GIS partial discharge digital twin rapid forward method and system and medium

Abstract

The present application relates to the technical field of electromagnetic field digital simulation, and particularly relates to a GIS partial discharge digital twin rapid forward method and system and a medium, the method comprising: constructing a differentiated equivalent source model according to the discharge mechanism of different partial discharge defects; constructing a GIS cavity digital twin model, precomputing the transfer function from multiple discrete source positions to sensors, and constructing a spatial transfer function library; obtaining GIS operating environment parameters and insulation aging state parameters, correcting electromagnetic parameters and updating the spatial transfer function library; obtaining a target source position and a target defect type, retrieving and spatially interpolating to obtain a target transfer function, and calling the differentiated equivalent source model to generate a target source signal; and convolving the target source signal with the target transfer function to obtain a partial discharge electromagnetic wave forward signal. Through the present application, the problem that the existing GIS partial discharge forward model is difficult to efficiently maintain the authenticity of electromagnetic response under dynamic operating conditions is effectively solved.

Description

Technical Field

[0001] This invention relates to the field of electromagnetic field digital simulation technology, and in particular to a rapid forward modeling method, system and medium for GIS partial discharge digital twin. Background Technology

[0002] Gas-insulated switchgear utilizes a sealed metal cavity, insulating gas, and solid insulating components to achieve insulation and breaking of high-voltage electrical equipment. Partial discharge is an early electrical phenomenon in the formation and deterioration process of insulation defects. The ultra-high frequency electromagnetic waves generated by partial discharge can be received by sensors placed on the equipment and used for defect identification, fault location, and condition assessment. Since it is difficult to obtain and accurately label a large number of partial discharge samples in the field, existing technologies typically rely on electromagnetic simulation or digital twin models to generate virtual signals. CN121257128A discloses the establishment of a digital twin model based on GIS structure, operating parameters, and abnormal operating conditions, and the simulation of different partial discharge fault scenarios to generate diagnostic data; CN118226211A discloses the establishment of a digital twin model based on historical partial discharge data, and the updating of the model in conjunction with current detection data to obtain partial discharge identification results.

[0003] However, existing solutions are mainly geared towards fault identification or model result correction. They typically do not adequately distinguish the electromagnetic radiation mechanisms of different defects during forward modeling and require repeated electromagnetic calculations for different source locations. Furthermore, the model parameters are difficult to change synchronously with the operating environment and insulation aging state. Therefore, it is difficult to enable the GIS partial discharge forward modeling model to efficiently maintain the authenticity of the electromagnetic response under dynamic operating conditions.

[0004] The information disclosed in this background section is intended only to enhance the understanding of the general background of this disclosure and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0005] This invention provides a rapid forward modeling method, system, and medium for GIS partial discharge digital twins, which can effectively solve the problems in the background art.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A fast forward modeling method for GIS partial-exposure digital twins, the method comprising: Based on the discharge mechanism of different partial discharge defects, a differentiated equivalent source model corresponding to the defect type is constructed; A digital twin model of a GIS cavity is constructed, and the transfer functions from multiple discrete source locations to sensors are pre-calculated based on the digital twin model of the GIS cavity, and a spatial transfer function library is constructed. Obtain GIS operating environment parameters and insulation aging state parameters, correct the electromagnetic parameters of the GIS cavity digital twin model based on the operating environment parameters and insulation aging state parameters, and update the spatial transfer function library; Obtain the target source location and target defect type; retrieve and spatially interpolate the updated spatial transfer function library based on the target source location to obtain the target transfer function; and call the corresponding differentiated equivalent source model to generate the target source signal based on the target defect type. The target source signal is convolved with the target transfer function to obtain the forward modeling signal of the partial discharge electromagnetic wave received by the sensor.

[0007] Furthermore, the differentiated equivalent source model includes a time-varying point dipole source corresponding to corona discharge, a line source array distributed along the discharge path on the insulator surface corresponding to surface discharge, and a dielectric-enclosed source corresponding to internal air gap discharge. The time-varying point dipole source, the line source array, and the dielectric-enclosed source each have time-domain waveforms, spatial distribution characteristics, and electromagnetic wave coupling characteristics that match the corresponding discharge mechanisms.

[0008] Furthermore, the GIS cavity digital twin model includes a metal shell cavity, conductors and their supporting structures, gas chamber partitions, insulators, gas insulating medium domains, and sensor installation positions, and configures the electromagnetic parameters corresponding to the metal shell cavity, the insulators, and the gas insulating medium domains.

[0009] Furthermore, the construction of the space transfer function library includes: In the GIS cavity digital twin model, multiple discrete source locations are set to cover the free space region, the region near the insulator, and the conductor edge region; By applying broadband reference excitation at each of the discrete source locations and performing electromagnetic full-wave calculations, the time-domain impulse response from each of the discrete source locations to each of the sensors is obtained. The time-domain impulse response is transformed in the frequency domain to obtain the corresponding frequency-domain transfer function, and the frequency-domain transfer function is associated with and stored with the corresponding discrete source location and the sensor to form the spatial transfer function library.

[0010] Furthermore, obtaining the target transfer function and generating the partial discharge electromagnetic wave forward modeling signal include: Based on the target source location, multiple neighboring discrete source locations are retrieved from the spatial transfer function library using a spatial index; The target transfer function is obtained by performing inverse distance weighted interpolation on the frequency domain transfer function corresponding to the location of the nearest discrete source. The target source signal is generated according to the target defect type, and the target source signal is convolved with the target transfer function by a fast Fourier transform to obtain the forward modeling signal of the partial discharge electromagnetic wave.

[0011] Furthermore, the electromagnetic parameter correction of the GIS cavity digital twin model includes: The temperature and pressure of the gas insulating medium domain are periodically acquired through a GIS online monitoring system to form the operating environment parameters; Based on the pre-defined mapping relationship between temperature, pressure and gas dielectric parameters, the dielectric parameters of the gas insulating medium domain are corrected using the operating environment parameters. The spatial transfer function library is updated when the dielectric parameters of the modified gas insulating medium domain meet the preset update conditions.

[0012] Furthermore, the electromagnetic parameter correction of the GIS cavity digital twin model also includes: The cumulative discharge energy is determined based on the partial discharge signal, and the cumulative discharge energy is used to characterize the insulation aging state parameters. Based on the pre-calibrated mapping relationship between the cumulative discharge energy and the dielectric loss parameters, the dielectric loss parameters of the insulator are corrected. When the corrected dielectric loss parameters meet the preset update conditions, the frequency domain transfer function corresponding to the region near the insulator in the spatial transfer function library is updated.

[0013] Furthermore, it also includes: By changing the location of the partial discharge source, the type of partial discharge defect, the operating environment parameters, and the noise conditions, the forward modeling signals of the partial discharge electromagnetic waves are generated in batches. The partial discharge source location, the partial discharge defect type, the model parameters of the differentiated equivalent source model, the operating environment parameters, the version identifier of the spatial transfer function library, and the time-domain waveform and frequency-domain spectrum corresponding to each sensor are associated and stored to form a traceable partial discharge electromagnetic wave forward modeling signal dataset.

[0014] A rapid forward modeling system for GIS partial-field digital twins, the system comprising: The differentiated source modeling module constructs differentiated equivalent source models corresponding to different partial discharge defects based on their discharge mechanisms. The transfer function construction module builds a digital twin model of the GIS cavity and pre-calculates the transfer functions from multiple discrete source locations to the sensors based on the digital twin model of the GIS cavity, thus constructing a spatial transfer function library. The twin parameter update module acquires GIS operating environment parameters and insulation aging status parameters, corrects the electromagnetic parameters of the GIS cavity digital twin model based on the operating environment parameters and insulation aging status parameters, and updates the spatial transfer function library. The forward modeling element acquisition module obtains the target source location and target defect type, searches and performs spatial interpolation on the updated spatial transfer function library based on the target source location to obtain the target transfer function, and calls the corresponding differentiated equivalent source model to generate the target source signal based on the target defect type. The forward modeling signal generation module convolves the target source signal with the target transfer function to obtain the forward modeling signal of the partial discharge electromagnetic wave received by the sensor.

[0015] A computer-readable storage medium storing a computer program, the computer program including program instructions that, when executed by a processor, can implement the GIS partial discharge digital twin rapid forward modeling method.

[0016] The technical solution of this invention can achieve the following technical effects: By constructing differentiated equivalent source models for different partial discharge defects and a GIS cavity spatial transfer function library, and adaptively correcting model parameters based on the operating environment and insulation aging state, and by retrieving and interpolating the transfer function at the target location and convolving it with the corresponding source signal, rapid adaptive forward modeling of GIS partial discharge electromagnetic waves is achieved. This effectively solves the problem that existing GIS partial discharge forward modeling models cannot efficiently maintain the authenticity of electromagnetic response under dynamic operating conditions.

[0017] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

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

[0019] Figure 1 A schematic diagram of a rapid forward modeling method for partial-exposure digital twins in GIS; Figure 2 Schematic diagram of equivalent source models for the physical discharge mechanism of different defect types; Figure 3Flowchart of the method for pre-calculation and fast forward modeling of cavity space transfer function; Figure 4 A closed-loop flowchart for real-time self-correction of environmental data-driven model parameters. Detailed Implementation

[0020] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0022] Example 1; like Figure 1 As shown, this application provides a fast forward modeling method for GIS partial-exposure digital twins, the method including: S10: Based on the discharge mechanism of different partial discharge defects, construct differentiated equivalent source models corresponding to the defect types; S20: Construct a digital twin model of the GIS cavity, and pre-calculate the transfer functions from multiple discrete source locations to the sensors based on the digital twin model of the GIS cavity, and construct a spatial transfer function library; S30: Obtain GIS operating environment parameters and insulation aging status parameters, correct the electromagnetic parameters of the GIS cavity digital twin model based on the operating environment parameters and insulation aging status parameters, and update the spatial transfer function library; S40: Obtain the target source location and target defect type, search and spatially interpolate the updated spatial transfer function library based on the target source location to obtain the target transfer function, and call the corresponding differentiated equivalent source model to generate the target source signal based on the target defect type; S50: Convolve the target source signal with the target transfer function to obtain the forward modeling signal of the partial discharge electromagnetic wave received by the sensor.

[0023] Specifically, this embodiment takes a 550kV gas-insulated switchgear as the object. The equipment includes a circuit breaker, disconnector, grounding switch, current transformer, busbar and basin insulator. A digital twin model of the GIS cavity is established using the COMSOL Multiphysics RF module. The simulation frequency band is 300MHz to 3GHz. Four ultra-high frequency sensors are set in the cavity. The center frequency of the sensors is 1.0GHz and the bandwidth is 0.3GHz to 2.0GHz. Based on the discharge mechanisms of corona discharge, surface discharge, and internal air gap discharge, corresponding differentiated equivalent source models are constructed. Corona discharge occurs at the tip of a metallic conductor needle. The discharge channel size is much smaller than the wavelength of ultra-high frequency electromagnetic waves. Corona discharge can be equivalently represented as a time-varying point dipole, with the dipole moment being: ; in, Represents the time-varying dipole moment. Represents time-varying charge. This indicates the equivalent dipole arm length; the corona discharge uses a unipolar Gaussian pulse with a rise time of less than 1ns and a pulse width of less than 5ns, with the main peak of the spectrum concentrated in the range of 300MHz to 800MHz. Surface discharge propagates along the insulator surface. It can be equivalently represented as a line source array distributed along the discharge path. The line source array consists of eight equally spaced radiating elements. Each radiating element uses a damped oscillating pulse with a rise time greater than 5 ns and a pulse width greater than 20 ns, with an oscillation frequency of 10 MHz to 50 MHz. The radiation directivity coefficient of the line source array is: ; in, This indicates the angle between the radiation direction and the normal to the insulator surface, where the normal to the insulator surface is the direction of strongest radiation. Internal air gap discharge occurs inside the epoxy resin insulator. This internal air gap discharge can be considered equivalent to a spherical, enclosed source surrounded by a dielectric material. Electromagnetic waves couple to the GIS cavity through the dielectric boundary between the epoxy resin and SF6 gas. The dielectric boundary transmission coefficient is: ; in, Indicates the SF6 gas wave impedance. The complex wave impedance of epoxy resin and the approximate gas wave impedance of SF6 are as follows: ; in, The dielectric constant of SF6 gas is represented; the internal air gap discharge uses an oscillating decay wave with an oscillation frequency of 10MHz to 50MHz and a pulse group duration of 50ns to 200ns, and the high-frequency components are attenuated by the epoxy resin dielectric loss. A scaled-down 3D model was created based on the actual geometric dimensions of the GIS. The model includes the inner cavity of the metal casing, the mounting slot for the basin-type insulator, the conductor support, the gas chamber baffle, the epoxy resin insulator, and the SF6 gas filling area. The sensor installation positions were set in the model coordinate system. The conductivity of the metal casing was set to 5.8 × 10⁻⁶. 7S / m, the dielectric constant of epoxy resin insulator is set to 4.2, the reference loss tangent is set to 0.005, and the dielectric constant of SF6 gas under 20℃ and 0.5MPa conditions is set to 1.0026; the global mesh size is no greater than one-tenth of the minimum wavelength corresponding to the highest simulation frequency, and the mesh size of the three-phase interface region of metal, insulator and gas is refined to one-twentieth of the minimum wavelength. Inside the GIS cavity, 1200 discrete source positions are set at 2cm intervals. Among them, 800 discrete source positions are set in the free space area, 250 discrete source positions are set in the area less than 5cm from the insulator surface, and 150 discrete source positions are set in the area less than 3cm from the conductor surface. A broadband unit pulse covering the frequency band of 300MHz to 3GHz was sequentially applied at each discrete source location, and a three-dimensional finite element full-wave simulation was performed to obtain the discrete source locations. r i To the j Time-domain impulse response of each sensor h ij ( t The frequency domain transfer function is obtained by performing a fast Fourier transform on the time-domain impulse response. ; in, Indicates the amplitude response. Represents the phase response; the group delay is: ; The 4800 sets of frequency domain transfer functions associated with 1200 discrete source locations and four UHF sensors are stored, and a three-dimensional KD-Tree spatial index is established with the coordinates of the discrete source locations to form a spatial transfer function library. During GIS operation, the air chamber temperature is acquired every three minutes via the SCADA interface. T ( t and pressure P ( t This forms the operating environment parameters; the dielectric constant of SF6 gas is dynamically corrected based on temperature and pressure. ; in, Indicates the reference dielectric constant. , and These represent the temperature correction factor, pressure correction factor, and temperature-pressure interaction correction factor, respectively. Each correction factor is obtained by least-squares fitting of measurement data from a standard experimental chamber under multiple temperature and pressure conditions. The cumulative discharge energy is determined based on the partial discharge process: ; in, , and These represent the discharge voltage, discharge current, and discharge duration corresponding to the k-th discharge, respectively. The discharge voltage and discharge current are obtained by inverse calculation using the amplitude of the ultra-high frequency signal. The cumulative discharge energy is used to characterize the insulation aging state, and the loss tangent of the epoxy resin insulator is updated. ; in, Indicates the reference loss tangent. This represents the maximum change in the loss tangent. This represents the energy characteristic parameters calibrated through accelerated aging experiments; When the relative rate of change of the dielectric constant of SF6 gas satisfies: ; Recalculate the frequency domain transfer function corresponding to all discrete source locations; when the relative rate of change of the loss tangent of the epoxy resin insulator satisfies: ; Only the frequency domain transfer function corresponding to the region near the insulator is recalculated; after the update is completed, a new version of the spatial transfer function library is generated. When performing forward modeling of partial discharge electromagnetic waves, the target source location and target defect type are input. The six discrete source locations closest to the target source location are retrieved using a three-dimensional KD-Tree, and the inverse distance weighted interpolation weights are calculated based on the distances between the target source location and each discrete source location. ; in, The distance between the target source location and the kth nearest discrete source location is represented by K, where K represents the number of nearest discrete source locations involved in the interpolation. Spatial interpolation is performed on the amplitude response and phase response using interpolation weights to synthesize the target transfer function from the target source location to each sensor. Based on the target defect type, the corresponding differentiated equivalent source model is invoked to generate the target source signal, and the target source signal is then fast-convolved with the target transfer function: ; in, Indicates the target source signal. This represents the target transfer function from the target source location to the sensor. This represents the forward modeling signal of the partial discharge electromagnetic wave received by the sensor; through this process, the time-domain waveform and frequency-domain spectrum corresponding to each ultra-high frequency sensor are obtained.

[0024] The technical solution of this invention constructs differentiated equivalent source models and GIS cavity space transfer function libraries for different partial discharge defects, adaptively corrects model parameters based on operating environment and insulation aging state, and convolves the transfer function of the target location with the corresponding source signal after retrieval and interpolation, thereby realizing rapid adaptive forward modeling of electromagnetic waves of GIS partial discharge. This effectively solves the problem that existing GIS partial discharge forward modeling models are unable to maintain the authenticity of electromagnetic response efficiently under dynamic operating conditions.

[0025] Furthermore, such as Figure 2 As shown, the differentiated equivalent source model includes a time-varying point dipole source corresponding to corona discharge, a line source array distributed along the discharge path on the insulator surface corresponding to surface discharge, and a dielectric-enclosed source corresponding to internal air gap discharge. The time-varying point dipole source, the line source array, and the dielectric-enclosed source have time-domain waveforms, spatial distribution characteristics, and electromagnetic wave coupling characteristics that match the corresponding discharge mechanisms, respectively.

[0026] As a preferred embodiment of the above, a corresponding differentiated equivalent source model is selected according to the type of partial discharge defect, and the spatial structure, time-domain waveform and radiation parameters of the equivalent source are configured according to the location of discharge occurrence, the shape of discharge channel and the boundary conditions of the medium. When the defect type is corona discharge, the partial discharge source is configured as a time-varying point dipole source located at the tip of a metal conductor needle. The length of the corona discharge channel is much smaller than the wavelength of ultra-high frequency electromagnetic waves, so a concentrated point source is used to describe the discharge region, and the dipole axis is determined according to the needle tip direction. The point dipole source uses a unipolar Gaussian pulse with a pulse rise time of less than 1 ns, a pulse width of less than 5 ns, and a main peak in the spectrum located between 300 MHz and 800 MHz. The model parameters of the point dipole source include dipole moment amplitude, pulse rise time, pulse width, and dipole direction. When the defect type is surface discharge, the partial discharge source is configured as a line source array distributed along the discharge path on the insulator surface. The discharge path is determined according to the shape of the insulator surface, and eight radiating elements are set at equal intervals along the discharge path. Each radiating element is arranged sequentially according to the development direction of the surface discharge to characterize the spatial extension of the discharge channel. Each radiating element uses a damped oscillating pulse with a pulse rise time greater than 5ns, a pulse width greater than 20ns, and an oscillation frequency of 10MHz to 50MHz. The radiation intensity varies with the angle between the radiation direction and the normal to the insulator surface. The radiation intensity is the largest in the direction normal to the insulator surface, and the radiation intensity along the tangential direction of the surface is about 30% of the normal radiation intensity. The model parameters of the line source array include the number of radiating elements, the discharge path length, the radiation directionality parameter, the oscillation frequency, the pulse rise time, and the pulse width. When the defect type is internal air gap discharge, the partial discharge source is configured as a dielectric-encapsulated spherical closed source located inside the epoxy resin insulator; the air gap diameter is set to 0.5mm to 3mm. The electromagnetic waves generated by the spherical closed source propagate through the dielectric boundary between the air gap and the epoxy resin and are further coupled to the SF6 gas-filled domain. The reflection and transmission effects of the dielectric boundary on the electromagnetic waves are considered in the modeling, and the additional attenuation of high-frequency components in the dielectric is characterized by the epoxy resin dielectric constant and loss tangent. The dielectric-encapsulated spherical closed source uses an oscillating decay wave with an oscillation frequency of 10MHz to 50MHz and a pulse group duration of 50ns to 200ns. The model parameters of the dielectric-encapsulated spherical closed source include the air gap radius, epoxy resin dielectric constant, epoxy resin loss tangent, oscillation frequency, and decay time constant. After completing defect type identification or defect type input, the corresponding model is called from the preset set of differentiated equivalent source models, the target source position is assigned to the corresponding equivalent source, and the target source signal is generated according to the corresponding model parameters for subsequent transfer function interpolation and convolution forward modeling.

[0027] Furthermore, the GIS cavity digital twin model includes the inner cavity of the metal shell, conductors and their supporting structures, gas chamber partitions, insulators, gas insulating medium domains, and sensor installation locations, and configures the electromagnetic parameters corresponding to the inner cavity of the metal shell, insulators, and gas insulating medium domains.

[0028] As a preferred embodiment of the above, a digital twin model of the GIS cavity is established based on a real GIS whole-bay equipment; a unified three-dimensional coordinate system is constructed according to the equipment design drawings and physical dimensions, and the inner cavity of the metal shell, the central conductor, the conductor support, the insulator mounting groove, the gas chamber partition, the epoxy resin insulator and the SF6 gas filling domain are established in the unified coordinate system so that the size, relative position and connection relationship of each component are consistent with the actual equipment. The inner cavity of the metal shell is constructed according to the actual air chamber outline, retaining the local cavity features formed by the insulator mounting slot, tee connection structure, and air chamber partition. The central conductor and conductor support are arranged according to the actual installation position to characterize the reflection and propagation of electromagnetic waves on the conductor surface, the inner wall of the shell, and the structural turning areas. The epoxy resin insulator is modeled using a solid dielectric domain, and the configurable positions of the creepage path on the insulator surface and the internal air gap defects are retained. The SF6 gas filling domain is jointly enclosed by the metal shell, central conductor, conductor support, air chamber partition, and epoxy resin insulator, and is used to simulate the propagation of partial discharge electromagnetic waves in the GIS cavity. Establish UHF sensor mounting holes at the corresponding positions in the GIS cavity, and arrange the sensor receiving surface or receiving point inside the mounting holes. Record the spatial coordinates of each sensor in a unified coordinate system to form a definite spatial correspondence between the discrete source position and each sensor. Conductive material properties are configured for the metal casing and the center conductor, and frequency-dependent impedance boundaries are used to describe the propagation loss of high-frequency electromagnetic waves on the metal surface; dielectric constant and loss tangent are configured for the epoxy resin insulator; dielectric constant is configured for the SF6 gas-filled domain. After completing the geometric model and material domain configuration, the global mesh size is determined based on the highest simulation frequency, and local mesh refinement is performed in the interface region of metal, insulator and gas and the region near the insulator to ensure the accuracy of electromagnetic field calculation at complex boundaries. Subsequently, the electromagnetic field solution boundary, the local discharge power source loading position and the sensor signal extraction position are set to form a GIS cavity digital twin model for transfer function pre-calculation and forward modeling of partial discharge electromagnetic waves.

[0029] Furthermore, such as Figure 3 As shown, the construction of the space transfer function library includes: In the GIS cavity digital twin model, multiple discrete source locations are set up to cover the free space region, the region near the insulator, and the conductor edge region; By applying broadband reference excitation at each discrete source location and performing electromagnetic full-wave calculations, the time-domain impulse response from each discrete source location to each sensor is obtained. The time-domain impulse response is transformed in the frequency domain to obtain the corresponding frequency-domain transfer function. The frequency-domain transfer function is then associated with and stored with the corresponding discrete source location and sensor to form a spatial transfer function library.

[0030] As a preferred embodiment of the above, a discrete source location grid is set inside the GIS cavity digital twin model. The discrete source locations cover the gas free space, the area near the insulator, and the conductor edge area to characterize the propagation differences of partial discharge electromagnetic waves in different structural areas. Each discrete source location is represented by a unified three-dimensional coordinate system and a corresponding relationship is established with the sensor location. A broadband unit pulse covering the simulation frequency band is loaded at each discrete source location, and a three-dimensional finite element full-wave simulation is executed sequentially. The time-domain impulse response generated when the unit pulse propagates from the current discrete source location to each sensor is recorded. During the simulation, the GIS cavity structure, material parameters, boundary conditions and sensor settings are kept consistent so that the propagation response corresponding to different discrete source locations has a unified benchmark. Fast Fourier transform is performed on each time-domain impulse response to obtain the frequency domain transfer function from the discrete source location to each sensor, and the amplitude response, phase response and group delay are extracted; each set of frequency domain transfer functions is associated with the discrete source location coordinates, sensor identifier and electromagnetic parameter status for retrieval according to source location and sensor; A KD-Tree spatial index is constructed using the three-dimensional coordinates of discrete source locations. The spatial index is then mapped to the corresponding frequency domain transfer function to form a spatial transfer function library. The spatial transfer function library is generated using an offline pre-calculation method, and full-wave simulations of each discrete source location can be executed in parallel. After the pre-calculation is completed, any target source location can retrieve neighboring discrete source locations through the spatial index, providing a foundation for subsequent spatial interpolation of transfer functions and rapid forward modeling of partial discharge electromagnetic waves.

[0031] Furthermore, such as Figure 3 As shown, the acquisition of the target transfer function and the generation of the forward modeling signal of the partial discharge electromagnetic wave include: Based on the target source location, multiple nearby discrete source locations are retrieved from the spatial transfer function library using a spatial index; The target transfer function is obtained by performing inverse distance weighted interpolation on the frequency domain transfer function corresponding to the location of the nearest discrete source. The target source signal is generated based on the target defect type, and the target source signal is convolved with the target transfer function by a fast Fourier transform to obtain the forward modeling signal of the partial discharge electromagnetic wave.

[0032] As a preferred embodiment of the above, when performing electromagnetic wave forward modeling of partial discharge, the target source location coordinates and target defect type are input, the target source location coordinates are used as spatial retrieval conditions to input the KD-Tree spatial index, the multiple discrete source locations closest to the target source location are retrieved from the spatial transfer function library, and the frequency domain transfer function from each discrete source location to each ultra-high frequency sensor is read. The interpolation weights are determined based on the distance between the target source location and each of the neighboring discrete source locations. The smaller the distance between the discrete source locations, the greater the interpolation weight. The inverse distance weighting method is used to spatially interpolate the amplitude response and phase response of each frequency sampling point. The interpolated amplitude response and phase response are then combined to form the target transfer function from the target source location to each UHF sensor. According to the target defect type, the corresponding model is called from the set of differentiated equivalent source models, and the source model parameters corresponding to the target defect type are read to generate the target source signal; the corona discharge calls the time-varying point dipole source model, the surface discharge calls the line source array model distributed along the discharge path on the insulator surface, and the internal air gap discharge calls the dielectric-enclosed spherical closed source model. The target source signal is subjected to a fast Fourier transform, and the transform result is multiplied with the target transfer function corresponding to each UHF sensor. Then, an inverse fast Fourier transform is performed to obtain the time-domain forward modeling signal received by each UHF sensor. At the same time, the corresponding frequency domain spectrum is obtained based on the frequency domain calculation results, thereby avoiding the need to re-perform three-dimensional finite element full-wave simulation for the target source location.

[0033] Furthermore, such as Figure 4As shown, the electromagnetic parameter correction of the GIS cavity digital twin model includes: The temperature and pressure of the gas insulating medium domain are periodically acquired through the GIS online monitoring system to form operating environment parameters; Based on the pre-defined mapping relationship between temperature, pressure and gas dielectric parameters, the dielectric parameters of the gas insulating medium domain are corrected using operating environment parameters. The space transfer function library is updated when the dielectric parameters of the modified gas insulating medium domain meet the preset update conditions.

[0034] As a preferred embodiment of the above, during GIS operation, the temperature and pressure of the air chamber are read through the SCADA interface of the online monitoring system; the temperature sensor has an accuracy of ±0.5℃, the pressure sensor has an accuracy of ±0.01MPa, the sampling period is three minutes, and the temperature, pressure and timestamp are continuously recorded according to the sampling time to form a time series of operating environment parameters; The relationship between temperature, pressure and the dielectric constant of SF6 gas was calibrated using a standard GIS experimental chamber. The dielectric constant of SF6 gas was measured under multiple temperature and pressure conditions, and the measurement results were fitted with least squares to obtain temperature correction coefficient, pressure correction coefficient and temperature-pressure interaction correction coefficient. Each time new temperature and pressure data were received, the calibrated temperature-pressure correction relationship was called to calculate the dielectric constant of SF6 gas under the current condition, and the calculation result was written into the digital twin model of the GIS chamber to replace the original dielectric constant of the corresponding gas filling domain. Calculate the rate of change of the current SF6 gas dielectric constant relative to the reference dielectric constant; if the rate of change does not reach the update threshold, continue to use the current spatial transfer function library for partial discharge electromagnetic wave forward modeling; if the rate of change reaches the update threshold, recalculate the frequency domain transfer function from all discrete source locations to each ultra-high frequency sensor according to the corrected SF6 gas dielectric constant, and replace the corresponding data in the spatial transfer function library. After the transfer function is updated, a new version identifier is generated for the space transfer function library. The version identifier is then associated with and stored with temperature, pressure, update time, and the corrected SF6 gas dielectric constant, so that subsequent forward modeling signals can be traced back to the corresponding operating conditions and space transfer function library version.

[0035] Furthermore, such as Figure 4 As shown, the electromagnetic parameter correction of the GIS cavity digital twin model also includes: The cumulative discharge energy is determined based on the partial discharge signal, and the cumulative discharge energy is used to characterize the insulation aging state parameters. Based on the pre-calibrated mapping relationship between cumulative discharge energy and dielectric loss parameters, the dielectric loss parameters of the insulator are corrected. When the corrected dielectric loss parameters meet the preset update conditions, update the frequency domain transfer function in the spatial transfer function library corresponding to the region near the insulator.

[0036] As a preferred embodiment of the above, during the operation of GIS, the partial discharge signal output by the UHF sensor is continuously collected, the discharge voltage and discharge current corresponding to each discharge are deduced based on the amplitude of the UHF signal, and the energy of a single discharge is calculated in combination with the discharge duration; the energy of each discharge is accumulated in chronological order to obtain the cumulative discharge energy, and the cumulative discharge energy is used as an insulation aging state parameter characterizing the aging degree of epoxy resin insulators. Accelerated aging tests on epoxy resin insulators were conducted to calibrate the relationship between cumulative discharge energy and the change in loss tangent. The maximum change in loss tangent was set to 0.003, and the energy characteristic parameter was set to 500J. The cumulative discharge energy was read on a monthly basis, and the current loss tangent was calculated according to the calibration relationship. The calculation results were then written into the epoxy resin insulator material parameters of the GIS cavity digital twin model. After each update of the loss tangent, the current loss tangent is compared with the reference loss tangent. When the degree of change reaches the preset update condition, the three-dimensional finite element full-wave simulation is re-executed only for the discrete source positions in the region near the insulator to obtain the new frequency domain transfer function from the corresponding discrete source position to each UHF sensor. The frequency domain transfer function of other regions in the spatial transfer function library remains unchanged. After the partial update is completed, the corresponding data in the spatial transfer function library is replaced with the new frequency domain transfer function, and a version identifier containing the update time, cumulative discharge energy, loss tangent and parameter hash value is generated, so that the subsequent partial discharge electromagnetic wave forward modeling signal can correspond to the current aging state of the insulator.

[0037] Furthermore, it also includes: By changing the location of the partial discharge source, the type of partial discharge defect, the operating environment parameters and noise conditions, the forward modeling signals of partial discharge electromagnetic waves can be generated in batches. The partial discharge source location, partial discharge defect type, model parameters of the differentiated equivalent source model, operating environment parameters, version identifier of the spatial transfer function library, and time-domain waveforms and frequency-domain spectra of each sensor are associated and stored to form a traceable partial discharge electromagnetic wave forward modeling signal dataset.

[0038] As a preferred embodiment of the above, after completing the construction of the spatial transfer function library and the calibration of the model parameters, the location of the partial discharge source, the type of partial discharge defect, the temperature, the pressure and the noise level are changed respectively, and the forward modeling of the electromagnetic waves of partial discharge is performed in batches. The location of the partial discharge source covers all discrete source locations in the GIS cavity. The types of partial discharge defects include corona discharge, surface discharge and internal air gap discharge. The environmental conditions are formed by a combination of multiple temperature and multiple pressure levels. The noise is superimposed on the forward modeling signal using a mixture of Gaussian noise and impulse noise. For each group of source locations, defect types and environmental conditions, the corresponding differentiated equivalent source model is first called to generate the target source signal. Then, the target transfer function is obtained from the spatial transfer function library that matches the current environmental parameters and insulation aging state. The time-domain waveform and frequency-domain spectrum of each UHF sensor are generated, and multiple sets of forward modeling samples are formed according to different noise levels. Each forward modeling sample records the three-dimensional coordinates of the partial discharge source location, the type of partial discharge defect, the type and parameters of the differentiated equivalent source model, temperature, pressure, timestamp, spatial transfer function library version identifier, and the time-domain waveform and frequency-domain spectrum of each UHF sensor. The spatial transfer function library version identifier consists of a version number and a parameter hash value, which is used to determine the state of electromagnetic parameters used when generating the current forward modeling sample. A structured partial discharge electromagnetic wave forward modeling signal dataset is formed by combining all discrete source locations, three types of partial discharge defects, multiple temperature and pressure combinations, and multiple noise levels. The dataset can be retrieved according to source location, defect type, environmental conditions, and noise level, and can be directly used for training, verification, and testing of partial discharge localization models and defect type identification models.

[0039] Example 2; Based on the same inventive concept as the GIS partial discharge digital twin fast forward modeling method in the foregoing embodiments, the present invention also provides a GIS partial discharge digital twin fast forward modeling system, the system comprising: The differentiated source modeling module constructs differentiated equivalent source models corresponding to different partial discharge defects based on their discharge mechanisms. The transfer function construction module builds a digital twin model of the GIS cavity and pre-calculates the transfer functions from multiple discrete source locations to the sensors based on the digital twin model of the GIS cavity, thus constructing a spatial transfer function library. The twin parameter update module acquires GIS operating environment parameters and insulation aging status parameters, corrects the electromagnetic parameters of the GIS cavity digital twin model based on the operating environment parameters and insulation aging status parameters, and updates the spatial transfer function library. The forward modeling element acquisition module obtains the target source location and target defect type, searches and performs spatial interpolation on the updated spatial transfer function library based on the target source location to obtain the target transfer function, and calls the corresponding differentiated equivalent source model to generate the target source signal based on the target defect type. The forward modeling signal generation module convolves the target source signal with the target transfer function to obtain the forward modeling signal of the partial discharge electromagnetic wave received by the sensor.

[0040] The system described above in this invention can effectively implement a rapid forward modeling method for GIS partial discharge digital twins, and the technical effects it can achieve are as described in the above embodiments, and will not be repeated here.

[0041] Example 3; Based on the same inventive concept as the GIS partial discharge digital twin fast forward modeling method in the foregoing embodiments, the present invention also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, can realize the GIS partial discharge digital twin fast forward modeling method.

[0042] Although this application has been described in conjunction with specific features and embodiments, it is apparent that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and accompanying drawings are merely exemplary illustrations of the application as defined herein, and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Thus, if such modifications and modifications fall within the scope of this application and its equivalents, this application is intended to include such modifications and modifications.

Claims

1. A fast forward modeling method for GIS partial-exposure digital twins, characterized in that, The method includes: Based on the discharge mechanism of different partial discharge defects, a differentiated equivalent source model corresponding to the defect type is constructed; A digital twin model of a GIS cavity is constructed, and the transfer functions from multiple discrete source locations to sensors are pre-calculated based on the digital twin model of the GIS cavity, and a spatial transfer function library is constructed. Obtain GIS operating environment parameters and insulation aging state parameters, correct the electromagnetic parameters of the GIS cavity digital twin model based on the operating environment parameters and insulation aging state parameters, and update the spatial transfer function library; Obtain the target source location and target defect type; retrieve and spatially interpolate the updated spatial transfer function library based on the target source location to obtain the target transfer function; and call the corresponding differentiated equivalent source model to generate the target source signal based on the target defect type. The target source signal is convolved with the target transfer function to obtain the forward modeling signal of the partial discharge electromagnetic wave received by the sensor. The differentiated equivalent source model includes a time-varying point dipole source corresponding to corona discharge, a line source array distributed along the discharge path on the insulator surface corresponding to surface discharge, and a dielectric-enclosed source corresponding to internal air gap discharge. The time-varying point dipole source, the line source array, and the dielectric-enclosed source respectively have time-domain waveforms, spatial distribution characteristics, and electromagnetic wave coupling characteristics that match the corresponding discharge mechanisms. The GIS cavity digital twin model includes a metal shell cavity, conductors and their supporting structures, gas chamber partitions, insulators, gas insulating medium domains, and sensor installation positions, and configures the electromagnetic parameters corresponding to the metal shell cavity, the insulators, and the gas insulating medium domains.

2. The GIS partial-effect digital twin fast forward modeling method according to claim 1, characterized in that, The construction of the space transfer function library includes: In the GIS cavity digital twin model, multiple discrete source locations are set to cover the free space region, the region near the insulator, and the conductor edge region; By applying broadband reference excitation at each of the discrete source locations and performing electromagnetic full-wave calculations, the time-domain impulse response from each of the discrete source locations to each of the sensors is obtained. The time-domain impulse response is transformed in the frequency domain to obtain the corresponding frequency-domain transfer function, and the frequency-domain transfer function is associated with and stored with the corresponding discrete source location and the sensor to form the spatial transfer function library.

3. The GIS partial-effect digital twin fast forward modeling method according to claim 1, characterized in that, Obtaining the target transfer function and generating the forward modeling signal of the partial discharge electromagnetic wave include: Based on the target source location, multiple neighboring discrete source locations are retrieved from the spatial transfer function library using a spatial index; The target transfer function is obtained by performing inverse distance weighted interpolation on the frequency domain transfer function corresponding to the location of the nearest discrete source. The target source signal is generated according to the target defect type, and the target source signal is convolved with the target transfer function by a fast Fourier transform to obtain the forward modeling signal of the partial discharge electromagnetic wave.

4. The GIS partial discharge digital twin fast forward modeling method according to claim 1, characterized in that, The electromagnetic parameter correction of the GIS cavity digital twin model includes: The temperature and pressure of the gas insulating medium domain are periodically acquired through a GIS online monitoring system to form the operating environment parameters; Based on the pre-defined mapping relationship between temperature, pressure and gas dielectric parameters, the dielectric parameters of the gas insulating medium domain are corrected using the operating environment parameters. The spatial transfer function library is updated when the dielectric parameters of the modified gas insulating medium domain meet the preset update conditions.

5. The GIS partial-effect digital twin fast forward modeling method according to claim 4, characterized in that, The electromagnetic parameter correction of the GIS cavity digital twin model also includes: The cumulative discharge energy is determined based on the partial discharge signal, and the cumulative discharge energy is used to characterize the insulation aging state parameters. Based on the pre-calibrated mapping relationship between the cumulative discharge energy and the dielectric loss parameters, the dielectric loss parameters of the insulator are corrected. When the corrected dielectric loss parameters meet the preset update conditions, the frequency domain transfer function corresponding to the region near the insulator in the spatial transfer function library is updated.

6. The GIS partial-effect digital twin fast forward modeling method according to claim 1, characterized in that, Also includes: By changing the location of the partial discharge source, the type of partial discharge defect, the operating environment parameters, and the noise conditions, the forward modeling signals of the partial discharge electromagnetic waves are generated in batches. The partial discharge source location, the partial discharge defect type, the model parameters of the differentiated equivalent source model, the operating environment parameters, the version identifier of the spatial transfer function library, and the time-domain waveform and frequency-domain spectrum corresponding to each sensor are associated and stored to form a traceable partial discharge electromagnetic wave forward modeling signal dataset.

7. A rapid forward modeling system for GIS partial-exposure digital twins, characterized in that, The system employs the GIS partial discharge digital twin fast forward modeling method as described in claim 1, wherein the system comprises: The differentiated source modeling module constructs differentiated equivalent source models corresponding to different partial discharge defects based on their discharge mechanisms. The transfer function construction module builds a digital twin model of the GIS cavity and pre-calculates the transfer functions from multiple discrete source locations to the sensors based on the digital twin model of the GIS cavity, thus constructing a spatial transfer function library. The twin parameter update module acquires GIS operating environment parameters and insulation aging status parameters, corrects the electromagnetic parameters of the GIS cavity digital twin model based on the operating environment parameters and insulation aging status parameters, and updates the spatial transfer function library. The forward modeling element acquisition module obtains the target source location and target defect type, searches and performs spatial interpolation on the updated spatial transfer function library based on the target source location to obtain the target transfer function, and calls the corresponding differentiated equivalent source model to generate the target source signal based on the target defect type. The forward modeling signal generation module convolves the target source signal with the target transfer function to obtain the forward modeling signal of the partial discharge electromagnetic wave received by the sensor.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which includes program instructions that, when executed by a processor, can implement the GIS partial discharge digital twin rapid forward modeling method as described in any one of claims 1-6.

Images