Method and system for identifying medium voltage cable faults based on multi-physical field coupling

Abstract

This invention provides a method and system for medium-voltage cable fault identification based on multi-physics coupling, applied in the field of medium-voltage cable fault identification technology. The method specifically involves: allocating aging effect weights, operating condition weights, and synergistic effect weights based on the spatiotemporal collaborative mapping relationship between aging gradient and operating condition gradient; correcting each physical field parameter based on the material physical parameters and historical operating parameters of the medium-voltage cable, combined with the aging effect weights; and constructing a multi-physics coupling model by introducing environmental influencing factors and inter-field feedback coefficients. Based on the real-time operating condition parameters of the target cable, the response field quantity of the target cable is solved by combining the operating condition effect weights and synergistic effect weights, and compared with the actual physical field quantity to output the fault identification result. This invention effectively improves the accuracy and reliability of fault identification by quantifying the independent effects and nonlinear coupling effects of aging and operating conditions during the fault identification process.

Description

Technical Field

[0001] This invention relates to the field of medium-voltage cable fault identification technology, and in particular to a medium-voltage cable fault identification method and system based on multi-physics coupling. Background Technology

[0002] During the long-term operation of medium-voltage cables, multiple physical field responses, such as electric field, temperature field, and dielectric loss field, are generated inside the cables. The distribution and evolution of these physical fields accurately reflect the cable insulation degradation, the initiation and development of local defects. Compared to single-parameter monitoring methods, fault identification based on multi-physical field coupling can comprehensively characterize the cable's health status from multiple dimensions, extracting early, subtle fault features more comprehensively. This effectively improves the comprehensiveness and sensitivity of fault identification and has become the main method for identifying medium-voltage cable faults.

[0003] However, in actual operation, cables are simultaneously subjected to load fluctuations, short-term overloads, and other operational disturbances, as well as aging effects such as insulation performance degradation and material deterioration. These operational changes and aging processes are coupled, directly altering the distribution and evolution of the multiphysics field within the cable, thus affecting the formation and extraction of fault features. Currently, most medium-voltage cable fault identification methods based on multiphysics coupling employ independent modeling and separate analysis. They either analyze the impact of operational fluctuations on the multiphysics field or simulate field anomalies caused by insulation aging, with these two analysis methods operating independently. This independent analysis approach prevents the multiphysics coupling model from accurately reproducing the field distribution in actual cable operation, making it difficult to isolate the interference caused by the coupling of operational conditions and aging, and failing to accurately extract fault features under their combined effect. Consequently, the accuracy and reliability of fault identification are low, failing to meet the accurate fault identification requirements of medium-voltage cables in complex operating scenarios. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing multi-physics coupling fault identification methods, which can only independently analyze the effects of operating condition fluctuations or insulation aging, ignoring the coupling effect between operating condition and aging, resulting in low fault identification accuracy. This invention provides a method and system for medium-voltage cable fault identification based on multi-physics coupling. By introducing the synergistic effect weight of aging and operating condition, the coupling effect between operating condition and aging is quantified. Furthermore, by introducing the quantified coupling effect when solving the field quantities in the multi-physics coupling model, the normal response field quantity reflecting aging, operating condition, and coupling effect is output. This field quantity is then compared with the actual field quantity to achieve accurate fault identification of the cable.

[0005] The objective of this invention is achieved through the following technical solution: A method for identifying medium-voltage cable faults based on multi-physics coupling includes: A gradient spatiotemporal collaborative mapping relationship is constructed based on the aging gradient and the operating condition gradient, and the weights of aging effect, operating condition effect and collaborative effect are allocated according to the gradient spatiotemporal collaborative mapping relationship. Based on the material physical parameters and historical operating parameters of medium-voltage cables, the physical field parameters are corrected by combining the weight of aging effects. Based on the corrected physical field parameters, combined with environmental influencing factors and inter-field feedback coefficients, a corresponding multi-physics coupling model is constructed. Based on the real-time operating parameters, operating condition weights, and synergistic effect weights of the target cable, the field quantity calculation is performed through a multi-physics coupling model to obtain the response field quantity of the target cable. The response field quantity is compared with the actual physical field quantity of the target cable, and the fault identification result of the target cable is output based on the comparison result.

[0006] By assigning weights to aging, operating conditions, and synergistic effects, the impact of aging, operating conditions, and their coupling on the scenario is quantified. The aging weight is used to correct material physical parameters to reflect the cable's aging state, while the operating condition and synergistic weights are used for real-time field quantity calculations to quantify operating condition disturbances and their coupling with aging. This allows the multiphysics coupling model to output a normal response field quantity that simultaneously includes the effects of aging, operating conditions, and coupling. By comparing the acquired normal response field quantity under the current operating condition with the measured scenario, rapid cable fault identification can be achieved. This effectively overcomes the problems of neglected coupling effects and easily interfered fault characteristics, improving the accuracy of cable fault diagnosis.

[0007] Furthermore, the step of constructing a gradient spatiotemporal collaborative mapping relationship based on the aging gradient and the operating condition gradient, and allocating aging effect weights, operating condition effect weights, and collaborative effect weights according to the gradient spatiotemporal collaborative mapping relationship, includes: Spatiotemporal alignment of aging gradient data and operating condition gradient data is performed to establish a spatiotemporal co-mapping matrix between aging gradient and operating condition gradient. Based on the spatiotemporal co-mapping matrix, the influence amplitudes of aging gradient and operating condition gradient on the physical field parameters of the cable at each spatiotemporal node are extracted. The product of aging gradient and operating condition gradient is used as the coupling interaction term. The influence coefficients of aging gradient, operating condition gradient and coupling interaction term at each spatiotemporal node are obtained by multivariate regression fitting. The initial influence weights of aging gradient, operating condition gradient and coupling interaction term are calculated based on the entropy weight method. Based on historical cable operation data and fault sample data, the initial influence weights are corrected to obtain the aging effect weight, operating condition effect weight, and synergistic effect weight of the coupling interaction term.

[0008] Furthermore, the aging gradient data includes insulation degradation gradient data and long-term cumulative aging gradient data distributed spatially along the cable axis, and the operating condition gradient data includes transient gradient data and short-term fluctuation gradient data of the cable operating condition parameters distributed spatially.

[0009] Furthermore, the physical parameters of the medium-voltage cable material and historical operating parameters, combined with the weighting of aging effects, are used to correct each physical field parameter, including: The intrinsic reference values ​​of each physical field parameter are determined based on the physical parameters of medium-voltage cable materials; The cumulative aging degree index is extracted from historical operating parameters, and the corresponding effective aging factor is calculated using the aging effect weight as the spatial modulation factor. The intrinsic reference values ​​of each physical field parameter are weighted and corrected point by point based on the effective aging factor to obtain the corrected physical field parameters.

[0010] Furthermore, based on the corrected physical field parameters, combined with environmental influencing factors and inter-field feedback coefficients, a corresponding multiphysics coupling model is constructed, including: Based on the corrected physical field parameters, a two-dimensional axisymmetric geometric model of the cable is established and a finite element mesh is generated. Calculate the environmental impact coefficients of each physical field based on environmental impact factors, and use the environmental impact coefficients as correction terms to construct the environmental correction field control equations for each physical field. Based on the preset inter-field feedback coefficients, and combined with the corrected physical field parameters, a two-way mutual feedback coupling correlation between different physical fields is established. By combining the environmental modified field control equations and the two-way mutual feedback coupling correlation, a set of coupled control equations is formed. The set of coupled control equations is then discretized using finite element methods based on the two-dimensional axisymmetric geometric model of the cable and the divided finite element mesh to obtain a multiphysics coupling model.

[0011] Furthermore, based on the real-time operating condition parameters, operating condition weights, and synergistic effect weights of the target cable, the field quantity calculation is performed through a multiphysics coupling model to obtain the response field quantity of the target cable, including: The transient operating condition gradient of the target cable is calculated based on the real-time operating condition parameters of the target cable. The transient operating condition gradient is weighted according to the operating condition effect weight to obtain the operating condition excitation. The working condition incentives are weighted according to the synergy weights to obtain effective working condition incentives; The effective operating condition excitation is mapped to the corresponding physical field parameters, and the source terms and boundary conditions of the multiphysics coupling model are set according to the physical field parameters and environmental parameters obtained by mapping. Based on the set source terms and boundary conditions, the multiphysics coupling model is iteratively solved using the finite element method to obtain the response field quantity of the target cable under the current operating conditions.

[0012] Furthermore, the comparison of the response field quantity with the actual physical field quantity of the target cable, and the output of the fault identification result of the target cable based on the comparison result, includes: The actual physical field quantities at each preset point of the target cable are continuously collected at each time. The difference between the actual physical field quantities at each preset point at each time and the response field quantities output by the multi-physics coupling model at the corresponding time is calculated to obtain the spatiotemporal distribution sequence of the residual of each physical field. Based on the spatiotemporal distribution sequence of residuals, the slow-varying residual components and transient residual components of each preset point are extracted; The faults of the target cable are determined based on the slow-varying residual components and transient residual components at each preset point, and the corresponding fault types are identified.

[0013] Furthermore, the step of determining the fault of the target cable based on the slow-varying residual component and transient residual component at each preset point, and identifying the corresponding fault type, includes: Based on the preset sampling period, calculate the slow-varying residual change rate of the slow-varying residual component at each preset point, identify the sampling period where the slow-varying residual change rate at each preset point exceeds the preset aging threshold, and at the same time, calculate the instantaneous amplitude of the transient residual component at each preset point. If the continuous length of the sampling period corresponding to a preset point exceeds the preset length threshold, or the instantaneous amplitude of the transient residual component at a preset point exceeds the preset transient threshold, a fault is determined to exist. Fault features are extracted based on the slow-varying residual component and the transient residual component, and the corresponding fault type is identified through a pre-trained classifier. If the rate of change of slow residuals and the instantaneous amplitude of transient residuals at all preset points do not exceed the corresponding thresholds, then it is determined that there is no fault.

[0014] Furthermore, after determining that a fault exists, the following steps are also executed: When the continuous length of the sampling period corresponding to a preset point exceeds the preset length threshold, the fault section is located based on the slow residual change rate of the corresponding preset point and adjacent preset points. When the instantaneous amplitude of the transient residual component at a preset point exceeds a preset transient threshold, the fault location is located based on the corresponding preset point.

[0015] A medium-voltage cable fault identification system based on multi-physics coupling is used to perform any of the above-mentioned identification methods, including: The parameter processing module is used to allocate the weights of aging effect, working condition effect, and synergistic effect, and to correct each physical field parameter based on the material physical parameters and historical operating parameters of the medium-voltage cable, combined with the aging effect weights. The model building module is used to construct a corresponding multiphysics coupling model based on the corrected physical field parameters, combined with environmental influencing factors and inter-field feedback coefficients. The response field quantity calculation module is used to calculate the response field quantity of the target cable based on the real-time operating condition parameters, operating condition action weights, and synergistic action weights of the target cable through a multi-physics coupling model. The fault identification module is used to compare the response field quantity with the actual physical field quantity of the target cable, and output the fault identification result of the target cable based on the comparison result.

[0016] The beneficial effects of this invention are: By assigning weights to aging, operating conditions, and synergistic effects, the influence of aging, operating conditions, and the coupling effect of aging and operating conditions on the field quantity is quantified. The aging weight is used to correct material physical parameters to reflect the cable's aging state, while the operating condition and synergistic weights are used for real-time field quantity calculation to quantify operating condition disturbances and their coupling effect with aging. This allows the multiphysics coupling model to output a normal response field quantity that simultaneously includes the effects of aging, operating conditions, and coupling. By comparing the obtained normal response field quantity under the current operating condition with the measured scenario, rapid cable fault identification can be achieved. This effectively overcomes the problems of neglected coupling effects and easily interfered fault characteristics, improving the accuracy of cable fault diagnosis. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of a process of the present invention; Figure 2 This is a schematic diagram of the calculation process of a response field quantity according to an embodiment of the present invention; Figure 3 This is a schematic diagram of a fault identification process according to an embodiment of the present invention. Detailed Implementation

[0018] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0019] Example: A method for fault identification of medium-voltage cables based on multi-physics coupling, such as Figure 1 As shown, it includes: A gradient spatiotemporal collaborative mapping relationship is constructed based on the aging gradient and the operating condition gradient, and the weights of aging effect, operating condition effect and collaborative effect are allocated according to the gradient spatiotemporal collaborative mapping relationship. Based on the material physical parameters and historical operating parameters of medium-voltage cables, the physical field parameters are corrected by combining the weight of aging effects. Based on the corrected physical field parameters, combined with environmental influencing factors and inter-field feedback coefficients, a corresponding multi-physics coupling model is constructed. Based on the real-time operating parameters, operating condition weights, and synergistic effect weights of the target cable, the field quantity calculation is performed through a multi-physics coupling model to obtain the response field quantity of the target cable. The response field quantity is compared with the actual physical field quantity of the target cable, and the fault identification result of the target cable is output based on the comparison result.

[0020] Based on cable fault identification through multi-physics coupling, the coupling effect of working conditions and aging is further quantified, and a multi-physics coupling model that fits the actual operating state under the influence of coupling effect is constructed. The normal response field quantity that simultaneously reflects aging, working conditions and coupling effect is output by the constructed multi-physics coupling model. This is then directly compared with the measured field quantity to extract reliable fault characteristics, thereby ensuring the accuracy and reliability of medium-voltage cable fault identification.

[0021] In medium-voltage cables, the distribution and evolution of various physical fields are simultaneously influenced by both aging and operating condition gradients, with both types of gradients directly altering key parameters and distribution states of the physical fields. Furthermore, the aging and operating condition gradients exhibit coupling characteristics, further amplifying their impact on the physical field's evolution. Analyzing only one type of gradient fails to fully capture the true field characteristics under the combined influence of these two factors. Therefore, in multi-physics coupling analysis and fault identification, it is crucial to consider not only the independent effects of aging and operating condition gradients on the physical fields but also the additional impact of their coupling.

[0022] However, aging gradients and operating condition gradients differ fundamentally in their temporal variation characteristics. Aging gradients tend to exhibit long-term accumulation and slow, gradual changes, while operating condition gradients show short-term fluctuations and transient abrupt changes. Their spatiotemporal variation patterns are quite different. If they are simply superimposed linearly, the coupling characteristics will be distorted due to the misalignment of temporal features and differences in spatial distribution, failing to reflect the true interaction patterns.

[0023] Therefore, when quantifying the influence of aging gradient and operating condition gradient on the physical field, we first construct a spatiotemporal collaborative mapping relationship between aging gradient and operating condition gradient, and uniformly associate and match the spatial distribution and change characteristics of the two types of gradients under different time dimensions. Then, by allocating the corresponding weights of aging gradient, operating condition gradient and coupling term of aging and operating condition, we can quantify the influence of aging, operating condition and coupling effect of aging and operating condition on the physical field.

[0024] Specifically, the step of constructing a gradient spatiotemporal collaborative mapping relationship based on the aging gradient and the operating condition gradient, and allocating aging effect weights, operating condition effect weights, and collaborative effect weights according to the gradient spatiotemporal collaborative mapping relationship, includes: Spatiotemporal alignment of aging gradient data and operating condition gradient data is performed to establish a spatiotemporal co-mapping matrix between aging gradient and operating condition gradient. Based on the spatiotemporal co-mapping matrix, the influence amplitudes of aging gradient and operating condition gradient on the physical field parameters of the cable at each spatiotemporal node are extracted. The product of aging gradient and operating condition gradient is used as the coupling interaction term. The influence coefficients of aging gradient, operating condition gradient and coupling interaction term at each spatiotemporal node are obtained by multivariate regression fitting. The initial influence weights of aging gradient, operating condition gradient and coupling interaction term are calculated based on the entropy weight method. Based on historical cable operation data and fault sample data, the initial influence weights are corrected to obtain the aging effect weight, operating condition effect weight, and synergistic effect weight of the coupling interaction term.

[0025] The gradient in aging gradient and operating condition gradient refers to the magnitude, non-uniformity, and trend of changes in the spatial distribution or temporal variation of cable characteristic parameters. It is used to quantify the strength and speed of parameter changes with location or time. Specifically, the aging gradient characterizes the spatial and temporal gradient of parameters related to the degree of cable insulation degradation, while the operating condition gradient characterizes the spatial and temporal gradient of parameters related to cable operating conditions.

[0026] When collecting aging gradient data and operating condition gradient data, the target cable's own axial spatial distribution measured data and long and short time series historical operating data are used as the basic data. The data is supplemented by historical operating data of similar cables and common fault samples to ensure the sample size of the data.

[0027] Specifically, the aging gradient data includes insulation degradation gradient data and long-term cumulative aging gradient data distributed spatially along the cable axis, and the operating condition gradient data includes transient gradient data and short-term fluctuation gradient data of the cable operating condition parameters distributed spatially.

[0028] The aging gradient data reflects the distribution differences of insulation aging along the spatial location of the cable, as well as the long-term cumulative deterioration trend. The operating condition gradient data reflects the distribution differences of operating condition parameters along the spatial location, as well as the rapid fluctuation trend in the short term. During the data collection process, the insufficient coverage of the measured data for a single target cable can be supplemented based on standardized historical gradient data of similar cables, ensuring the integrity of the gradient data in both spatial and temporal dimensions.

[0029] In the process of collecting insulation degradation gradient data of the cable along the axial spatial distribution, insulation degradation parameters such as insulation resistance, dielectric loss factor, partial discharge, and equivalent resistivity of insulation layer are first collected from multiple preset points along the target cable axis. Then, the quantitative value of the corresponding point is obtained by processing the ratio of the insulation parameter difference between adjacent points to the point spacing.

[0030] When collecting long-term cumulative aging gradient data, the following data are collected first: long-term continuous monitoring values ​​of insulation degradation parameters, years of operation, cumulative load duration, and historical ambient temperature records of the target cable since it was put into operation. Then, the quantitative value of the corresponding time node is obtained by processing the ratio of the change in insulation degradation parameters at different time nodes to the time span.

[0031] After completing gradient data acquisition, the aging gradient data and operating condition gradient data are spatiotemporally aligned based on unified cable axial spatial nodes and timestamps. The two types of gradient data under the same spatiotemporal node are matched, and a spatiotemporal co-mapping matrix for aging and operating condition gradients is established accordingly. In the constructed spatiotemporal co-mapping matrix, each row represents a unique spatiotemporal node, and each row is an independent spatiotemporal unit formed by a fixed spatial detection point along the cable axial direction and its corresponding timestamp. This unit includes spatial location information along the cable axial direction and matches the sampling time under a unified time series, thus mapping long-term cumulative aging changes and short-term fluctuating operating condition changes to the same spatiotemporal dimension. The columns of the matrix represent the characteristic parameters of each gradient, corresponding to the axial insulation degradation characteristic value and long-term cumulative aging change under the aging gradient, and the spatial transient operating condition parameter value and short-term operating condition fluctuation change under the operating condition gradient. Each element in the resulting spatiotemporal collaborative mapping matrix is ​​the actual quantified value of each gradient feature under the corresponding spatiotemporal node, that is, the specific calculated value of the insulation degradation gradient, cumulative aging gradient, transient gradient and fluctuation gradient of the cable spatial point under the spatiotemporal conditions.

[0032] Measured results of physical field parameters such as electric field, temperature field, and stress field of the cable were collected synchronously at corresponding spatiotemporal nodes. Using the aging gradient, operating condition gradient, and the product of the aging gradient and operating condition gradient as independent variables, and the physical field parameters as dependent variables, a quantitative correlation model between gradient changes and physical field responses was constructed using multiple regression. The model was then solved using the least squares method, with the optimization objective being to minimize the sum of squared residuals between the measured data and the model-fitted data. The influence coefficients of the aging gradient, operating condition gradient, and coupling interaction term at each spatiotemporal node were calculated.

[0033] The influence coefficients obtained at each spatiotemporal node are grouped into sequences. Influence coefficients of different dimensions and magnitudes are normalized and standardized, mapped to a unified numerical range. For each coefficient sequence, its corresponding weight is calculated based on its standardized numerical distribution, and the information entropy of the sequence is further calculated using these weights. Information entropy reflects the dispersion and information contribution of the coefficient sequence; higher dispersion indicates a stronger distinguishing effect on the physical field, and a lower information entropy value. Based on the information entropy values ​​of each sequence, its independent entropy weights are calculated, yielding the initial influence weights for the aging gradient, operating condition gradient, and coupling interaction terms.

[0034] After obtaining the initial influence weights using the entropy weight method, these initial influence weights can objectively reflect the degree of influence of various factors on physical field parameters under normal operating conditions. However, due to the limited data range, they cannot reflect the gradient effect law during long-term cable operation and may deviate from the actual influence law in cable operation. Therefore, historical operating data and fault sample data of the target cable and similar cables are further introduced to iteratively correct and optimize the initial influence weights, ensuring that the final weights can accurately reflect the actual interaction between aging, operating conditions, and their coupling.

[0035] Specifically, the initial influence weights are substituted into the historical operating data of the target cable and similar cables. The theoretical influence values ​​of each factor on the physical field parameters are calculated based on these weights, i.e., a weighted sum. This theoretical influence value is then compared with the measured physical field parameter values ​​in the historical operating data. By calculating indicators such as mean square error and absolute error, the degree of deviation between the initial weight calculation results and the actual data is quantified. Simultaneously, combined with fault sample data, the proportion of influence of aging, operating conditions, and coupling interactions on physical field parameter anomalies in different fault types and different fault development stages is determined to extract the action patterns of various factors. Based on the degree of deviation and action patterns, the initial influence weights are initially adjusted, increasing the weight proportions that conform to the actual patterns and decreasing the weight proportions with larger deviations.

[0036] The adjusted weights are then substituted back into the historical operating data and fault sample data to recalculate the deviation between the theoretical impact value and the actual data. If the deviation is still greater than the preset threshold, the weight ratio is adjusted again based on the direction and magnitude of the deviation, combined with the patterns in the historical data and the impact characteristics in the fault samples. The deviation comparison and weight adjustment process is repeated until the deviation between the theoretical impact value calculated by the weights and the historical measured data and fault sample data is less than the preset threshold, thus obtaining the final aging effect weight, operating condition effect weight, and synergistic effect weight. Among them, the synergistic effect weight corresponds to the corrected weight of the coupling interaction term and is used to characterize the additional field quantity influence generated by the mutual coupling of the aging gradient and the operating condition gradient.

[0037] Considering the distribution and evolution of physical fields such as electric field, temperature field, and stress field inside medium-voltage cables, which are not only determined by the inherent physical parameters of the cable material but also continuously shift with the accumulation of insulation aging, physical field calculations based solely on material physical parameters can only yield theoretical field quantities under ideal cable conditions, failing to reflect the dynamic impact of aging factors in actual operation. Furthermore, corrections based solely on historical operating parameters will deviate from the actual cable condition due to the failure to distinguish between the different effects of aging and operating conditions on the physical fields. Therefore, this paper proposes a method to correct the corresponding physical field parameters based on the weighting of aging effects, and then incorporate the spatial differences and overall degree of aging accumulation into the material properties during the model construction stage. This enables a multi-physics coupled model to characterize the degree of aging.

[0038] Specifically, the physical parameters of the medium-voltage cable, based on its material physical parameters and historical operating parameters, are adjusted by weighting the aging effect to correct each physical field parameter, including: The intrinsic reference values ​​of each physical field parameter are determined based on the physical parameters of medium-voltage cable materials; The cumulative aging degree index is extracted from historical operating parameters, and the corresponding effective aging factor is calculated using the aging effect weight as the spatial modulation factor. The intrinsic reference values ​​of each physical field parameter are weighted and corrected point by point based on the effective aging factor to obtain the corrected physical field parameters.

[0039] Using the dielectric constant, conductivity, thermal conductivity, elastic modulus, and insulation activation energy of the medium-voltage cable itself as inherent physical constraints, the intrinsic reference values ​​of each physical field parameter are determined, such as the dielectric constant and thermal conductivity in the unaged state.

[0040] Combining historical operating parameters of the cable, including load current time-series data, ambient temperature records, service life, and insulation degradation monitoring data, a cumulative aging index is extracted from these historical operating parameters. Its expression is: ; Where A represents the cumulative aging level indicator. For the number of years of operation, To accumulate load power, The number of severe overload events. , and These represent the maximum values ​​corresponding to the years of operation, cumulative load power, and the number of severe overload events, respectively.

[0041] Then, using the aging weight as the spatial modulation factor, the corresponding effective aging factor is calculated, and its expression is: ; in, For the first Effective aging factors at preset locations A preset fusion coefficient is used to balance spatial differences and overall aging trends. For the first Aging weights for each preset point.

[0042] Then, based on the calculated effective aging factor, the intrinsic reference values ​​of each physical field parameter are weighted and corrected for the location of the dwell point. The calculation expression is as follows: ; in, For the first Physical field parameters at preset points, These are the intrinsic reference values ​​for the corresponding physical field parameters. These are preset aging-sensitive parameters corresponding to different physical field parameters, with different coefficients for different physical field parameters.

[0043] When determining the aforementioned correction relationships, the variation patterns of material parameters under different aging degrees can be verified through multiphysics analytical calculations combined with finite element numerical simulations. For example, based on the structure of a medium-voltage cable, including its conductor, insulation layer, inner and outer shields, and sheath, a two-dimensional axisymmetric finite element simulation model of the cable can be established, using the governing equations of each physical field as the field control conditions for solving the problem. For different aging degrees, the corresponding corrected material parameters are substituted, and finite element solutions are performed with the rated reference operating conditions as boundary conditions to obtain the electric field, temperature field, and stress field distributions under this aging state, thereby calibrating the value of the aging sensitivity coefficient.

[0044] After correcting the physical field parameters, the final physical field parameters not only retain the intrinsic distribution law determined by the inherent physical properties of the cable material, but also reflect the spatial differences and overall process of insulation degradation through the aging effect weight and cumulative aging degree index, so that the material properties match the actual aging state of the cable.

[0045] However, in the actual operation of medium-voltage cables, there is a natural bidirectional feedback between their internal electric field, temperature field, and mechanical stress field. Electric field loss and heat generation will change the dielectric properties of the insulation material, and temperature changes will in turn adjust the electric field distribution. The structural deformation caused by thermal stress will further disturb various physical fields. At the same time, the cable is in a complex external environment with varying temperatures, humidity, soil thermal resistance, and heat dissipation conditions, and its physical field distribution will be continuously affected by environmental factors.

[0046] If the model is built solely based on the modified physical field parameters, it can only reflect the static impact of aging on each physical field, and cannot reflect the dynamic feedback between fields and the continuous disturbances of environmental factors. Therefore, environmental influence factors and inter-field feedback coefficients are further introduced, and a multi-physics coupling model is constructed in combination with the modified physical field parameters. This allows the model to closely match the dynamic evolution of the physical fields during actual cable operation, ensuring the reliability of the established multi-physics coupling model.

[0047] Specifically, based on the corrected physical field parameters, combined with environmental influencing factors and inter-field feedback coefficients, a corresponding multiphysics coupling model is constructed, including: Based on the corrected physical field parameters, a two-dimensional axisymmetric geometric model of the cable is established and a finite element mesh is generated. Calculate the environmental impact coefficients of each physical field based on environmental impact factors, and use the environmental impact coefficients as correction terms to construct the environmental correction field control equations for each physical field. Based on the preset inter-field feedback coefficients, and combined with the corrected physical field parameters, a two-way mutual feedback coupling correlation between different physical fields is established. By combining the environmental modified field control equations and the two-way mutual feedback coupling correlation, a set of coupled control equations is formed. The set of coupled control equations is then discretized using finite element methods based on the two-dimensional axisymmetric geometric model of the cable and the divided finite element mesh to obtain a multiphysics coupling model.

[0048] Using the cable axis as the axis of symmetry, geometric regions for the conductor layer, inner semiconducting layer, insulation layer, outer semiconducting layer, metallic shielding layer, and outer sheath are constructed respectively. The geometric dimensions of each region are set according to the actual specifications of the target cable to obtain a basic geometric model. Then, the corrected physical field parameters are spatially distributed according to preset points and assigned to the corresponding regions in the geometric model through interpolation or partitioning, so that the material properties exhibit a differentiated distribution along the axial direction.

[0049] The constructed geometric model is then subjected to a multi-layer structured mesh in the radial direction, with the mesh being densified in areas such as the insulation layer. In the axial direction, the mesh seed is set according to the spacing of the preset points to ensure that each preset point corresponds to at least one node or element. The model is then freely meshed using quadrilateral or triangular elements to obtain a mesh model that can be used for finite element solution.

[0050] Historical operating data of the target cable and corresponding environmental temperature and humidity monitoring data were collected. Linear regression fitting was used, with historical operating data and environmental monitoring data at corresponding time and space nodes as the data source for fitting. A quantitative mapping relationship between environmental influencing factors and physical field parameters was established, and the proportional coefficient characterizing the influence of environmental influencing factors on physical field parameters was obtained as the corresponding environmental impact coefficient.

[0051] Then, using the environmental impact coefficient as a correction term, the environmental correction field control equations for each physical field are constructed.

[0052] Among them, the steady-state Poisson equation of the insulation domain of medium-voltage cable is used as the basic electric field control equation. The basic dielectric constant of the current target cable after aging correction is determined according to the corrected physical field parameters. The product of the dielectric constant of the cable and the environmental influence coefficient corresponding to the electric field is used as the equivalent dielectric constant to replace the dielectric constant in the basic electric field control equation, thus forming the electric field control equation.

[0053] The transient heat conduction equation of the cable insulation domain is used as the basic temperature control equation. The heat source power and thermal conductivity parameters after aging correction are determined according to the corrected physical field parameters. The corrected heat source power and thermal conductivity parameters are multiplied by the environmental influence coefficient corresponding to the temperature field to obtain the equivalent heat source and thermal conductivity parameters. These are substituted into the basic heat conduction equation to replace the original fixed heat source and thermal conductivity parameters, thus forming the temperature field control equation.

[0054] Based on the linear elasticity equilibrium equation as the fundamental stress control equation, the elastic modulus and thermal expansion coefficient after aging correction are determined according to the modified physical field parameters. The modified elastic modulus and thermal expansion coefficient are multiplied by the environmental influence coefficient corresponding to the stress field to obtain the equivalent elastic modulus and equivalent thermal expansion coefficient. The solution result of the temperature field control equation is used as the temperature load, which, together with the equivalent elastic modulus and equivalent thermal expansion coefficient, is substituted into the fundamental equilibrium equation to replace the original mechanical parameters and form the stress control equation.

[0055] Using the same fitting method as the environmental impact coefficient, various inter-field feedback coefficients between the electric field and the temperature field, and between the temperature field and the stress field, were determined. Specifically, the electrothermal feedback coefficient was obtained by fitting the dielectric loss power of the electric field as the independent variable and the temperature rise rate of the temperature field as the dependent variable. The thermoelectric feedback coefficient was obtained by fitting the temperature value of the temperature field as the independent variable and the rate of change of the equivalent dielectric constant of the electric field as the dependent variable. The thermodynamic feedback coefficient was obtained by fitting the temperature difference distribution of the temperature field as the independent variable and the thermal strain amplitude of the stress field as the dependent variable.

[0056] Multiplying the dielectric loss power field value of the electric field by the electrothermal feedback coefficient, the result is used as the value of the additional dielectric loss heat source in the temperature field, establishing a quantitative coupling correlation between the electric field and the temperature field. Multiplying the temperature distribution field value of the temperature field by the thermoelectric feedback coefficient, the result is used as the coupling correction value of the equivalent dielectric constant of the electric field, establishing a quantitative coupling correlation between the temperature field and the electric field. Multiplying the temperature difference distribution field value of the temperature field by the thermodynamic feedback coefficient, the result is used as the temperature load value of the stress field, establishing a quantitative coupling correlation between the temperature field and the stress field. Finally, combining the above three quantitative coupling correlations forms a bidirectional mutual feedback coupling correlation between different physical fields.

[0057] By combining the environmental modified field control equations and the two-way mutual feedback coupling correlation, a set of coupled control equations is formed. Using the established two-dimensional axisymmetric geometric model of the cable and the generated finite element mesh, the set of coupled control equations is discretized using the finite element method. This transforms the continuous partial differential equations describing the multiphysics behavior of the cable into a solvable set of discrete algebraic equations, forming a reusable numerical simulation model, which is the multiphysics coupled model.

[0058] In actual operation, the response field of a cable is simultaneously affected by aging accumulation, operating condition disturbances, and the coupling effect of these two factors. It is impossible to directly separate the response field from measured data in isolation. If the effects of aging and coupling are eliminated for separation, it would introduce a physical contradiction related to linear decomposition. Therefore, the coupling effect is incorporated into the model solution process in the form of excitation amplification to obtain a response field that includes the combined effects of aging, operating conditions, and coupling. This eliminates the need for separation; comparing it with measured physical quantities allows for cable fault identification.

[0059] Specifically, such as Figure 2 As shown, the response field quantities of the target cable are obtained by performing field quantity calculations through a multiphysics coupling model based on the real-time operating condition parameters, operating condition weights, and synergistic effect weights of the target cable, including: The transient operating condition gradient of the target cable is calculated based on the real-time operating condition parameters of the target cable. The transient operating condition gradient is weighted according to the operating condition effect weight to obtain the operating condition excitation. The working condition incentives are weighted according to the synergy weights to obtain effective working condition incentives; The effective operating condition excitation is mapped to the corresponding physical field parameters, and the source terms and boundary conditions of the multiphysics coupling model are set according to the physical field parameters and environmental parameters obtained by mapping. Based on the set source terms and boundary conditions, the multiphysics coupling model is iteratively solved using the finite element method to obtain the response field quantity of the target cable under the current operating conditions.

[0060] Real-time operating parameters, including load current, operating voltage, conductor temperature, and sheath temperature, are collected synchronously at various preset points on the target cable. At the same time, environmental parameters, such as ambient temperature, ambient humidity, wind speed, or soil thermal resistance, are also collected.

[0061] For each sampling time and each preset point, the spatial rate of change of current at that point is obtained by calculating the difference in load current between adjacent points and dividing it by the distance between the points. The corresponding spatial rates of change of voltage and temperature are calculated in the same way. In this embodiment, when calculating the transient operating condition gradient, the spatial gradient is preferred to reflect the non-uniform distribution of operating conditions along the cable length. If the cable cross-section is uniform and the operating parameters change little along the axial direction, the instantaneous rate of change of the current point with respect to time can also be used as a simplified alternative.

[0062] The transient operating condition gradient at each point is multiplied by the operating condition weight to obtain the operating condition excitation. If there are multiple operating condition parameters, such as current, voltage, and temperature, they can be calculated separately and then weighted and summed according to their respective preset sensitivity coefficients to obtain the excitation.

[0063] The synergistic effect weight is used to characterize the amplification effect of aging on operating condition disturbances. Based on this, the operating condition excitation is multiplied by an amplification factor, which is equal to 1 plus the synergistic effect weight, or 1 plus a preset coefficient multiplied by the synergistic effect weight. When the cable is completely unaged, the synergistic effect weight is zero, and the effective operating condition excitation is equal to the operating condition excitation. As the degree of aging increases, the synergistic effect weight increases, and the effective operating condition excitation is amplified accordingly, thus reflecting the objective law that the more severe the aging, the more significant the physical response to the same operating condition disturbance.

[0064] After obtaining the effective operating condition excitation, it is mapped to the corresponding physical field parameters according to the type of effective operating condition excitation, serving as internal source terms or boundary perturbations for model solution. For example, for the temperature field, the effective operating condition excitation is mapped to an additional Joule heat source density, which can be applied to the internal units of the cable conductor and insulation layer as a volumetric heat source to participate in heat conduction calculations. At the same time, based on the collected real-time environmental parameters, the external boundary conditions of the model are independently set. For example, a convective heat transfer boundary is set on the outer surface of the cable. The heat transfer coefficient applied to the convective heat transfer boundary can be determined by looking up tables or empirical formulas based on the ambient wind speed or laying method, while the ambient temperature is taken as the real-time collected value.

[0065] The established source terms and boundary conditions are substituted into the discretized multiphysics coupling model, and the solution is iteratively obtained using a finite element solver. Specifically, the field quantities are first initialized, with the temperature field set to ambient temperature, the electric field set to zero, and the stress field set to zero. Within each time step, the temperature field, electric field, and stress field are solved sequentially. First, the heat conduction equation is solved based on the current heat source distribution and boundary conditions to obtain the temperature distribution. Then, the temperature field result is substituted into the electric field control equation to solve the Poisson equation, obtaining the electric field intensity. Finally, the temperature field result is used as a thermal load to solve the linear elasticity equations, obtaining the stress distribution.

[0066] In each iteration, the field quantities are updated according to the pre-established bidirectional mutual feedback coupling correlation. For example, the dielectric loss power is fed back to the heat source of the temperature field, and the temperature distribution is fed back to the dielectric constant of the electric field.

[0067] Compare the changes in field quantity between two adjacent iterations, and stop iterating when the maximum relative change is less than the preset convergence accuracy.

[0068] Finally, the electric field intensity, temperature, and stress at all preset points at each time point are extracted to form a set of response field quantities. This response field quantity is output as the response field quantity of the target cable under the current operating conditions. This response field quantity has been modified by the environmental correction field control equation and the bidirectional mutual feedback coupling correlation built into the model, and can reflect the comprehensive influence of environmental factors, inter-field mutual feedback, aging state, real-time operating condition disturbances, and the coupling effect of the two. No further separation or correction is required, and it can be directly used for comparison with the measured field quantity for fault identification.

[0069] Although the response field output by the model already incorporates the combined effects of aging, operating conditions, and coupling, if only a single or short-term comparison is performed, the residuals may contain both slow drifts caused by insulation degradation and instantaneous jumps caused by load fluctuations, which cannot be distinguished. Therefore, through continuous monitoring and time-domain filtering, the slow-varying residuals and transient residuals can be separated, thereby avoiding misjudging normal operating condition fluctuations as faults, while accurately identifying the weak characteristics of early aging and ensuring the accuracy of fault identification.

[0070] Among them, such as Figure 3 As shown, the comparison of the response field quantity with the actual physical field quantity of the target cable, and the output of the fault identification result of the target cable based on the comparison result, includes: The actual physical field quantities at each preset point of the target cable are continuously collected at each time. The difference between the actual physical field quantities at each preset point at each time and the response field quantities output by the multi-physics coupling model at the corresponding time is calculated to obtain the spatiotemporal distribution sequence of the residual of each physical field. Based on the spatiotemporal distribution sequence of residuals, the slow-varying residual components and transient residual components of each preset point are extracted; The faults of the target cable are determined based on the slow-varying residual components and transient residual components at each preset point, and the corresponding fault types are identified.

[0071] Actual physical field quantities, including electric field strength, temperature, and stress, are continuously collected at multiple time points along the target cable. For each sampling time and each preset point, the difference between the actual physical field quantity and the response field quantity output by the multiphysics coupling model at the same time is calculated to obtain the residual value at that time and point. The residual values ​​of all points at all time points are organized according to spatial location and time order to form a spatiotemporal distribution sequence of residuals for each physical field. This sequence presents a two-dimensional data structure, with time on the horizontal axis and spatial location along the cable axis on the vertical axis.

[0072] For each preset point, the residual time series is low-pass filtered using methods such as moving average filtering. The filtered series is used as the slowly varying residual component to reflect long-term trends caused by insulation aging, slow degradation of material properties, etc. The original residual time series is then subtracted from the corresponding slowly varying residual component to obtain the transient residual component, which reflects short-term rapid changes caused by load fluctuations, voltage disturbances, sudden changes in ambient temperature, etc.

[0073] After component separation is completed, the corresponding fault judgment logic is further adopted to distinguish between gradual faults caused by insulation aging accumulation and sudden faults caused by load fluctuations, overloads and other operating condition disturbances based on the slow-changing residual component and the transient residual component. This avoids misjudging normal operating condition fluctuations as faults and prevents the weak characteristics of early aging from being submerged, thereby achieving accurate identification of fault types.

[0074] Specifically, the faults in the target cable are determined based on the slowly varying residual components and transient residual components at each preset point, and the corresponding fault types are identified, including: Based on the preset sampling period, calculate the slow-varying residual change rate of the slow-varying residual component at each preset point, identify the sampling period where the slow-varying residual change rate at each preset point exceeds the preset aging threshold, and at the same time, calculate the instantaneous amplitude of the transient residual component at each preset point. If the continuous length of the sampling period corresponding to a preset point exceeds the preset length threshold, or the instantaneous amplitude of the transient residual component at a preset point exceeds the preset transient threshold, a fault is determined to exist. Fault features are extracted based on the slow-varying residual component and the transient residual component, and the corresponding fault type is identified through a pre-trained classifier. If the rate of change of slow residuals and the instantaneous amplitude of transient residuals at all preset points do not exceed the corresponding thresholds, then it is determined that there is no fault.

[0075] For the slowly varying residual components, the rate of change of the slowly varying residual at each preset point is calculated based on the preset sampling period. Specifically, it is obtained by dividing the difference between the slowly varying residual values ​​of adjacent sampling periods by the time interval.

[0076] For each sampling point, the slow-varying residual value of the previous sampling period is subtracted from the current sampling period's slow-varying residual value, and then divided by the sampling period interval to obtain the slow-varying residual change rate of that point within that period. Then, it is determined whether the slow-varying residual change rate of each point exceeds a preset aging threshold. This aging threshold can be calibrated based on historical aging rate data of the target cable insulation material or set according to actual needs. The number of sampling periods in which each preset point continuously exceeds the aging threshold is recorded, i.e., the continuous length.

[0077] At the same time, for each preset point, the instantaneous amplitude of the transient residual component is directly calculated, that is, the absolute value of the transient residual at the sampling time, and it is determined whether it exceeds the preset transient threshold. The transient threshold can be set according to the fluctuation range of normal operating conditions, such as using three times the standard deviation of the mean of the transient residual under normal operating conditions as the transient threshold.

[0078] If there exists a preset point where the sampling period length for which the rate of change of the slow-varying residual continuously exceeds the aging threshold is greater than the preset length threshold, then the point is determined to have a gradual fault.

[0079] If there is a preset point where the instantaneous amplitude of the transient residual exceeds the preset transient threshold, then it is immediately determined that there is a sudden fault at that point.

[0080] A fault is determined to exist if either of the above two conditions is met.

[0081] For cases where a fault is identified, further fault features are extracted to identify the specific fault type. These fault features include the slow-varying residual change rate sequence, the instantaneous amplitude sequence of the transient residual, the energy ratio of the slow-varying residual to the transient residual, the duration of the residual peak, and the residual spatial expansion rate for the preset location and its adjacent locations. The energy ratio of the slow-varying residual to the transient residual is the ratio of the sum of squares of the slow-varying residual to the sum of squares of the transient residual over a period of time. The residual spatial expansion rate is the ratio of the time difference between adjacent locations exceeding a threshold residual to the distance between them. The extracted fault features are combined into a feature vector and input into a pre-trained classifier, such as a support vector machine or random forest, based on historical cable fault sample data to obtain the specific fault type.

[0082] If the rate of change of slow residual at all preset points does not exceed the aging threshold, or exceeds the threshold but the continuous length does not reach the preset length threshold, and the instantaneous amplitude of transient residual at all points does not exceed the transient threshold, then the cable is judged to be in a healthy and fault-free state.

[0083] To further improve the efficiency of handling cable faults, after confirming the existence of a fault, the following steps are also taken: When the continuous length of the sampling period corresponding to a preset point exceeds the preset length threshold, the fault section is located based on the slow residual change rate of the corresponding preset point and adjacent preset points. When the instantaneous amplitude of the transient residual component at a preset point exceeds a preset transient threshold, the fault location is located based on the corresponding preset point.

[0084] For gradual faults, when the continuous length of the sampling period corresponding to a preset point exceeds a preset length threshold, it is determined that a gradual fault exists in the continuous area where the preset point is located. During location, the slow-varying residual change rate values ​​of the preset point and its adjacent preset points are first extracted. Along the cable axis, with the point coordinates as the horizontal axis and the slow-varying residual change rate as the vertical axis, a change rate distribution curve is plotted to obtain the point with the largest change rate, which is taken as the fault center point.

[0085] Simultaneously, continuous intervals where the rate of change exceeds the aging threshold are identified, and the starting and ending points of these intervals are used as the boundaries of the fault segment. If the point with the largest rate of change is located in the middle of the interval, then the fault segment is expanded outward from this point to both sides until the rate of change decays below the threshold, thus forming a complete fault segment.

[0086] For sudden faults, when the instantaneous amplitude of the transient residual at a preset point exceeds a preset transient threshold, the coordinates of that preset point are directly output as the fault location. If multiple consecutive adjacent points exceed the transient threshold simultaneously, the average coordinates of these points are calculated as the fault center location, and the minimum and maximum coordinates of the region exceeding the threshold are output as the fault influence range.

[0087] For cases where both gradual and sudden faults exist simultaneously, i.e., the same cable section satisfies both the condition of continuous over-threshold slow residual and the condition of transient residual over-threshold, the locations of the gradual fault section and the sudden fault are output respectively.

[0088] Another aspect of this embodiment provides a medium-voltage cable fault identification system based on multi-physics coupling, including: The parameter processing module is used to allocate the weights of aging effect, working condition effect, and synergistic effect, and to correct each physical field parameter based on the material physical parameters and historical operating parameters of the medium-voltage cable, combined with the aging effect weights. The model building module is used to construct a corresponding multiphysics coupling model based on the corrected physical field parameters, combined with environmental influencing factors and inter-field feedback coefficients. The response field quantity calculation module is used to calculate the response field quantity of the target cable based on the real-time operating condition parameters, operating condition action weights, and synergistic action weights of the target cable through a multi-physics coupling model. The fault identification module is used to compare the response field quantity with the actual physical field quantity of the target cable, and output the fault identification result of the target cable based on the comparison result.

[0089] The parameter processing module is connected to the model building module, the model building module is connected to the response field quantity calculation module, and the response field quantity calculation module is connected to the fault identification module.

[0090] Furthermore, multiple preset points are set on the target cable, and multiple related sensing devices for collecting environmental parameters and physical quantities are set at each preset point. Each sensing device can communicate with the parameter processing module, model building module, response field quantity calculation module and fault identification module to realize data transmission.

[0091] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any way. Other variations and modifications are possible without departing from the technical solutions described in the claims.

Claims

1. A method for identifying medium-voltage cable faults based on multi-physics coupling, characterized in that, include: A gradient spatiotemporal collaborative mapping relationship is constructed based on the aging gradient and the operating condition gradient, and the weights of aging effect, operating condition effect and collaborative effect are allocated according to the gradient spatiotemporal collaborative mapping relationship. Based on the material physical parameters and historical operating parameters of medium-voltage cables, the physical field parameters are corrected by combining the weight of aging effects. Based on the corrected physical field parameters, combined with environmental influencing factors and inter-field feedback coefficients, a corresponding multi-physics coupling model is constructed. Based on the real-time operating parameters, operating condition weights, and synergistic effect weights of the target cable, the field quantity calculation is performed through a multi-physics coupling model to obtain the response field quantity of the target cable. The response field quantity is compared with the actual physical field quantity of the target cable, and the fault identification result of the target cable is output based on the comparison result.

2. The method for medium-voltage cable fault identification based on multi-physics coupling according to claim 1, characterized in that, The step of constructing a gradient spatiotemporal collaborative mapping relationship based on aging gradient and operating condition gradient, and allocating aging effect weight, operating condition effect weight, and collaborative effect weight according to the gradient spatiotemporal collaborative mapping relationship, includes: Spatiotemporal alignment of aging gradient data and operating condition gradient data is performed to establish a spatiotemporal co-mapping matrix between aging gradient and operating condition gradient. Based on the spatiotemporal co-mapping matrix, the influence amplitudes of aging gradient and operating condition gradient on the physical field parameters of the cable at each spatiotemporal node are extracted. The product of aging gradient and operating condition gradient is used as the coupling interaction term. The influence coefficients of aging gradient, operating condition gradient and coupling interaction term at each spatiotemporal node are obtained by multivariate regression fitting. The initial influence weights of aging gradient, operating condition gradient and coupling interaction term are calculated based on the entropy weight method. Based on historical cable operation data and fault sample data, the initial influence weights are corrected to obtain the aging effect weight, operating condition effect weight, and synergistic effect weight of the coupling interaction term.

3. The medium-voltage cable fault identification method based on multi-physics coupling according to claim 2, characterized in that, The aging gradient data includes insulation degradation gradient data and long-term cumulative aging gradient data distributed spatially along the cable axis. The operating condition gradient data includes transient gradient data and short-term fluctuation gradient data of the cable operating condition parameters distributed spatially.

4. The method for medium-voltage cable fault identification based on multi-physics coupling according to claim 1, characterized in that, The physical parameters of the medium-voltage cable, based on its material properties and historical operating parameters, are adjusted by weighting the aging process, including: The intrinsic reference values ​​of each physical field parameter are determined based on the physical parameters of medium-voltage cable materials; The cumulative aging degree index is extracted from historical operating parameters, and the corresponding effective aging factor is calculated using the aging effect weight as the spatial modulation factor. The intrinsic reference values ​​of each physical field parameter are weighted and corrected point by point based on the effective aging factor to obtain the corrected physical field parameters.

5. The method for medium-voltage cable fault identification based on multi-physics coupling according to claim 1, characterized in that, The corresponding multiphysics coupling model is constructed based on the corrected physical field parameters, combined with environmental influencing factors and inter-field feedback coefficients, including: Based on the corrected physical field parameters, a two-dimensional axisymmetric geometric model of the cable is established and a finite element mesh is generated. Calculate the environmental impact coefficients of each physical field based on environmental impact factors, and use the environmental impact coefficients as correction terms to construct the environmental correction field control equations for each physical field. Based on the preset inter-field feedback coefficients, and combined with the corrected physical field parameters, a two-way mutual feedback coupling correlation between different physical fields is established. By combining the environmental modified field control equations and the two-way mutual feedback coupling correlation, a set of coupled control equations is formed. The set of coupled control equations is then discretized using finite element methods based on the two-dimensional axisymmetric geometric model of the cable and the divided finite element mesh to obtain a multiphysics coupling model.

6. The method for medium-voltage cable fault identification based on multi-physics coupling according to claim 5, characterized in that, The real-time operating condition parameters, operating condition weights, and synergistic effect weights of the target cable are used to calculate the response field quantities of the target cable through a multiphysics coupling model, including: The transient operating condition gradient of the target cable is calculated based on the real-time operating condition parameters of the target cable. The transient operating condition gradient is weighted according to the operating condition effect weight to obtain the operating condition excitation. The working condition incentives are weighted according to the synergy weights to obtain effective working condition incentives; The effective operating condition excitation is mapped to the corresponding physical field parameters, and the source terms and boundary conditions of the multiphysics coupling model are set according to the physical field parameters and environmental parameters obtained by mapping. Based on the set source terms and boundary conditions, the multiphysics coupling model is iteratively solved using the finite element method to obtain the response field quantity of the target cable under the current operating conditions.

7. The method for medium-voltage cable fault identification based on multi-physics coupling according to claim 1, characterized in that, The process of comparing the response field quantity with the actual physical field quantity of the target cable and outputting the fault identification result of the target cable based on the comparison result includes: The actual physical field quantities at each preset point of the target cable are continuously collected at each time. The difference between the actual physical field quantities at each preset point at each time and the response field quantities output by the multi-physics coupling model at the corresponding time is calculated to obtain the spatiotemporal distribution sequence of the residual of each physical field. Based on the spatiotemporal distribution sequence of residuals, the slow-varying residual components and transient residual components of each preset point are extracted; The faults of the target cable are determined based on the slow-varying residual components and transient residual components at each preset point, and the corresponding fault types are identified.

8. The method for medium-voltage cable fault identification based on multi-physics coupling according to claim 7, characterized in that, The method of determining the fault of the target cable based on the slow-varying residual component and transient residual component at each preset point, and identifying the corresponding fault type, includes: Based on the preset sampling period, calculate the slow-varying residual change rate of the slow-varying residual component at each preset point, identify the sampling period where the slow-varying residual change rate at each preset point exceeds the preset aging threshold, and at the same time, calculate the instantaneous amplitude of the transient residual component at each preset point. If the continuous length of the sampling period corresponding to a preset point exceeds the preset length threshold, or the instantaneous amplitude of the transient residual component at a preset point exceeds the preset transient threshold, a fault is determined to exist. Fault features are extracted based on the slow-varying residual component and the transient residual component, and the corresponding fault type is identified through a pre-trained classifier. If the rate of change of slow residuals and the instantaneous amplitude of transient residuals at all preset points do not exceed the corresponding thresholds, then it is determined that there is no fault.

9. The method for medium-voltage cable fault identification based on multi-physics coupling according to claim 8, characterized in that, After determining that a fault exists, the following is also executed: When the continuous length of the sampling period corresponding to a preset point exceeds the preset length threshold, the fault section is located based on the slow residual change rate of the corresponding preset point and adjacent preset points. When the instantaneous amplitude of the transient residual component at a preset point exceeds a preset transient threshold, the fault location is located based on the corresponding preset point.

10. A medium-voltage cable fault identification system based on multi-physics coupling, used to execute the identification method according to any one of claims 1 to 9, characterized in that, include: The parameter processing module is used to allocate the weights of aging effect, working condition effect, and synergistic effect, and to correct each physical field parameter based on the material physical parameters and historical operating parameters of the medium-voltage cable, combined with the aging effect weights. The model building module is used to construct a corresponding multiphysics coupling model based on the corrected physical field parameters, combined with environmental influencing factors and inter-field feedback coefficients. The response field quantity calculation module is used to calculate the response field quantity of the target cable based on the real-time operating condition parameters, operating condition action weights, and synergistic action weights of the target cable through a multi-physics coupling model. The fault identification module is used to compare the response field quantity with the actual physical field quantity of the target cable, and output the fault identification result of the target cable based on the comparison result.

Images