Power transmission line transient scale combined discrimination pre-diagnosis and health management method and system

Abstract

The present application relates to the technical field of power transmission line fault pre-diagnosis, health management, online monitoring and relay protection auxiliary decision, in particular to a power transmission line transient scale joint discrimination pre-diagnosis and health management method and system, which realizes comprehensive identification of fault state, type, line, phase and position through multi-source sample construction, transient feature extraction, joint discrimination and reliability verification. It includes data acquisition and synchronous preprocessing; transient scale representation and feature fusion; decorrelation compression and joint discrimination model construction; physical consistency verification and result output. The present application can improve the recognition stability under complex operation mode, high resistance grounding, lightning flashover and few-class fault scenarios, and provide basis for power transmission line fault pre-diagnosis, health state evaluation and operation maintenance decision.

Description

Technical Field

[0001] This invention belongs to the technical field of transmission line fault pre-diagnosis, health management, online monitoring and relay protection auxiliary decision-making, and particularly relates to a method and system for transmission line transient scale joint discrimination pre-diagnosis and health management. Background Technology

[0002] Transmission lines, as a crucial component of the power system, undertake the task of transmitting electricity over long distances and in large volumes. Because transmission lines typically traverse mountainous areas, rivers, and regions with complex weather conditions, their operating environment is open, complex, and uncertain. Their operational status is susceptible to factors such as the natural environment, external disturbances, and equipment performance degradation, resulting in faults that are somewhat random and sudden. If the fault state, type, line, phase, and location cannot be determined promptly and accurately after a fault occurs, it can easily lead to relay protection malfunctions or failures to operate, reclosing strategy mismatches, and an expansion of the power outage area, thereby affecting the safe and stable operation of the power system.

[0003] Existing methods for fault identification in transmission lines mainly include distance measurement methods based on impedance characteristics, location methods based on transient arrival times, discrimination methods based on steady-state or transient thresholds, and intelligent identification methods based on historical samples. While these methods can achieve fault detection, classification, or location under certain conditions, they still suffer from insufficient stability and engineering adaptability under complex operating conditions such as high-resistance grounding, transformer saturation, strong noise interference, asynchronous sampling, multi-terminal power supply access, line parameter deviations, and changes in operating topology.

[0004] In recent years, intelligent identification methods based on sample data have been gradually applied to the field of transmission line fault diagnosis. These methods can extract nonlinear features from electrical quantities such as voltage and current, and automatically identify complex fault states. However, existing methods still have the following shortcomings: First, fault detection, fault classification, and fault location are often implemented using discrete models, which can easily lead to error propagation between multiple processing stages. Second, there is a lack of unified feature representation methods among transient components, steady-state amplitudes, protection action information, and environmental state information, making it difficult to fully reflect the fault evolution process. Third, when expanding samples for a few types of faults, the lack of electrical mechanism constraints may generate samples inconsistent with the physical characteristics of the line. Fourth, the model output results usually lack a credibility verification mechanism that combines impedance ranging, traveling wave timing, sequence component laws, and protection action logic, resulting in insufficient interpretability and engineering credibility of the identification results.

[0005] Therefore, it is necessary to propose an intelligent fault identification method for transmission lines that is structurally clear, has unified feature representation, and possesses strong engineering interpretability. Based on the comprehensive utilization of transient scale information and the concept of joint discrimination, this method forms a systematic fault diagnosis process consisting of data construction, feature representation, joint fault identification, and credibility verification. This improves the accuracy, stability, and engineering applicability of transmission line fault identification under complex operating conditions and provides status information for transmission line pre-diagnosis and health management. Summary of the Invention

[0006] This invention provides a method and system for joint transient scale discrimination for pre-diagnosis and health management of transmission lines. It addresses the shortcomings of existing technologies, such as the easy propagation of errors between multiple processing stages, difficulty in fully reflecting the fault evolution process, potential generation of samples inconsistent with the physical characteristics of the line, and insufficient interpretability and engineering reliability of the identification results. This invention improves the accuracy, stability, and engineering applicability of fault identification in transmission lines under complex operating conditions, and provides status information for pre-diagnosis and health management of transmission lines.

[0007] This invention provides a method for combined transient scale discrimination for pre-diagnosis and health management of transmission lines. Through multi-source sample construction, transient feature extraction, joint discrimination, and reliability verification, it achieves comprehensive identification of fault status, type, line, phase, and location. The method includes the following steps: S1. Data Acquisition and Synchronous Preprocessing: Collect electrical quantities, protection action quantities, reclosing records, environmental status quantities and line parameters of multiple terminals of the transmission line, and perform unified time-scale alignment, fault event window division, quality marking and label construction on data from different sources; S2. Transient Scale Characterization and Feature Fusion: Multi-scale pairwise filtering decomposition is performed on the voltage and current waveforms within the fault event window to extract amplitude statistical features, sequence component features, transient scale features and auxiliary state features, and form a fused feature vector. S3. Decorrelation Compression and Joint Discriminant Model Construction: Orthogonal decorrelation compression is performed on the fused feature vectors, and an electrical constraint sample supplementation strategy is used to handle class bias; the processed samples are input into the joint discriminant model, and the outputs are fault presence, fault category, fault line, fault phase, and fault location. S4. Physical Consistency Verification and Result Output: Based on the line impedance model, transient arrival time, sequence component law, and protection action logic, the output results of the joint discrimination model are verified, and the final fault identification result, confidence level, and handling suggestions are output.

[0008] According to the present invention, a method for combined transient scale discrimination for pre-diagnosis and health management of transmission lines is provided, wherein S1 specifically includes the following steps: S1.1 Acquire data collected by transmission line monitoring devices, protection devices, waveform recording devices, reclosing devices, and environmental monitoring devices to form the original observation vector:

[0009] In the formula, This is a three-phase voltage vector. For three-phase current vectors, For operating quantities such as frequency or rate of change of frequency, To protect the action state, For reclosing records, For environmental state quantities, For line length, The line impedance is the unit impedance. The superscript indicates the line topology status. This represents the vector transpose operation; S1.2. Establish a fault event window based on the protection start-up time, waveform recording start-up time, or the moment when the current surge exceeds the threshold as the event trigger time:

[0010] In the formula, For the first A sample of fault event windows, For the first The triggering time of an event, Relative to the event trigger time The time offset, The length of the background window before the failure. The lengths of the transient and steady-state windows after the fault; S1.3. Quality marking is performed on missing values, abnormal jump points, communication delay segments, and transformer saturation segments, and labels are constructed based on fault detection, fault category, fault line, fault phase, and fault location.

[0011] According to the present invention, a method for combined transient scale discrimination for pre-diagnosis and health management of transmission lines is provided, wherein S2 specifically includes the following steps: S2.1 Perform symmetrical component transformation on the three-phase voltage or three-phase current to obtain zero-sequence, positive-sequence, and negative-sequence characterization:

[0012] In the formula, , , It represents any three-phase voltage or three-phase current. For zero-order components, For positive order components, For negative order components, For phase rotation operator; S2.2 Perform multi-scale pairwise filtering decomposition on the transient waveform in the fault event window:

[0013] In the formula, For the first Low-frequency approximation coefficients of the layer For the first High-frequency detail coefficients of the layer and These are the paired low-pass and high-pass filter coefficients, respectively. This is the discrete sampling index of the approximate coefficient sequence of the previous layer. This is the index of the coefficients after filtering and downsampling in the current layer; S2.3 Calculate the energy, energy percentage, and transient arrival characteristics based on the high-frequency detail coefficients at each scale:

[0014] In the formula, For the first Scale energy, For the first Scale energy percentage To decompose the scale number, For the first The arrival time of the scale transient peak. To prevent constants with a denominator of zero; S2.4. The amplitude statistical features, sequence component features, transient scale features, protection action features, reclosing features, environmental state features, and quality markers are spliced ​​together to form a fused feature vector.

[0015] According to the present invention, a method for combined transient scale discrimination for pre-diagnosis and health management of transmission lines is provided. Specifically, S3 includes the following steps: S3.1. Standardize and orthogonally decorrelate the fused feature vector:

[0016] In the formula, To fuse feature vectors, and These are the mean vector and standard deviation vector of the training samples, respectively. This indicates element-wise division. It is an orthogonal projection matrix. This is the compressed fault representation vector. To prevent constants with zero denominators, a broadcast mechanism and vector are used here. Each element is added separately; the superscript T indicates the matrix transpose operation. It is the identity matrix. This indicates that the projection matrix satisfies the orthogonality constraint; S3.2 For fault categories with a small number of samples, candidate samples are generated among neighboring samples of the same type, and candidate samples that do not meet the phase sequence rules, impedance range and fault location range are removed. S3.3 Input the compressed fault representation vector into the joint discriminant model to obtain the multi-task output:

[0017] In the formula, For joint discrimination model, This refers to the line topology adjacency relationship. For faults, is there any output? For the probability of fault categories, This represents the probability of a faulty line. For fault phase probability, This refers to the relative location of the fault along the length of the line. The fault representation vector is obtained after feature fusion, standardization, and orthogonal decorrelation compression of the sample. S3.4. The model is trained using a joint objective of fault detection, fault category, fault line, fault phase and fault location, so that it can complete multiple outputs in one inference.

[0018] According to the present invention, a method for combined transient scale discrimination for pre-diagnosis and health management of transmission lines is provided. Specifically, S4 includes the following steps: S4.1 The relative location of the fault output by the joint discrimination model. Converted to fault distance Based on the impedance ranging value, the ranging value obtained based on the transient arrival time, the sequence component law, and the protection action logic, a physical consistency residual is constructed, and the output confidence level is corrected.

[0019] In the formula, To identify physical consistency deviations in the results, To determine the fault location through joint identification, The fault location is obtained by impedance ranging. The fault location is obtained based on the transient arrival time. , , These are non-negative weighting coefficients. The length of the faulty line. The electrical constraint residual vector is composed of the recognition results and the input state variables. for Norm; , , These are probability vectors for fault type, faulty line, and faulty phase, respectively. , , These represent the number of categories of the corresponding tasks. , , These represent the maximum confidence probabilities for the three tasks. The final credibility index is λ, where λ is the credibility decay coefficient. S4.2 When the final confidence level is not lower than the preset confidence threshold and the physical consistency deviation is not higher than the preset deviation threshold, output the automatic confirmation result; otherwise, output the verification mark and give the suspected fault section and suggested inspection objects.

[0020] This invention also provides a transmission line transient scale joint discrimination pre-diagnosis and health management system, used to execute the transmission line transient scale joint discrimination pre-diagnosis and health management method, including: The data acquisition and synchronization preprocessing module is used to collect electrical quantities, protection action quantities, reclosing records, environmental status quantities, and line parameters of multiple terminals of the transmission line; and to perform unified time-scale alignment, fault event window division, quality marking, and label construction on data from different sources. The transient scale characterization and feature fusion module is used to perform multi-scale pairwise filtering decomposition on voltage and current waveforms within the fault event window, extract amplitude statistical features, sequence component features, transient scale features and auxiliary state features, and form a fused feature vector. The cloud-related compression and joint discrimination module is used to perform orthogonal decorrelation compression on the fused feature vector and to handle class bias using an electrical constraint sample supplementation strategy. The processed samples are input into the joint discrimination model, which outputs the presence or absence of fault, fault category, fault line, fault phase, and fault location. The physical consistency verification and result output module is used to verify the output results of the joint discrimination model based on the line impedance model, transient arrival time, sequence component law and protection action logic, and output the final fault identification result, confidence level and handling suggestions.

[0021] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements the transmission line transient scale joint discrimination pre-diagnosis and health management method as described above.

[0022] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the transmission line transient scale joint discrimination pre-diagnosis and health management method as described above.

[0023] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the transmission line transient scale joint discrimination pre-diagnosis and health management method as described above.

[0024] Compared with existing technologies, this invention has the following advantages: It unifies multi-source acquired data into an event window, extracts local change information from fault waveforms through transient scale characterization, and outputs the presence, type, line, phase, and location of the fault in a single output using a joint discrimination model, reducing error propagation caused by multiple models connected in series. Simultaneously, this invention establishes physical consistency verification through impedance, transient arrival time, sequence components, and protection action logic, giving the intelligent discrimination results higher engineering reliability. Compared with identification methods that rely solely on thresholds or single models, this invention improves the identification stability under complex operating conditions, high-resistance grounding, lightning flashover, and a few types of fault scenarios, and provides a basis for transmission line fault pre-diagnosis, health status assessment, and operation and maintenance decisions. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0026] Figure 1 This is a flowchart of the method for combined transient scale discrimination, pre-diagnosis, and health management of transmission lines provided in this embodiment of the invention; Figure 2 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0028] The implementation plan is based on multi-terminal monitoring data of transmission lines and follows the process of "data access—event segmentation—feature characterization—joint discrimination—physical verification". The sampling frequency, threshold, window length and weight parameters can be adjusted according to the line voltage level, on-site sampling conditions and operating mode.

[0029] like Figure 1The diagram shown is a flowchart of a method for combined transient scale discrimination pre-diagnosis and health management of transmission lines according to this embodiment. The specific implementation steps are as follows: 1) Multi-source data access.

[0030] Collect three-phase voltage, three-phase current, frequency or rate of frequency change, protection operation status, reclosing records, environmental status quantities, and line parameters to form the original observation vector:

[0031] In the formula, For three-phase voltage vectors, For three-phase current vectors, For operating quantities such as frequency or rate of change of frequency, To protect the action state, For reclosing records, For environmental state quantities, For line length, The line impedance is the unit impedance. The superscript indicates the line topology status. This represents the vector transpose operation.

[0032] 2) Time stamp alignment and event window creation.

[0033] Data from different sources are aligned to a unified timescale, with the trigger time being the moment of protection start-up, waveform recording start-up, or current surge exceeding the limit. And create a fault event window:

[0034] In the formula, For the first A sample of fault event windows, For the first The triggering time of an event, Relative to the event trigger time The time offset, The length of the background window before the failure. This represents the length of the transient and steady-state windows after a fault.

[0035] 3) Data quality labeling and multi-task labeling construction.

[0036] Missing values, jump points, communication delays, sampling asynchrony, and transformer saturation segments are quality-labeled, and these quality labels are used as part of subsequent features. Multi-task labels are constructed based on fault records, protection reports, and manual verification results.

[0037] In the formula, For whether the fault is labeled, For fault category labels, For faulty line labels, For fault phase labeling, The superscript is used to label the relative location of the fault along the line length. This represents the vector transpose operation.

[0038] 4) Sequence component transformation and phase initial characterization.

[0039] Symmetrical component transformation is performed on the three-phase voltage or three-phase current within the event window to obtain zero-sequence, positive-sequence, and negative-sequence components, which are used to reflect the characteristics of ground faults, phase-to-phase faults, and unbalanced disturbances.

[0040] In the formula, , , It represents any three-phase voltage or three-phase current. For zero-order components, For positive order components, For negative order components, This is a phase rotation operator.

[0041] 5) Transient waveform multi-scale decomposition.

[0042] The voltage, current, or sequence component waveforms of the transient interval after the fault are selected as the input sequence, and layer-by-layer decomposition is performed using paired low-pass and high-pass filters:

[0043] In the formula, For the first Low-frequency approximation coefficients of the layer For the first High-frequency detail coefficients of the layer and These are the paired low-pass and high-pass filter coefficients, respectively. This is the discrete sampling index of the approximate coefficient sequence of the previous layer. This is the index for the coefficients of the current layer. This processing is used to extract information on fault initial abrupt changes, traveling wave impacts, and damped oscillations.

[0044] 6) Calculation of scale energy and transient arrival characteristics.

[0045] Based on the high-frequency detail coefficients at each scale, the scale energy, energy proportion, and transient peak arrival time are calculated to form transient scale characteristics:

[0046] In the formula, For the first Scale energy, For the first Scale energy percentage Where Δt is the decomposition scale number, and Δt is the sampling interval. To prevent constants with a denominator of zero.

[0047] 7) Feature fusion and decorrelation compression.

[0048] The amplitude statistical characteristics, sequence component characteristics, transient scale characteristics, protection action characteristics, reclosing characteristics, environmental state characteristics, and quality markers are spliced ​​together in a fixed order, and then standardized and orthogonally decorrelated compressed.

[0049] In the formula, To fuse feature vectors, For quality label vectors, and These are the mean vector and standard deviation vector of the training samples, respectively. This indicates element-wise division. It is an orthogonal projection matrix. The compressed fault representation vector, with superscript... This represents the vector transpose operation.

[0050] 8) Supplementing minority samples and screening for electrical constraints.

[0051] For a few types of samples, such as high-resistance grounding, lightning flashover, and transitional faults, candidate samples can be generated among neighboring samples of the same type:

[0052] In the formula, and For the representation vector of the nearest neighbor samples of the same class, These are the interpolation coefficients. After candidate samples are generated, they are screened based on phase sequence rules, impedance range, line length range, and protection action logic, eliminating samples that are inconsistent with the electrical mechanism.

[0053] 9) Joint discriminant model training and online inference.

[0054] Fault characterization vector and line topology adjacency Input joint discriminant model In a single inference operation, the output includes whether a fault exists, the fault type, the faulty circuit, the faulty phase, and the relative location of the fault.

[0055] In the formula, For faults, is there any output? For the probability of fault categories, This represents the probability of a faulty line. For fault phase probability, This refers to the relative location of the fault along the length of the line. For the first The fault representation vector is obtained by feature fusion, standardization, and orthogonal decorrelation compression of each sample.

[0056] 10) Output of physical consistency verification results.

[0057] Based on line length Convert relative position to fault distance This is combined with impedance ranging, transient arrival time, sequence component patterns, and protection action logic calculations to determine consistency deviation and final reliability.

[0058] In the formula, For impedance ranging results, This is the distance measurement result at the transient arrival time. For electrical constraint residual vectors, , , denoted by C, ρ represents the maximum probability of the three discrimination tasks, and λ is the confidence decay coefficient. When C is not lower than a preset threshold and ρ is not higher than a preset threshold, an automatic confirmation result is output; otherwise, a review mark, a suspected section, and a suggested inspection object are output.

[0059] The final output may include the presence or absence of a fault, fault type, faulty line, faulty phase, fault distance, confidence level, and handling recommendations. This embodiment, through the above steps, forms a closed loop by integrating multi-source data, transient scale representation, joint discrimination, and physical consistency verification. It can be used for intelligent identification of transmission line faults under complex operating conditions, high-resistance grounding, lightning flashover, strong noise interference, and a limited number of fault sample conditions.

[0060] The following describes a transmission line transient scale joint discrimination pre-diagnosis and health management system provided by the present invention. The transmission line transient scale joint discrimination pre-diagnosis and health management system described below can be referred to in correspondence with the transmission line transient scale joint discrimination pre-diagnosis and health management method described above.

[0061] A transmission line transient scale joint discrimination pre-diagnosis and health management system, used to execute the transmission line transient scale joint discrimination pre-diagnosis and health management method, including: The data acquisition and synchronization preprocessing module is used to collect electrical quantities, protection action quantities, reclosing records, environmental status quantities, and line parameters of multiple terminals of the transmission line; and to perform unified time-scale alignment, fault event window division, quality marking, and label construction on data from different sources. The transient scale characterization and feature fusion module is used to perform multi-scale pairwise filtering decomposition on voltage and current waveforms within the fault event window, extract amplitude statistical features, sequence component features, transient scale features and auxiliary state features, and form a fused feature vector. The cloud-related compression and joint discrimination module is used to perform orthogonal decorrelation compression on the fused feature vector and to handle class bias using an electrical constraint sample supplementation strategy. The processed samples are input into the joint discrimination model, which outputs the presence or absence of fault, fault category, fault line, fault phase, and fault location. The physical consistency verification and result output module is used to verify the output results of the joint discrimination model based on the line impedance model, transient arrival time, sequence component law and protection action logic, and output the final fault identification result, confidence level and handling suggestions.

[0062] Figure 2 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 2 As shown, the electronic device may include a processor, a communications interface, a memory, and a communication bus, wherein the processor, communications interface, and memory communicate with each other via the communication bus. The processor can call logical instructions in the memory to execute the transmission line transient scale joint discrimination pre-diagnosis and health management method.

[0063] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and sold or used as independent products, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0064] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the transmission line transient scale joint discrimination pre-diagnosis and health management method provided by the above methods.

[0065] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the transmission line transient scale joint discrimination pre-diagnosis and health management method provided by the above methods.

[0066] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0067] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0068] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for combined transient-scale discrimination for pre-diagnosis and health management of transmission lines, characterized in that, Includes the following steps: S1. Data Acquisition and Synchronous Preprocessing: Acquire electrical quantities, protection action quantities, reclosing records, environmental status quantities, and line parameters of multiple terminals of the transmission line; perform unified time-scale alignment, fault event window division, quality marking, and label construction on data from different sources; S2. Transient Scale Characterization and Feature Fusion: Multi-scale pairwise filtering decomposition is performed on the voltage and current waveforms within the fault event window to extract amplitude statistical features, sequence component features, transient scale features and auxiliary state features, and form a fused feature vector. S3. Decorrelation Compression and Joint Discriminant Model Construction: Orthogonal decorrelation compression is performed on the fused feature vectors, and an electrical constraint sample supplementation strategy is used to handle class bias; the processed samples are input into the joint discriminant model, and the outputs are fault presence, fault category, fault line, fault phase, and fault location. S4. Physical Consistency Verification and Result Output: Based on the line impedance model, transient arrival time, sequence component law, and protection action logic, the output results of the joint discrimination model are verified, and the final fault identification result, confidence level, and handling suggestions are output.

2. The method for combined transient scale discrimination, pre-diagnosis, and health management of transmission lines according to claim 1, characterized in that, The specific implementation of S1 includes the following steps: S1.1 Acquire data collected by transmission line monitoring devices, protection devices, waveform recording devices, reclosing devices, and environmental monitoring devices to form the original observation vector: In the formula, For three-phase voltage vectors, For three-phase current vectors, For operating quantities such as frequency or rate of change of frequency, To protect the action state, For reclosing records, For environmental state quantities, For line length, The line impedance is the unit impedance. The superscript indicates the line topology status. This represents the vector transpose operation; S1.

2. Establish a fault event window based on the protection start-up time, waveform recording start-up time, or the moment when the current surge exceeds the threshold as the event trigger time: In the formula, For the first A sample of fault event windows, For the first The triggering time of an event, Relative to the event trigger time The time offset, The length of the background window before the failure. The lengths of the transient and steady-state windows after the fault; S1.

3. Quality marking is performed on missing values, abnormal jump points, communication delay segments, and transformer saturation segments, and labels are constructed based on fault detection, fault category, fault line, fault phase, and fault location.

3. The method for combined transient scale discrimination, pre-diagnosis, and health management of transmission lines according to claim 1, characterized in that, The specific implementation of S2 includes the following steps: S2.1 Perform symmetrical component transformation on the three-phase voltage or three-phase current to obtain zero-sequence, positive-sequence, and negative-sequence characterization: In the formula, , , It represents any three-phase voltage or three-phase current. For zero-order components, For positive order components, For negative order components, For phase rotation operator; S2.2 Perform multi-scale pairwise filtering decomposition on the transient waveform in the fault event window: In the formula, For the first Low-frequency approximation coefficients of the layer For the first High-frequency detail coefficients of the layer and These are the paired low-pass and high-pass filter coefficients, respectively. This serves as the discrete sampling index for the approximate coefficient sequence of the previous layer. This is the index of the coefficients after filtering and downsampling in the current layer; S2.3 Calculate the energy, energy percentage, and transient arrival characteristics based on the high-frequency detail coefficients at each scale: In the formula, For the first Scale energy, For the first Scale energy percentage To decompose the scale number, For the first The arrival time of the scale transient peak. To prevent constants with a denominator of zero; S2.

4. The amplitude statistical features, sequence component features, transient scale features, protection action features, reclosing features, environmental state features, and quality markers are spliced ​​together to form a fused feature vector.

4. The method for combined transient scale discrimination, pre-diagnosis, and health management of transmission lines according to claim 1, characterized in that, The specific implementation of S3 includes the following steps: S3.

1. Standardize and orthogonally decorrelate the fused feature vector: In the formula, To fuse feature vectors, and These are the mean vector and standard deviation vector of the training samples, respectively. This indicates element-wise division. It is an orthogonal projection matrix. This is the compressed fault representation vector. To prevent constants with zero denominators, a broadcast mechanism and vector are used here. Each element is added separately; the superscript T indicates the matrix transpose operation. It is the identity matrix. This indicates that the projection matrix satisfies the orthogonality constraint; S3.2 For fault categories with a small number of samples, candidate samples are generated among neighboring samples of the same type, and candidate samples that do not meet the phase sequence rules, impedance range and fault location range are removed. S3.3 Input the compressed fault representation vector into the joint discriminant model to obtain the multi-task output: In the formula, For joint discrimination model, This refers to the line topology adjacency relationship. For faults, is there any output? For the probability of fault categories, This represents the probability of a faulty line. For fault phase probability, This refers to the relative location of the fault along the length of the line. The fault representation vector is obtained after feature fusion, standardization, and orthogonal decorrelation compression of the sample. S3.

4. The model is trained using a joint objective of fault detection, fault category, fault line, fault phase and fault location, so that it can complete multiple outputs in one inference.

5. The method for combined transient scale discrimination, pre-diagnosis, and health management of transmission lines according to claim 1, characterized in that, The specific implementation of S4 includes the following steps: S4.1 The relative location of the fault output by the joint discrimination model. Convert to fault distance Based on the impedance ranging value, the ranging value obtained based on the transient arrival time, the sequence component law, and the protection action logic, a physical consistency residual is constructed, and the output confidence level is corrected. In the formula, To identify physical consistency deviations in the results, To determine the fault location through joint identification, The fault location is obtained by impedance ranging. The fault location is obtained based on the transient arrival time. , , These are non-negative weighting coefficients. The length of the faulty line. The electrical constraint residual vector is composed of the recognition results and the input state variables. for Norm; , , These are probability vectors for fault type, faulty line, and faulty phase, respectively. , , These represent the number of categories of the corresponding tasks. , , These represent the maximum confidence probabilities for the three tasks. The final credibility index is λ, where λ is the credibility decay coefficient. S4.2 When the final confidence level is not lower than the preset confidence threshold and the physical consistency deviation is not higher than the preset deviation threshold, output the automatic confirmation result; otherwise, output the verification mark and give the suspected fault section and suggested inspection objects.

6. A transmission line transient scale joint discrimination pre-diagnosis and health management system, used to execute the transmission line transient scale joint discrimination pre-diagnosis and health management method according to any one of claims 1-5, characterized in that, include: The system includes a data acquisition and synchronization preprocessing module, a transient scale characterization and feature fusion module, a cloud-related compression and joint discrimination module, and a physical consistency verification and result output module.

7. The transmission line transient scale joint discrimination pre-diagnosis and health management system according to claim 6, characterized in that, The data acquisition and synchronization preprocessing module is used to collect electrical quantities, protection action quantities, reclosing records, environmental status quantities, and line parameters of multiple terminals of the transmission line; and to perform unified time-scale alignment, fault event window division, quality marking, and label construction on data from different sources. The transient scale characterization and feature fusion module is used to perform multi-scale pairwise filtering decomposition on voltage and current waveforms within the fault event window, extract amplitude statistical features, sequence component features, transient scale features and auxiliary state features, and form a fused feature vector. The cloud-related compression and joint discrimination module is used to perform orthogonal decorrelation compression on the fused feature vector and to handle class bias using an electrical constraint sample supplementation strategy. The processed samples are input into the joint discrimination model, which outputs the presence or absence of fault, fault category, fault line, fault phase, and fault location. The physical consistency verification and result output module is used to verify the output results of the joint discrimination model based on the line impedance model, transient arrival time, sequence component law and protection action logic, and output the final fault identification result, confidence level and handling suggestions.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method as described in any one of claims 1 to 5.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method according to any one of claims 1 to 5.

Images