A Calculation Method for Magnetic Field Aging Characteristics of Three-Phase Integrated High-Temperature Superconducting Cables

Abstract

This invention discloses a method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable, comprising: acquiring cable geometric parameters and superconducting tape current information; establishing an alternating magnetic field simulation model and setting typical operating conditions to calculate the alternating magnetic field distribution and magnetic loss density; collecting online monitoring data and performing filtering and noise reduction processing; aligning the preprocessed monitoring data with the simulation model output according to a time series; extracting multi-scale candidate features from the aligned time series data based on time-frequency transformation; using principal component analysis to reduce the dimensionality and filter the feature parameters to obtain feature vectors; constructing a magnetic field characteristic aging assessment model and training the feature vectors with aging level and remaining service life labels; inputting the real-time extracted features into the magnetic field characteristic aging assessment model to output the magnetic field aging level and remaining service life of the cable. This invention solves the technical problem in the prior art that the magnetic field aging assessment of superconducting cables cannot be achieved online and quantitatively.

Description

Technical Field

[0001] This invention belongs to the field of high-temperature superconducting power transmission technology, specifically a method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable. Background Technology

[0002] With the transformation of the global energy structure and the acceleration of urbanization, the load density of power grids continues to grow, especially in megacities and industrial clusters, where the capacity bottleneck of transmission lines is becoming increasingly apparent. Traditional copper or aluminum core cables exhibit significant resistive losses under high-current transmission conditions, leading to conductor temperature rise, accelerated insulation aging, and a decline in overall energy efficiency. Furthermore, transmission channel resources are limited, and the investment and construction of new lines are highly challenging. Therefore, improving the transmission capacity of existing channels has become an urgent need for the power industry.

[0003] High-temperature superconducting (HTS) cables, with their near-zero resistance below the critical temperature, can carry significantly higher currents than traditional conductors within a relatively small cross-sectional area, while exhibiting extremely low losses. This has made them a crucial technological solution for short-distance, high-capacity power transmission. In scenarios such as power plant outgoing lines, urban main grid connections, subway power supply systems, and renewable energy aggregation, three-phase integrated high-temperature superconducting cables demonstrate significant advantages due to their compact structure and high power density.

[0004] However, high-temperature superconducting cables still face performance degradation under alternating electromagnetic fields during long-term operation. This is especially true in three-phase integrated structures, where the three superconducting strips share a shield and sheath, leading to a more complex magnetic field distribution. Under rated operation, short-term overload, or even fault pulse conditions, the shield may experience magnetic flux penetration and localized magnetic flux accumulation, resulting in eddy current losses, localized heating, and material performance degradation. Existing superconducting cable operation status monitoring and diagnostic technologies mainly rely on single physical parameters such as temperature measurement, pressure detection, or partial discharge monitoring, which have the following limitations: First, temperature and pressure sensors are greatly affected by environmental coupling, resulting in relatively slow dynamic responses; second, partial discharge characteristics are mostly used for insulation fault detection, but are insufficient for capturing the magnetic loss and fatigue aging mechanisms of superconductors under alternating magnetic fields; third, traditional diagnostic methods struggle to maintain accuracy and stability under short-term fault pulses and high-frequency load fluctuations, posing challenges to online, quantitative prediction of superconducting cable life and health assessment.

[0005] Especially for three-phase integrated high-temperature superconducting cables, where the three superconducting strips share a common shielding layer and sheath structure, the complex magnetic field distribution makes magnetic flux penetration and eddy current losses in the shielding layer and sheath potential factors accelerating aging. Current research lacks mature solutions combining magnetic field simulation and online monitoring, making it difficult to accurately characterize the magnetic field aging characteristics under diverse operating conditions such as continuous operation and pulse overload. Therefore, there is an urgent need for a dedicated aging characteristic analysis method for superconducting cable systems under magnetic field effects to improve diagnostic accuracy and prediction reliability, supporting intelligent decision-making in power grid operation and maintenance. Thus, a calculation method for the magnetic field aging characteristics of three-phase integrated high-temperature superconducting cables is urgently needed. Summary of the Invention

[0006] To address the aforementioned problems in the existing technology, this invention provides a method for calculating the magnetic field aging characteristics of three-phase integrated high-temperature superconducting cables, solving the technical challenge of online and quantitative assessment of the magnetic field aging of superconducting cables in the existing technology.

[0007] The technical solution to achieve the above objectives is:

[0008] A method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable includes:

[0009] Step S1: Obtain cable geometric parameters and superconducting tape current information, establish an alternating magnetic field simulation model, set typical operating conditions, and calculate the alternating magnetic field distribution and magnetic loss density.

[0010] Step S2: Synchronously collect online monitoring data and perform filtering and noise reduction processing on the monitoring data;

[0011] Step S3: Align the preprocessed monitoring data with the simulation model output according to the time series.

[0012] Step S4: Extract multi-scale candidate features from the aligned time-series data based on time-frequency transformation;

[0013] Step S5: Principal component analysis is used to reduce the dimensionality and filter the feature parameters to obtain the feature vectors;

[0014] Step S6: Construct a magnetic field characteristic aging assessment model, and train, cross-validate and optimize the feature vectors, aging levels and remaining life labels;

[0015] Step S7: Input the real-time extracted features into the magnetic field feature aging assessment model, and output the magnetic field aging level and remaining service life of the cable.

[0016] Preferably, in step S1, the geometric structural parameters include, but are not limited to, the width, thickness, phase spacing, material type and physical properties of the superconducting tape, shielding layer and sheath;

[0017] The current information for superconducting tapes includes peak current and operating frequency range.

[0018] Preferably, in step S1, the obtained cable geometric parameters and superconducting tape current information are used as input conditions, and an alternating magnetic field simulation model is constructed using the finite element analysis method to simulate the electromagnetic field distribution and dynamic changes of a three-phase integrated high-temperature superconducting cable under given operating conditions.

[0019] Preferably, in step S1, typical operating conditions include rated current, overload current, and short-time fault pulse current;

[0020] According to typical operating conditions, the corresponding current waveform or amplitude and phase parameters are input into the alternating magnetic field simulation model. Under each operating condition, the transient electromagnetic field solution method is used to perform numerical calculations on the model to obtain the cross-section of the cable and the alternating magnetic field distribution along the length direction. During the calculation process, the nonlinear current density-electric field relationship of the superconducting tape, the eddy current effect of the shielding layer, and the interphase coupling magnetic field are comprehensively considered. The magnetic flux density distribution in the calculation domain is integrated or volume averaged through post-processing to obtain the volume magnetic loss density distribution under each operating condition.

[0021] The results are used for subsequent analysis of cable thermal characteristics and safety margin assessment.

[0022] Preferably, in step S2, the monitoring data includes current waveform, high-frequency partial discharge pulse signal, and magnetic field strength on the surface of the shielding layer; the cleaning signal is output after bandpass filtering and wavelet threshold denoising.

[0023] Preferably, in step S4, the multi-scale candidate features include higher-order harmonic amplitude, magnetic loss factor, and spectral center shift.

[0024] Preferably, in step S4, the aligned time-series data is subjected to multi-scale analysis using short-time Fourier transform or continuous wavelet transform to obtain a two-dimensional time-frequency distribution spectrum, and then multi-scale candidate features are extracted.

[0025] in,

[0026] When using short-time Fourier transform, the signal is segmented according to the preset window function length and step size, and fast Fourier transform is performed on each segment to obtain a two-dimensional time-frequency distribution spectrum.

[0027] When using continuous wavelet transform, a complex Morlet wavelet or other composite wavelet is selected as the mother wavelet function. Convolution operations with different scale factors are performed on the original signal to obtain the time-scale distribution spectrum, and then converted into a time-frequency distribution spectrum according to the correspondence between scale and frequency.

[0028] Preferably, in step S5, principal component analysis is used to reduce the dimensionality and filter the feature parameters to obtain feature vectors, including:

[0029] First, the extracted feature parameters are organized into a feature matrix, and then the matrix is ​​normalized or standardized to eliminate the differences in dimensions and scales between different features.

[0030] Subsequently, principal component analysis was used to reduce the dimensionality of the standardized feature matrix. The covariance matrix was calculated and its eigenvectors were extracted. The eigenvectors were sorted according to their eigenvalues ​​and the top few eigenvectors whose cumulative contribution rate reached a preset threshold were selected to construct the dimensionality reduction transformation matrix. The original features were then mapped to this dimensionality reduction space to obtain the set of principal component eigenvectors.

[0031] Finally, the thresholding method or the maximum variance method is applied to the dimensionality-reduced feature set to remove redundant features, and the magnetic field aging feature vector with the strongest discriminative ability is retained for subsequent model training or state evaluation.

[0032] Preferably, in step S6, a magnetic field feature aging assessment model is constructed using a convolutional neural network-long short-term memory network based on an attention mechanism, and the magnetic field aging feature vector with the strongest discriminative ability is trained, cross-validated, and hyperparameter optimized with the pre-labeled aging level and remaining life label.

[0033] The pre-labeled aging levels include mild, moderate, and severe.

[0034] Preferably, in step S7, the magnetic field aging level and remaining service life of the cable are output in real time, and an operation and maintenance warning and maintenance report are automatically triggered when the remaining service life is lower than a predetermined threshold.

[0035] The maintenance report includes the current trend of magnetic loss factor changes, aging level assessment results, and recommendations for the next steps in operation and maintenance.

[0036] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention combines focused magnetic field simulation with online monitoring, employs high-resolution time-frequency analysis and adaptive feature selection to effectively extract key indicators such as 3rd to 9th order harmonics and spectral shift, and improves the model's generalization ability and real-time online performance by using principal component analysis to reduce dimensionality. Then, it combines a convolutional neural network-long short-term memory network model based on attention mechanism to achieve online, quantitative, and high-accuracy assessment of magnetic field aging, and has built-in automatic early warning and maintenance report generation functions, significantly improving the efficiency of cable aging prediction and the ability to support maintenance decisions. Attached Figure Description

[0037] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0038] Figure 1 This is a flowchart of a calculation method for the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable according to the present invention;

[0039] Figure 2 This is a geometric schematic diagram of the alternating magnetic field simulation model in an embodiment of the present invention;

[0040] Figure 3 This is a schematic diagram of the finite element mesh generation of the alternating magnetic field simulation model in an embodiment of the present invention;

[0041] Figure 4 This is a simulation result diagram of the alternating magnetic field simulation model in an embodiment of the present invention;

[0042] Figure 5 This is a schematic diagram illustrating the comparison between the true and predicted values ​​of the training dataset in an embodiment of the present invention;

[0043] Figure 6 This is a schematic diagram illustrating another comparison result between the actual values ​​and predicted values ​​of the test dataset in an embodiment of the present invention. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] like Figure 1 As shown, a method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable includes:

[0046] Step S1: Obtain cable geometric parameters and superconducting tape current information, establish an alternating magnetic field simulation model, set typical operating conditions, and calculate the alternating magnetic field distribution and magnetic loss density.

[0047] In the embodiments, the geometric parameters include, but are not limited to, the width, thickness, phase spacing, material type and physical properties of the superconducting tape, shielding layer and sheath;

[0048] The superconducting tape current information includes the peak current (±I) and the operating frequency range (0~500 Hz).

[0049] In this embodiment, the acquired cable geometric parameters and superconducting tape current information are used as input conditions. An alternating magnetic field simulation model is constructed using the finite element analysis method to simulate the electromagnetic field distribution and dynamic changes of a three-phase integrated high-temperature superconducting cable under given operating conditions. The establishment of the finite element model may include steps such as geometric modeling, material parameter assignment, boundary condition setting, mesh generation, and solver configuration to ensure that the simulation results can accurately reflect the electromagnetic characteristics of the cable under actual operating conditions.

[0050] In the embodiments, typical operating conditions include rated current, overload current, and short-time fault pulse current;

[0051] According to typical operating conditions, the corresponding current waveform or amplitude and phase parameters are input into the alternating magnetic field simulation model. Under each operating condition, the transient electromagnetic field solution method is used to perform numerical calculations on the model to obtain the cross-section of the cable and the alternating magnetic field distribution along the length direction. During the calculation process, the nonlinear current density-electric field relationship of the superconducting tape, the eddy current effect of the shielding layer, and the interphase coupling magnetic field are comprehensively considered. The magnetic flux density distribution in the calculation domain is integrated or volume averaged through post-processing to obtain the volume magnetic loss density distribution under each operating condition.

[0052] The results are used for subsequent analysis of cable thermal characteristics and safety margin assessment.

[0053] Step S2: Synchronously collect online monitoring data and perform filtering and noise reduction processing on the monitoring data.

[0054] In this embodiment, the monitoring data includes current waveform (including harmonic components), high-frequency partial discharge pulse signal (bandwidth 100 kHz-10 MHz), and magnetic field strength on the surface of the shielding layer; the cleaning signal is output after bandpass filtering and wavelet threshold denoising.

[0055] Step S3: Align the preprocessed monitoring data with the simulation model output according to the time series to achieve one-to-one matching between the time domain and the operating conditions.

[0056] Step S4: Extract multi-scale candidate features from the aligned time-series data based on time-frequency transformation.

[0057] In the embodiments, the multi-scale candidate features include higher-order harmonic amplitudes (including the amplitudes of the 3rd to 9th harmonic components), magnetic loss factor, and spectral center shift.

[0058] In the embodiment, the aligned time series data is subjected to multi-scale analysis using short-time Fourier transform or continuous wavelet transform to obtain a two-dimensional time-frequency distribution spectrum, and then multi-scale candidate features are extracted.

[0059] in,

[0060] When using short-time Fourier transform, the signal is segmented according to the preset window function length and step size, and fast Fourier transform is performed on each segment to obtain a two-dimensional time-frequency distribution spectrum.

[0061] When using continuous wavelet transform, a complex Morlet wavelet or other composite wavelet is selected as the mother wavelet function. Convolution operations with different scale factors are performed on the original signal to obtain the time-scale distribution spectrum, and then converted into a time-frequency distribution spectrum according to the correspondence between scale and frequency.

[0062] Step S5: Principal component analysis is used to reduce the dimensionality and filter the feature parameters to obtain the feature vector.

[0063] In this embodiment, principal component analysis is used to reduce the dimensionality and filter the feature parameters to obtain feature vectors, including:

[0064] First, the extracted feature parameters are organized into a feature matrix, and then the matrix is ​​normalized or standardized to eliminate the differences in dimensions and scales between different features.

[0065] Subsequently, principal component analysis was used to reduce the dimensionality of the standardized feature matrix. The covariance matrix was calculated and its eigenvectors were extracted. The eigenvectors were sorted according to their eigenvalues ​​and the top few eigenvectors whose cumulative contribution rate reached a preset threshold were selected to construct the dimensionality reduction transformation matrix. The original features were then mapped to this dimensionality reduction space to obtain the set of principal component eigenvectors.

[0066] Finally, the thresholding method or the maximum variance method is applied to the dimensionality-reduced feature set to remove redundant features, and the magnetic field aging feature vector with the strongest discriminative ability is retained for subsequent model training or state evaluation.

[0067] Step S6: Construct a magnetic field characteristic aging assessment model, and train, cross-validate and optimize the feature vectors, aging levels and remaining life labels.

[0068] In this embodiment, a magnetic field feature aging assessment model is constructed using a convolutional neural network-long short-term memory network based on an attention mechanism. The magnetic field aging feature vector with the strongest discriminative ability is trained, cross-validated, and hyperparameter optimized with the pre-labeled aging level and remaining life label.

[0069] The pre-labeled aging levels include mild, moderate, and severe.

[0070] Step S7: Input the real-time extracted features into the magnetic field feature aging assessment model, and output the magnetic field aging level and remaining service life of the cable.

[0071] In this embodiment, the magnetic field aging level and remaining service life of the cable are output in real time, and an operation and maintenance warning and maintenance report are automatically triggered when the remaining service life is lower than a predetermined threshold.

[0072] The maintenance report includes the current trend of magnetic loss factor changes, aging level assessment results, and recommendations for the next steps in operation and maintenance.

[0073] This invention combines refined magnetic field simulation with multi-source monitoring data fusion, rigorous time-frequency feature extraction, and advanced machine learning algorithms to evaluate the aging status of three-phase integrated high-temperature superconducting cables under alternating magnetic fields online and quantitatively, providing reliable data support and early warning mechanisms for cable health management and operation and maintenance decisions.

[0074] Implementation Column 1

[0075] Step 1: Establish an alternating magnetic field simulation model. Obtain the geometric structural parameters (such as superconducting strip width, thickness, phase spacing, material properties of shielding layer and sheath) and operating current information (peak value ±I, frequency range 0~500 Hz) of the target three-phase integrated high-temperature superconducting cable, and establish an alternating magnetic field simulation model in finite element software.

[0076] Obtain the precise geometric parameters of the target cable, including: the width (e.g., 4 mm), thickness (e.g., 0.2 mm), three-phase spacing, shielding layer thickness and conductivity, and the thermo-electro-magnetic properties of the sheath material; simultaneously input the operating current information (peak ±I, frequency range 0–500 Hz). In the material property settings, the conductivity of the superconducting strip is set as a temperature-dependent function, and the relationship between the critical current density and the applied magnetic field is defined. Set the boundary conditions to magnetic insulation or periodic boundaries, determined based on the size of the computational region.

[0077] like Figure 2 As shown, a three-dimensional geometric model is constructed in finite element software (such as COMSOL Multiphysics). The three-phase superconducting strips are arranged in a 120° uniform circular pattern, with an outer shielding layer and a sheath to simulate the actual laying condition.

[0078] Simulation calculations were performed based on typical operating conditions, with the rated current set (…). ), overload current ( ) and short-time fault pulse current ( Under typical working conditions such as ), the distribution of alternating magnetic field and volumetric magnetic loss density under each working condition is obtained through finite element simulation calculation.

[0079] like Figure 3 As shown, the geometric model was meshed using finite element methods during the simulation. The mesh was locally refined at the edge of the superconducting band and in the shielding layer region to improve the calculation accuracy.

[0080] like Figure 4 As shown, the simulation results of the magnetic field distribution under rated operating conditions indicate that there is an obvious region of concentrated alternating magnetic flux on the surface of the shielding layer, and the magnetic loss is the largest between adjacent phase conductors.

[0081] Step 2 involves online monitoring data acquisition and preprocessing. At the actual operating site, current sensors, partial discharge detection devices, and magnetic field sensors are deployed to simultaneously acquire the following data: current waveform (including harmonic components), high-frequency partial discharge pulse signals (bandwidth 100 kHz to 10 MHz), and the magnetic field strength on the shielding layer surface. The acquired signals are then processed using bandpass filtering and wavelet thresholding to remove environmental interference and high-frequency noise.

[0082] Step 3: After preprocessing and cleaning the online monitoring data, the next step is to strictly align these high-quality monitoring data with the simulation results under the corresponding operating conditions using timestamps. By precisely matching the time series of both, it is ensured that the actual measurement data and the simulation data correspond completely in the time domain, thereby achieving accurate correspondence of operating conditions. This step not only ensures the synchronization between data but also lays a solid foundation for subsequent joint analysis and feature extraction. Through precise matching of the time domain and operating conditions, the interference of data deviations and differences in operating conditions on the analysis results can be effectively eliminated, improving the accuracy of subsequent model training and aging judgment.

[0083] Step 4: The aligned data then proceeds to the time-frequency analysis stage, using Short-Time Fourier Transform (STFT) or Continuous Wavelet Transform (CWT) to perform multi-scale time-frequency decomposition of the signal. These methods can effectively reveal the time-varying frequency components of the cable magnetic field signal and capture the instantaneous characteristics of the signal. Based on the time-frequency spectrum, key physical indicators, including the amplitudes of the 3rd to 9th harmonics, magnetic loss factor, and spectral center frequency offset, are extracted. These candidate features cover multiple dimensions of the cable magnetic field aging process, helping to comprehensively reflect the degradation state of the cable's internal materials and structure.

[0084] Step 5: Due to the high dimensionality and redundancy of the extracted features, Principal Component Analysis (PCA) is used to reduce the dimensionality of the multi-scale features to improve the model's discrimination efficiency and stability. PCA maps high-dimensional features to a low-dimensional space through linear transformation, retaining the main variation information and removing noise and highly correlated redundant features. Subsequently, the most discriminative subset of features is further selected using thresholding or maximum variance methods to form the final feature vector used for magnetic field aging state identification.

[0085] Step 6: In the model training phase, this invention employs a deep learning structure that integrates a convolutional neural network (CNN), an attention mechanism, and a bidirectional long short-term memory network (BiLSTM) to fully explore the spatial distribution patterns and temporal dependencies of magnetic field aging features. Specifically, the model input layer receives multi-scale feature vectors after dimensionality reduction and filtering; the convolutional layer is used to extract local time-frequency feature patterns, capturing the spatial local correlation of features through convolutional kernel sliding calculation; an attention mechanism layer is introduced after the convolutional layer, assigning different weights according to the importance of features to the output task, highlighting key features that contribute significantly to aging level prediction, thereby improving the model's discriminative and generalization abilities; subsequently, a bidirectional long short-term memory network (BiLSTM) structure is used to encode features along the forward and reverse directions of the time series, respectively, to simultaneously retain historical and future information, enhancing the model's ability to capture complex temporal patterns; then, a fully connected layer is used to achieve high-dimensional feature mapping and regression output, finally providing the aging level and remaining life prediction values ​​at the output layer.

[0086] During training, this invention uses mean squared error (MSE) as the loss function, selects the Adam algorithm as the optimizer, and sets the initial adaptive learning rate to 0.001 to achieve rapid convergence and stable optimization. To ensure the reliability and generalization of model performance, all samples are divided into training, validation, and test sets at a ratio of 70%, 15%, and 15%, respectively. The training set is used for updating model parameters, the validation set is used for hyperparameter tuning and preventing overfitting, and the test set is used for final performance evaluation. The number of training epochs is set to 200, and the batch size is 64 to balance training efficiency and model convergence stability. During training, an early stopping strategy and a learning rate decay mechanism are used to further improve the model's prediction accuracy and stability.

[0087] like Figure 5 As shown, the true values ​​in the training dataset are highly consistent with the model's predictions, indicating that the model has high accuracy in feature learning and aging level identification.

[0088] like Figure 6 As shown, the prediction results on the test dataset are in high agreement with the true values, verifying the model's ability to generalize on unknown data.

[0089] Step 7: During real-time cable monitoring, the system inputs the collected feature vectors into a pre-trained model. This model accurately assesses the current magnetic field aging level of the cable based on the input multi-dimensional data, determining whether the cable is in a healthy, mildly aging, or severely aging stage. Simultaneously, the model can combine historical operating data to predict the cable's remaining service life, providing a scientific basis for subsequent maintenance. When the model's predicted remaining service life falls below a set threshold, the system automatically triggers an operation and maintenance early warning, promptly sending alarm information to relevant maintenance personnel to ensure rapid response and handling of problems. This early warning mechanism effectively prevents cable failure risks caused by aging, ensuring the safe and stable operation of the power system.

[0090] Furthermore, the system automatically generates detailed maintenance reports based on the aging level and lifespan prediction results output by the model. These reports not only include an analysis of the cable's current condition but also provide targeted maintenance recommendations and risk assessments, helping the operations and maintenance team develop reasonable maintenance plans and resource allocation schemes to achieve intelligent operations and maintenance management. Through this process, cable condition monitoring and maintenance are highly automated and intelligent, improving the efficiency of preventative maintenance, extending cable lifespan, reducing maintenance costs and the risk of sudden failures, and ensuring the long-term safe operation of the power system.

[0091] The method proposed in this embodiment enables real-time monitoring and quantitative assessment of the magnetic field aging status of three-phase integrated high-temperature superconducting cables, without interrupting normal cable operation. This non-destructive online monitoring technology significantly improves the timeliness and accuracy of equipment status perception, allowing maintenance personnel to understand the cable's health condition and aging process. Based on the quantitative aging assessment results, the maintenance team can scientifically formulate maintenance plans and strategies, identify potential risks in advance, and avoid the impact of sudden failures on the power system. This method not only improves maintenance efficiency and response speed but also effectively extends the service life of high-temperature superconducting cables, ensuring the safe and stable transmission of power and promoting the development of intelligent operation and maintenance of power equipment.

[0092] Finally, it should be noted that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable, characterized in that, include: Step S1: Obtain cable geometric parameters and superconducting tape current information, establish an alternating magnetic field simulation model, set typical operating conditions, and calculate the alternating magnetic field distribution and magnetic loss density. Step S2: Synchronously collect online monitoring data and perform filtering and noise reduction processing on the monitoring data; Step S3: Align the preprocessed monitoring data with the simulation model output according to the time series. Step S4: Extract multi-scale candidate features from the aligned time-series data based on time-frequency transformation; Step S5: Principal component analysis is used to reduce the dimensionality and filter the feature parameters to obtain the feature vectors; Step S6: Construct a magnetic field characteristic aging assessment model, and train, cross-validate and optimize the feature vectors, aging levels and remaining life labels; Step S7: Input the real-time extracted features into the magnetic field feature aging assessment model, and output the magnetic field aging level and remaining service life of the cable. In step S2, the monitoring data includes current waveform, high-frequency partial discharge pulse signal and magnetic field strength on the surface of the shielding layer; after bandpass filtering and wavelet threshold denoising, the cleaning signal is output.

2. The method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable according to claim 1, characterized in that, In step S1, the geometric structural parameters include, but are not limited to, the width, thickness, phase spacing, material type and physical properties of the superconducting tape, shielding layer and sheath; The current information for superconducting tapes includes peak current and operating frequency range.

3. The method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable according to claim 1, characterized in that, In step S1, the obtained cable geometric parameters and superconducting tape current information are used as input conditions. An alternating magnetic field simulation model is constructed using the finite element analysis method to simulate the electromagnetic field distribution and dynamic changes of a three-phase integrated high-temperature superconducting cable under given operating conditions.

4. The method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable according to claim 3, characterized in that, In step S1, typical operating conditions include rated current, overload current and short-time fault pulse current. According to typical operating conditions, the corresponding current waveform or amplitude and phase parameters are input into the alternating magnetic field simulation model. Under each operating condition, the transient electromagnetic field solution method is used to perform numerical calculations on the model to obtain the cross-section of the cable and the alternating magnetic field distribution along the length direction. During the calculation process, the nonlinear current density-electric field relationship of the superconducting tape, the eddy current effect of the shielding layer, and the interphase coupling magnetic field are comprehensively considered. The magnetic flux density distribution in the calculation domain is integrated or volume averaged through post-processing to obtain the volume magnetic loss density distribution under each operating condition. The results are used for subsequent analysis of cable thermal characteristics and safety margin assessment.

5. The method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable according to claim 1, characterized in that, In step S4, the multi-scale candidate features include higher-order harmonic amplitude, magnetic loss factor, and spectral center shift.

6. The method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable according to claim 1, characterized in that, In step S4, the aligned time series data is subjected to multi-scale analysis using short-time Fourier transform or continuous wavelet transform to obtain a two-dimensional time-frequency distribution spectrum, and then multi-scale candidate features are extracted. in, When using short-time Fourier transform, the signal is segmented according to the preset window function length and step size, and fast Fourier transform is performed on each segment to obtain a two-dimensional time-frequency distribution spectrum. When using continuous wavelet transform, a complex Morlet wavelet or other composite wavelet is selected as the mother wavelet function. Convolution operations with different scale factors are performed on the original signal to obtain the time-scale distribution spectrum, and then converted into a time-frequency distribution spectrum according to the correspondence between scale and frequency.

7. The method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable according to claim 1, characterized in that, In step S5, principal component analysis is used to reduce the dimensionality and filter the feature parameters to obtain feature vectors, including: First, the extracted feature parameters are organized into a feature matrix, and then the matrix is ​​normalized or standardized to eliminate the differences in dimensions and scales between different features. Subsequently, principal component analysis was used to reduce the dimensionality of the standardized feature matrix. The covariance matrix was calculated and its eigenvectors were extracted. The eigenvectors were sorted according to their eigenvalues ​​and the top few eigenvectors whose cumulative contribution rate reached a preset threshold were selected to construct the dimensionality reduction transformation matrix. The original features were then mapped to this dimensionality reduction space to obtain the set of principal component eigenvectors. Finally, the thresholding method or the maximum variance method is applied to the dimensionality-reduced feature set to remove redundant features, and the magnetic field aging feature vector with the strongest discriminative ability is retained for subsequent model training or state evaluation.

8. The method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable according to claim 7, characterized in that, In step S6, a magnetic field feature aging assessment model is constructed using a convolutional neural network-long short-term memory network based on an attention mechanism. The magnetic field aging feature vector with the strongest discriminative ability is trained, cross-validated, and hyperparameter optimized with the pre-labeled aging level and remaining life label. The pre-labeled aging levels include mild, moderate, and severe.

9. The method for calculating the magnetic field aging characteristics of a three-phase integrated high-temperature superconducting cable according to claim 1, characterized in that, In step S7, the magnetic field aging level and remaining service life of the cable are output in real time, and an operation and maintenance warning and maintenance report are automatically triggered when the remaining service life is lower than a predetermined threshold. The maintenance report includes the current trend of magnetic loss factor changes, aging level assessment results, and recommendations for the next steps in operation and maintenance.

Images