation: At the parameter initialization, the number of relaxation terms K is determined according to the number of loss peaks in the frequency spectrum curve of the imaginary part of the complex dielectric constant, and the relaxation time is inversely deduced from the loss peak frequency: .

9. The method of claim 1, wherein, The method further comprises: the real part and the imaginary part of the complex dielectric constant obtained by the dielectric spectrum fitting processing in process S8 are compared with corresponding parameters in a dielectric model library to determine the insulation state of the insulating material of the high-voltage switch cabinet; wherein the dielectric model library stores historical model parameters after fitting.

10. The method of claim 9, wherein, The determination of the insulation state of the insulating material of the high-voltage switch cabinet comprises: For a real part of the complex dielectric constant of 10 Hz to 10 2 Hz, an increase beyond a first threshold value, as determined by the comparison, is an indication of aging of the insulating member material; For a complex dielectric constant imaginary part of 10 Hz to 10 2 Hz, when the increase exceeds a second threshold value through the comparison, it is determined that the insulation material is damp.

11. A computer device comprising a memory and a processor, The memory stores computer readable instructions, characterized in that, the processor executes the computer readable instructions to implement the method of any one of claims 1 to 10 or any one of the processes in the method.

Images