A LabVIEW-based simulation method and system for locating loop-in-the-loop faults in cable sheaths.

Abstract

This invention discloses a loop-in-the-loop simulation method and system for locating cable sheath circulating current faults based on LabVIEW. The method includes: real-time acquisition of circulating current data from cable sheath monitoring points under LabVIEW control, followed by digital display and graphical representation; decomposition of the circulating current data to obtain wavelet coefficients and calculate features; inputting abnormal features into a classification model to obtain the fault type; calculating compensation coefficients considering the influence of electromagnetic fields, and calculating the fault location using the double-ended traveling wave method; switching the model to a typical operating condition simulation mode, correcting the model's predictions using Bayesian inference, and simulating the cable's full life-cycle operating status; and providing fault early warning through an LSTM model and risk assessment. This invention can effectively improve cable operation and maintenance efficiency and reliability, and is widely applicable to cable fault detection and ensuring the safe and stable operation of power systems.

Description

Technical Field

[0001] This invention belongs to the field of cable sheath circulating current fault location technology, and relates to a loop-in-the-loop simulation method and system for cable sheath circulating current fault location based on LabVIEW. Background Technology

[0002] In modern power systems, cable lines are widely used in urban power grids, industrial power supply, and other fields due to their numerous advantages in power transmission, such as small footprint, aesthetic appeal, and relatively low susceptibility to natural environmental influences. They have become a critical infrastructure for power transmission. The circulating current condition of the cable sheath directly affects the safety and stability of cable operation. Any circulating current fault can lead to serious consequences, such as localized overheating causing accelerated insulation aging, cable short circuits, or even fires, resulting in large-scale power outages and causing huge economic losses and inconvenience to social production and daily life.

[0003] Traditional cable sheath circulating current fault location technology has revealed numerous problems in practical applications. Regarding data acquisition, most existing equipment uses sensors with insufficient sensitivity, making it difficult to accurately capture subtle changes in cable sheath circulating current, and generally lacks effective anti-interference measures. In complex electromagnetic environments, such as near substations or in areas with parallel high-voltage lines, power frequency interference and electromagnetic radiation interference frequently occur, causing the acquired signals to be mixed with a large amount of noise, severely affecting the accuracy and reliability of the data, and creating hidden dangers for subsequent fault diagnosis and location.

[0004] In fault diagnosis, traditional techniques mainly rely on simple threshold judgments or empirical rules. This single diagnostic method is difficult to cope with the increasingly complex and varied types of cable faults. For some complex faults or weak signal changes in the early stages of a fault, it is often impossible to accurately identify them, which can easily lead to missed or false diagnoses and delay the best time for maintenance.

[0005] Most existing fault location algorithms are based on idealized models and do not fully consider the many complex factors in actual cable operation. Dynamic changes in the electromagnetic field around the cable, differences in cable parameters in different sections, and complex laying environments can all significantly affect the circulating current data, leading to large deviations in the location results. Moreover, during long-term operation, the electrical parameters of the cable will change over time due to factors such as changes in ambient temperature and humidity. Traditional algorithms are unable to adapt to these dynamic changes and cannot consistently guarantee location accuracy.

[0006] Furthermore, traditional technologies typically focus only on detecting and locating faults after they occur, lacking the ability to systematically simulate and analyze the entire lifecycle of cable operation, and thus failing to predict potential fault risks in advance. During operation and maintenance, staff struggle to fully grasp the operating characteristics of cables under different conditions, making it difficult to formulate scientific and reasonable preventative maintenance strategies. This leaves cable operation and maintenance in a reactive state, which is detrimental to the safe and stable operation of the power system. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a LabVIEW-based in-loop simulation method and system for locating cable sheath circulating current faults. This system can accurately collect data in real time and operate with low power consumption. It enables precise location and type identification of cable sheath circulating current faults, simulates multi-condition operation for auxiliary diagnosis, and predicts potential fault risks in advance through systematic simulation and analysis of the cable's entire lifecycle operating status. Furthermore, it supports multilingual interaction and 3D display, optimizing the user experience and effectively improving cable operation and maintenance efficiency and reliability. It is widely applicable to cable fault detection and ensuring the safe and stable operation of power systems.

[0008] The present invention adopts the following technical solution.

[0009] The first aspect of this invention proposes a loop-in-the-loop simulation method for locating cable sheath circulating current faults based on LabVIEW, comprising:

[0010] Step 1: Under the control of LabVIEW, the current sensor collects the circulating current data of the cable sheath monitoring point in real time based on micro energy, and displays it digitally and graphically on LabVIEW;

[0011] Step 2: Decompose the circulation data using wavelet transform to obtain the corresponding wavelet coefficients, and calculate the wavelet coefficient features. When the wavelet coefficient features are abnormal, input them into the pre-trained classification model to obtain the fault type.

[0012] Step 3: Calculate the compensation coefficient that takes into account the influence of electromagnetic field based on the real-time monitored electromagnetic field data, and calculate the location of the fault point using the two-end traveling wave method.

[0013] Step 4: The operating model uses real-time data of fault type, fault location and cable operating parameters, and automatically switches to the corresponding typical operating condition simulation mode through preset operating condition discrimination logic and threshold judgment conditions. It also continuously corrects the operating model's prediction of cable operating status under different simulated operating conditions through Bayesian inference, simulates the cable's full life cycle operating status, and provides risk assessment reference.

[0014] Step 5: Input the real-time operating parameters of the cable into the trained LSTM model for real-time prediction, evaluate the risk value based on the prediction data, and trigger the fault warning mechanism when the risk value exceeds the safety threshold.

[0015] Preferably, step 1, which involves digital display and graphical representation on LabVIEW, includes:

[0016] The LabVIEW data display area updates the circulating current values, temperature, and voltage of the cable sheath monitoring points in real time in digital form; the LabVIEW graphical display area uses intuitive charts and 3D dynamic models to present the changing trends of the circulating current data and the internal structure and circulation direction of the cable.

[0017] Preferably, in step 2, wavelet transform is used to decompose the circulating current data of the cable sheath, as shown in the following formula:

[0018]

[0019] in, and Decomposition layers The k-th scaling factor and wavelet coefficient; This is the nth circulation data; and Decompose respectively Time-decomposition layer Next, the k-th parent wavelet and the mother wavelet; This represents the number of decomposition levels in the wavelet transform.

[0020] Preferably, in step 3, the formula for calculating the compensation coefficient considering the influence of the electromagnetic field based on the real-time monitored electromagnetic field data is as follows:

[0021]

[0022] in This is the compensation coefficient; The deviation between the real-time monitored electromagnetic field strength and the historical average; This represents the historical average electromagnetic field strength. This indicates the spatial attenuation characteristics of electromagnetic interference. The distance between the historical fault point and the electromagnetic interference source; , This is the calibration coefficient.

[0023] Preferably, in step 3, the compensation coefficient is used to calculate the fault location using the two-terminal traveling wave method, as shown in the following formula:

[0024] =

[0025] in, Location of the fault; , The circulating current at monitoring point t is the current at time t when there is a fault at ends A and B of the cable sheath. , This represents the circulating current at monitoring point t at points A and B of the cable sheath during normal operation. , The start and end times of the selected time interval;

[0026] This is the compensation coefficient; This refers to the length of the cable sheath. This represents the propagation speed of the traveling wave.

[0027] Preferably, in step 5, the formula for assessing the risk value based on the predicted data is as follows:

[0028]

[0029]

[0030] in, Risk value; For feature weights; The basic weights are determined through the analysis of historical fault data and the combined principal component analysis-linear discriminant analysis (PCA-LDA) algorithm. The degree of matching between the feature and the current fault type; For adjustment coefficients; It is a feature mapping function; Let be the i-th input feature obtained from the predicted data; n is the total number of features.

[0031] Preferably, the feature mapping function for:

[0032]

[0033]

[0034] in, These are normal reference values ​​for the input features; This is a sensitivity adjustment factor; As a baseline sensitivity, This is the magnification factor.

[0035] Preferably, the LSTM model is a fusion structure that incorporates an attention mechanism and residual connections. The attention mechanism assigns dynamic weights to the parameter sequences of different operating stages of the cable, and the residual connections add skip connections between the hidden layers of the LSTM.

[0036] Preferably, the LSTM model adopts a dynamic phased training strategy. In the first phase, it is pre-trained using historical normal operation data. In the second phase, it is fine-tuned by introducing fault simulation data containing fault samples under different operating conditions. In the third phase, it absorbs newly collected operation and maintenance data in real time through an online learning module and updates the model parameters through a sliding window.

[0037] Preferably, the loss function of the LSTM model is a weighted mixture loss function:

[0038]

[0039] in Mean square error, For cross-entropy, The mean absolute error, , , This is a dynamically adjusted coefficient.

[0040] The second aspect of this invention proposes a LabVIEW-based loop-in-the-loop simulation system for locating cable sheath circulating current faults. The system comprises hardware and software components. The hardware component includes a current sensor and its built-in miniature energy harvesting device, along with a matching adaptive anti-interference circuit, for real-time and accurate acquisition of circulating current data from cable sheath monitoring points based on the miniature energy harvesting device. The software component includes:

[0041] The operation display module is built on the LABVIEW development platform. The human-computer interaction interface of the operation display module includes a system control button area, a real-time data display area, and a graphical result display area. It is used to control the current sensor to collect the circulating current data of the cable sheath monitoring point in real time based on micro energy, and to display it digitally and graphically.

[0042] The fault type identification module is used to decompose the circulation data using wavelet transform to obtain the corresponding wavelet coefficients and calculate the wavelet coefficient features. When the wavelet coefficient features are abnormal, they are input into a pre-trained classification model to obtain the fault type.

[0043] The fault location module is used to calculate the compensation coefficient that takes into account the influence of electromagnetic field based on real-time monitored electromagnetic field data, and to calculate the fault location by combining the two-end traveling wave method.

[0044] The running model module is used to run the model by utilizing real-time data of the cable's operating parameters, automatically switching to the corresponding typical operating condition simulation mode through preset operating condition discrimination logic and threshold judgment conditions, and continuously correcting the running model's prediction of the cable's operating status under different simulated operating conditions through Bayesian inference, simulating the cable's operating status throughout its entire life cycle to provide a risk assessment reference.

[0045] The intelligent early warning module is used to input real-time data of the cable's operating parameters into a trained LSTM model for real-time prediction. Based on the predicted data, the risk value is evaluated, and when the risk value exceeds the safety threshold, a fault early warning mechanism is triggered.

[0046] A third aspect of the present invention provides a terminal, including a processor and a storage medium; the storage medium is used to store instructions; the processor is used to perform operations according to the instructions to execute the steps of the method.

[0047] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method.

[0048] Compared with the prior art, the beneficial effects of the present invention include at least the following:

[0049] 1. This invention uses wavelet transform to decompose the circulating current data of the cable sheath, obtaining corresponding wavelet coefficients and calculating wavelet coefficient features. When the wavelet coefficient features are abnormal, they are input into a pre-trained classification model to obtain the fault type. Based on real-time monitored electromagnetic field data, compensation coefficients considering the influence of the electromagnetic field are calculated. Combined with the two-end traveling wave method, the fault location is calculated, accurately identifying the fault type and precisely determining the fault location, effectively shortening fault investigation time, improving the efficiency of power cable operation and maintenance, and reducing power outage losses. The two-end traveling wave method, by comparing the circulating current during faults and normal operation, can more accurately extract the current change characteristics caused by the fault, eliminating the interference of current fluctuations during normal operation on fault judgment, further improving the accuracy of fault location. By incorporating the spatial attenuation characteristics of electromagnetic interference into the compensation, the interference effects at different locations are dynamically corrected, reducing the location error, especially suitable for areas with strong electromagnetic interference near substations.

[0050] 2. When assessing risk values ​​based on predicted data, this invention overcomes the limitations of fixed weights by dynamically adapting weights to the fault type, effectively improving the accuracy of risk assessment and preventing key features from being masked by secondary features. Feature mapping is achieved based on sensitivity adjustment coefficients; the greater the deviation, the more the sensitivity increases exponentially, strengthening the response to serious deviations from normal conditions, avoiding over-warning of minor fluctuations, and ensuring rapid identification of major anomalies.

[0051] 3. The LSTM model of this invention is a fusion structure that introduces an attention mechanism and residual connections, and adopts a dynamic phased training strategy. It is trained using a weighted hybrid loss function. It combines the attention mechanism to focus on key features, dynamically trains to adapt to the entire life cycle, and the hybrid loss function balances prediction accuracy and robustness. It can improve the sensitivity to early weak fault signals and enhance the model's generalization ability under complex working conditions.

[0052] 4. The current sensor of this invention is based on real-time acquisition of circulating current data at cable sheath monitoring points using micro energy. The current sensor has a built-in micro energy acquisition device that can use the electromagnetic field energy around the cable to convert into power supply, achieving low-power continuous monitoring, reducing dependence on external power supply, reducing maintenance costs, and enhancing the system's long-term stable monitoring capability in complex environments.

[0053] 5. This invention implements a multilingual human-computer interaction interface and 3D dynamic graph display function based on LabVIEW, which is convenient for personnel in different regions to operate, makes cable circulation current and fault information intuitive and visible, and makes it easy for maintenance personnel to quickly grasp the cable operating status, thereby improving the convenience of operation and the efficiency of information acquisition.

[0054] 6. This invention, through switching between typical operating condition simulation modes of the running model, updating the model using Bayesian inference, and predicting and assessing risks using the LSTM model, can adapt to changes in cable operating conditions, correct the model in real time, provide accurate early warnings, reduce interference from environmental factors, and significantly improve the accuracy and reliability of system fault diagnosis and location. Attached Figure Description

[0055] Figure 1 This is a flowchart of the LOBVIEW-based cable sheath circulating current fault location simulation method of the present invention. Detailed Implementation

[0056] 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 of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, other embodiments obtained by those skilled in the art without creative effort are all within the protection scope of this invention.

[0057] Embodiment 1 of the present invention provides a loop-in-the-loop simulation method for locating cable sheath circulating current faults based on LabVIEW, such as... Figure 1 As shown, it includes:

[0058] Step 1: Under the control of LabVIEW, the current sensor collects the circulating current data of the cable sheath monitoring point in real time based on micro energy, and displays it digitally and graphically on LabVIEW;

[0059] More preferably, the digital display and graphical presentation on LabVIEW includes:

[0060] The LabVIEW data display area updates the circulating current values, temperature, and voltage of the cable sheath monitoring points in real time in digital form; the LabVIEW graphical display area uses intuitive charts and 3D dynamic models to present the changing trends of the circulating current data and the internal structure and circulation direction of the cable.

[0061] Step 2: Decompose the cable sheath circulating current data using wavelet transform to obtain the corresponding wavelet coefficients, and calculate the wavelet coefficient features. When the wavelet coefficient features are abnormal, input them into the pre-trained classification model to obtain the fault type.

[0062] More preferably, the wavelet transform is used to decompose the cable sheath circulating current data, as shown in the following formula:

[0063]

[0064] in, and Decomposition layers The k-th scaling factor and wavelet coefficients; j represents the decomposition level index of the wavelet transform, which serves as a level identifier during the decomposition process, and its value ranges from 1 to the selected decomposition level. Different The values ​​correspond to the decomposition results at different scales, as... As the signal is increased, it is broken down into a finer scale, allowing for the acquisition of characteristic information across different frequency bands. For example, when analyzing cable sheath circulating current data, a larger... The value helps capture more subtle changes in fault characteristics within the data; k is an index variable related to the scaling coefficient and wavelet coefficients at each decomposition level. The following is used to traverse and determine the corresponding scaling factors. With wavelet coefficients In different layer, The range of values ​​will be determined according to the actual data processing needs. Its function is to accurately characterize the coefficient values ​​of the wavelet function at different locations or times. These coefficients contain the characteristic information of the cable sheath circulation data at different scales and locations.

[0065] This refers to the nth cable sheath circulation current data; n is the index of the time or data point in the cable sheath circulation current data. middle, Used to represent data at different times or different sampling points. By analyzing different... Wavelet transform of the corresponding circulation data can be used to analyze the changes in data over time or location, thereby identifying the type of fault.

[0066] and Decompose respectively Time-decomposition layer Next, the k-th parent wavelet and the mother wavelet;

[0067] This represents the number of decomposition levels in the wavelet transform.

[0068] Step 3: Calculate the compensation coefficient that takes into account the influence of electromagnetic field based on the real-time monitored electromagnetic field data, and calculate the location of the fault point using the two-end traveling wave method.

[0069] More preferably, the formula for calculating the fault location using the two-terminal traveling wave method is as follows:

[0070] =

[0071] in, This refers to the location of the fault point, i.e., the position information of the fault point on the cable line relative to the starting point;

[0072] , These are the circulating currents at monitoring points t when a fault occurs at ends A and B of the cable sheath. These two current values ​​are key monitoring data when a fault occurs, and their changes contain information about the fault. In the principle of the double-ended traveling wave method, the traveling wave generated by the fault propagates to both ends of the cable, causing changes in the circulating currents at the monitoring points at both ends. By analyzing the characteristics of these two currents (such as amplitude and phase), information related to the fault can be obtained for calculating the location of the fault point.

[0073] , The circulating current at points A and B of the cable sheath at time t during normal operation is used as reference data for comparison with the circulating current during a fault. The two-end traveling wave method, by comparing the circulating current during a fault and during normal operation, can more accurately extract the current change characteristics caused by the fault, eliminate the interference of current fluctuations during normal operation on fault diagnosis, and improve the accuracy of fault location.

[0074] , The selected time interval, defined by its start and end times, is a range of time chosen during the fault occurrence process. It is determined through analysis of extensive historical fault data and simulated fault experiments. Within this specific time interval, the circulating current signal contains rich fault traveling wave information. For example, for most cable short-circuit faults, a time interval of 10μs to 50μs after the fault occurs is often selected as the time range. to The selection of this time interval is crucial for calculating the fault location. An inappropriate selection may fail to accurately capture the fault traveling wave information, leading to fault location errors.

[0075] This is a compensation coefficient used to account for the impact of changes in the electromagnetic field around the cable on the circulating current data. Since dynamic changes in the electromagnetic field around the cable can interfere with the circulating current data, thus affecting the accuracy of fault location, a compensation coefficient is introduced for correction. This coefficient is calculated in real-time using a built-in function based on the monitored electromagnetic intensity parameters. In practical applications, high-precision electromagnetic field sensors installed around the cable monitor the electromagnetic intensity parameters in real time. A mathematical model established based on electromagnetic field theory and extensive experimental data is then used to calculate the compensation coefficient, resulting in an accurate compensation coefficient value and minimizing location errors caused by environmental factors.

[0076] And preferably, , To monitor the deviation between electromagnetic field strength and historical average in real time, The average historical electromagnetic field strength. The distance between the historical fault point and the electromagnetic interference source. , For calibration coefficients, , By utilizing the spatial attenuation characteristics of electromagnetic interference It incorporates compensation to dynamically correct the interference effects at different locations, thereby reducing positioning errors, and is especially suitable for areas with strong electromagnetic interference, such as near substations.

[0077] The line length is a fixed, known parameter. When calculating the location of a fault point, it serves as a reference length in the calculation, and together with other parameters, determines the relative position of the fault point within the line.

[0078] The traveling wave velocity, or the speed at which a traveling wave propagates within a cable, is determined by the cable's material and structural characteristics. Different types of cables have different traveling wave velocities. In practical calculations, this value can be determined using empirical formulas or actual measurement data. Its influence on the proportional relationship in the fault location calculation formula is that the higher the velocity, the farther the fault point may be from the monitoring point under the same time and other conditions.

[0079] Step 4: The running model uses real-time data on fault type, fault location, and cable operating parameters. Through preset operating condition discrimination logic and threshold judgment conditions, it automatically switches to the corresponding typical operating condition simulation mode. It continuously corrects the running model's prediction of cable operating status under different simulated operating conditions through Bayesian inference, simulates the cable's full life cycle operating status, provides risk assessment reference, and regularly generates data comparison reports between simulated operating conditions and actual operating conditions.

[0080] Step 2 above obtained the fault type by analyzing the cable sheath circulating current data through wavelet transform. Step 3 combined the double-ended traveling wave method and compensation coefficient to calculate the fault location. These two steps provide key basic data for step 4. The fault type and fault location information can be used to optimize the operation model to simulate fault conditions.

[0081] In step 4, the operating model uses real-time data of cable operating parameters to switch between typical operating condition simulation modes. Information on fault type and fault location helps to more accurately simulate the cable's operating state under fault conditions. For example, if the fault type is a short-circuit fault and the fault location is known, the operating model can more accurately simulate the changes in current, voltage, and other parameters near the fault point and throughout the entire cable line after the short-circuit fault occurs. It can also simulate the recovery process of the cable's operating state under different repair measures, providing a more reliable basis for predicting the operating state after fault repair. Simultaneously, when using Bayesian inference to correct the operating model's predictions of the cable's operating state under different simulated conditions, the fault type and fault location, as important factors affecting the cable's operating state, participate in the model's prediction correction process, making the model more closely match actual operating conditions and improving simulation accuracy.

[0082] In practice, step 4, simulating the cable's entire lifecycle operation, provides data support and risk assessment reference for step 5, improving the accuracy of early warnings. These two steps are closely linked and work together to ensure the safe and stable operation of the cable system. For example, the simulation results obtained in step 4 help to calculate risk values ​​more accurately in step 5. By comparing simulated and actual operating conditions, potential trends in cable operating status can be identified. When this information is incorporated into the risk assessment function, the risk assessment becomes more closely aligned with reality. If the simulation results show a gradual increase in cable circulating current under a certain condition, and similar changes begin to appear in the actual operating conditions, then this trend will be considered when calculating the risk value in step 5, correspondingly increasing the weight of the risk assessment and making the risk value more reflective of the actual fault risk.

[0083] Step 5: Input the real-time operating parameters of the cable into the trained LSTM model for real-time prediction, evaluate the risk value based on the prediction data, and trigger the fault warning mechanism when the risk value exceeds the safety threshold.

[0084] More preferably, the LSTM model is a fusion structure incorporating an attention mechanism and residual connections. The attention mechanism dynamically assigns weights to the parameter sequences of different operating stages of the cable, and the residual connections add skip connections between the hidden layers of the LSTM. The LSTM model employs a dynamic, phased training strategy: the first phase uses historical normal operation data for pre-training; the second phase introduces fault simulation data containing fault samples under different operating conditions for fine-tuning; and the third phase absorbs newly collected operation and maintenance data in real time through an online learning module and updates the model parameters through a sliding window. The loss function of the LSTM model is a weighted hybrid loss function.

[0085]

[0086] in Mean square error, For cross-entropy, The mean absolute error, , , This is a dynamically adjusted coefficient.

[0087] By combining attention mechanisms to focus on key features, dynamically training to adapt to the entire lifecycle, and using a hybrid loss function to balance prediction accuracy and robustness, the model can improve sensitivity to early weak fault signals and enhance its generalization ability under complex conditions.

[0088] More preferably, the step of calculating the risk value based on the comparison between the predicted value and a preset threshold using a risk assessment function is as follows:

[0089]

[0090] in, The risk value is a combination of various input features. via feature mapping function The processed result, and based on the feature weights Weighted summation, when the calculated risk value When the preset safety threshold is exceeded, a fault warning mechanism will be triggered. Risk value. It is calculated by the formula. The preset threshold is a boundary value set based on cable operation safety standards, historical fault data, and actual operation and maintenance experience. It is used to determine whether the current cable operation status is within the safe range, and thus decide whether to issue a warning signal.

[0091] The fault type and fault location are important references for risk assessment. When the intelligent early warning module performs real-time prediction and calculates risk values ​​based on the LSTM model, if the prediction results show abnormal trends in parameters associated with known fault types and fault locations, the assessment weight for that risk will be increased. For example, if a short-circuit fault has occurred at a known location, and abnormal fluctuations in the circulating current near that location reappear, the intelligent early warning module will, based on historical fault type and location information, more accurately determine the likelihood of fault recurrence, improving the accuracy and timeliness of early warnings, thereby more effectively ensuring the safe and stable operation of the cable system.

[0092] And preferably, , The basic weights are determined through the analysis of historical fault data and the combined Principal Component Analysis-Linear Discriminant Analysis (PCA-LDA) algorithm. The degree of matching between the feature and the current fault type, To adjust the coefficients, the weights are dynamically adapted according to the fault type, overcoming the limitations of fixed weights, thereby improving the accuracy of risk assessment and preventing key features from being masked by secondary features.

[0093] The prediction result obtained from the LSTM model based on the real-time data of cable operation parameters is transformed into the input feature, which is obtained from the i-th input feature. Participating in risk assessment calculations. For example, predicting parameters such as the circulating current amplitude and temperature of the cable. These predicted values ​​will be processed into corresponding input features based on the cable's operating characteristics and fault mechanisms. These may be used directly as input features, or they may be calculated and transformed before being substituted into the feature mapping function. Process it;

[0094] n is the total number of features;

[0095] This is a feature mapping function, designed based on cable operating characteristics and fault mechanisms, to map input features. (Including predicted values ​​and their transformed characteristics) are transformed into indicators more relevant to failure risk, thus reflecting the impact of predicted values ​​on risk assessment. Taking the cable circulating current amplitude as an example, the details are as follows:

[0096] Mapping function Based on operational characteristics, the design first considers the relationship between fault risk and circulating current amplitude: generally, the greater the deviation of the cable circulating current amplitude from the normal range, the higher the fault risk. During normal operation, the cable circulating current amplitude is within a certain range. When the circulating current amplitude exceeds this range, the fault risk increases rapidly. A modified form of the Gaussian function is chosen, utilizing its good symmetry and attenuation characteristics on both sides of the center value to simulate the trend of fault risk changing with circulating current amplitude. The specific formula is as follows:

[0097]

[0098] in, Indicates the input characteristics of cable circulating current amplitude;

[0099] and These parameters are determined based on the cable's operating characteristics and historical fault data. Specifically, the parameters... The normal reference value representing the cable circulating current amplitude can be obtained through statistical analysis of a large amount of historical normal operating data, such as calculating the mean or median of historical normal circulating current amplitudes. value. Used to adjust the sensitivity of the function and control the rate at which the risk of failure increases when the circulating current amplitude deviates from the normal reference value. The larger the value, the steeper the function curve, meaning that even a slight deviation in the circulation amplitude will increase the risk of failure more rapidly. The smaller the value, the flatter the curve. By fitting historical fault data and combining different The accuracy of the fault risk assessment is determined by the value of the function.

[0100] The function characteristics and fault risk assessment are matched as follows:

[0101] when That is, when the cable circulating current amplitude is within the normal reference value, This indicates that the risk of failure is low at this time.

[0102] along with Deviation Whether it increases or decreases, The values ​​will all increase. The value will decrease. An increase in the value of indicates an increased risk of failure.

[0103] Furthermore, due to the characteristics of the exponential function, when the circulation amplitude deviates significantly from the normal range, It will quickly approach 1, which well reflects the situation of high failure risk under high circulation amplitude;

[0104] in, , The baseline sensitivity ranges from 0.8 to 1.2. This is the amplification factor, ranging from 1.5 to 2.0. The larger the deviation, the higher the sensitivity. It increases exponentially, strengthening the response to severe deviations from normal conditions, thereby avoiding over-warning of minor fluctuations while ensuring rapid identification of major anomalies.

[0105] Embodiment 2 of this invention provides a LabVIEW-based in-the-loop simulation system for locating cable sheath circulating current faults, used to implement the method described in Embodiment 1. The system includes hardware and software components. The hardware component includes a cable sheath circulating current fault information acquisition device for accurately and in real-time acquiring cable sheath circulating current data. This device also incorporates a miniature energy acquisition device to utilize the weak electromagnetic field energy around the cable for power generation, achieving low-power continuous operation. The software component includes:

[0106] The operation display module has an intuitive and convenient human-computer interaction interface, which enables comprehensive control of the entire simulation system and clear display of various information. Users can easily issue commands and view real-time status through this interface.

[0107] The fault type identification module based on wavelet analysis applies Daubechies wavelet transform to the acquired cable sheath circulating current data. The decomposition formula is as follows:

[0108]

[0109] In the formula, The scaling factor. These are wavelet coefficients. For scaling function, For wavelet functions, For the selected decomposition level, the fault type is accurately determined by detailed analysis of the wavelet coefficients;

[0110] The fault location algorithm module constructs a fault location algorithm based on the improved double-ended traveling wave method. It builds an equivalent circuit model by combining the precise electrical parameters of the cable, and uses the collected circulating current data according to the formula:

[0111] =

[0112] in Represents wave speed. , These are the circulating currents at the monitoring points when a fault occurs at either end of the line. , These are the circulating currents at the monitoring points at both ends of the line during normal operation. , is the selected time interval, is the total length of the cable, thereby achieving accurate determination of the location of the fault point;

[0113] The operation model module is used to simulate the operation conditions of the cable under multiple working conditions such as daily, peak, and after fault repair, and provides a key reference basis for in-depth diagnosis and precise location of faults.

[0114] Specifically as follows:

[0115] I. System Hardware Construction and Debugging

[0116] Installation and Calibration of the Cable Sheath Circulating Current Fault Information Acquisition Device

[0117] The cable sheath circulating current fault information acquisition device selects a high-sensitivity and low-noise current sensor, and the sensor is equipped with an adaptive anti-interference circuit, which can automatically adjust the filtering parameters according to the intensity of external interference to ensure the purity of the collected signal.

[0118] Sensor Selection and Installation: Carefully select a high-sensitivity and low-noise current sensor, and its technical indicators should meet the accurate perception requirements of the微小变化 (slight changes) of the cable sheath circulating current. During the installation process, strictly follow the electrical equipment installation specifications to ensure that the sensor is tightly and firmly connected to the cable sheath, preventing measurement errors caused by improper installation. For example, use a professional fixing clamp to fix the sensor at the specified monitoring point of the cable sheath, ensure good contact between the sensor electrode and the cable sheath, and achieve efficient signal transmission.

[0119] Adaptive Anti-Interference Circuit Debugging: Conduct detailed debugging on the adaptive anti-interference circuit supporting the sensor. In the laboratory environment, simulate external interference sources of different intensities and types, such as power frequency interference and electromagnetic radiation interference, and observe the suppression effect of the circuit on the interference signal. By adjusting circuit parameters such as the filter cut-off frequency and gain, make the circuit automatically optimize the filtering performance according to the interference intensity, ensure that the collected cable sheath circulating current signal is pure and stable, and provide a reliable data basis for subsequent analysis.

[0120] II. System Software Configuration and Operation Process

[0121] (I) Construction and Function Realization of the Operation Display Module

[0122] The operation display module supports multi-language switching to meet the needs of users in different regions, and can present the internal circulating current trend of the cable and the schematic diagram of the fault occurrence location in the form of a 3D dynamic graph.

[0123] Human-Computer Interface Design: Using the LabVIEW development platform, a rich set of graphical programming tools is employed to construct the human-computer interface for the operation and display module. The interface layout should be concise and clear, with well-defined functional areas, including a system control button area, a real-time data display area, and a graphical result display area. For example, the control button area includes buttons such as "Start Acquisition," "Stop Acquisition," and "Parameter Setting" for user convenience; the data display area updates key parameters such as cable sheath circulating current, temperature, and voltage in real-time in digital form; and the graphical display area uses intuitive charts (such as line graphs and bar charts) and 3D dynamic models to present the changing trends of circulating current data and the internal structure and circulation direction of the cable.

[0124] Multilingual switching and 3D dynamic graph rendering functionality: For multilingual switching, interface text resources for various commonly used languages ​​(such as Chinese, English, Japanese, etc.) were first collected and organized, and stored as language pack files. In the program, a language selection dropdown menu triggers a language switching event, and the corresponding language pack text is called based on the user's selection to update the displayed interface content. For 3D dynamic graph rendering, LabVIEW's 3D drawing function library is used, combined with the cable's geometric parameters and real-time circulating current data, to construct a 3D model of the cable. By dynamically updating the direction and size of the vector arrows representing the circulating current in the model, and using color gradients to represent the distribution of circulating current intensity, a smooth animation effect is used to display the dynamic changes of the circulating current inside the cable and the location of the fault. To ensure a stable frame rate, the drawing algorithm was optimized, such as reducing unnecessary graphics rendering calculations and using an efficient data caching mechanism, to ensure a good visual experience for the user.

[0125] (II) Operation steps of the fault type identification module based on wavelet analysis

[0126] The fault type identification module based on wavelet analysis has an additional built-in machine learning-assisted discrimination unit. The model is trained using a large amount of fault sample data in the early stage. When the wavelet analysis results are questionable, this unit is activated for secondary discrimination to improve the accuracy of fault identification.

[0127] Wavelet transform parameter settings and data processing: The cable sheath circulating current data acquired by the acquisition device is input into the fault type identification module based on Daubechies wavelet. First, based on the characteristics of the cable and previous fault data analysis experience, the number of decomposition levels of the wavelet transform is reasonably determined. Generally, for complex cable systems and variable fault types, a larger decomposition level may be needed to extract richer fault feature information, but the limitations of computing resources and processing time must also be considered. For example, after multiple comparative experiments, for the analysis of common medium-voltage cable sheath circulating current faults, a preliminary selection can be made. Then, according to the formula

[0128]

[0129] Perform wavelet transform to obtain the scaling coefficients. and wavelet coefficients During the calculation process, the numerical calculation function and efficient array processing mechanism of LabVIEW are fully utilized to ensure the accuracy and speed of the calculation.

[0130] Fault Feature Extraction and Intelligent Discrimination: In-depth fault feature extraction and analysis are performed on the wavelet coefficients obtained from wavelet transform. Various statistical features of the wavelet coefficients are calculated, such as energy distribution, mean amplitude, variance, and kurtosis. These features reflect fault-related information such as the energy concentration and fluctuation characteristics of the circulating signal at different frequency bands. Simultaneously, a machine learning-assisted discrimination unit trains the model using a large amount of previously collected labeled fault sample data. A classification model is constructed using machine learning algorithms such as Support Vector Machine (SVM) and neural networks, with the extracted wavelet coefficient features used as model input. During model training, techniques such as cross-validation are used to optimize model parameters and improve the model's generalization ability. When the wavelet analysis results are abnormal or outside the preset confidence interval, the machine learning-assisted discrimination unit is activated. For example, when the energy distribution of the wavelet coefficients suddenly changes in a specific frequency band and exceeds the normal fluctuation range, the current wavelet coefficient feature vector is input into the trained model. The model performs secondary discrimination based on the learned fault modes and outputs the final fault type judgment result, such as short-circuit fault, open-circuit fault, insulation aging fault, etc., effectively improving the accuracy of fault type identification.

[0131] (III) Detailed Operation Process of Fault Location Algorithm Module

[0132] During the calculation process, the fault location algorithm module takes into account the influence of changes in the electromagnetic field around the cable on the circulating current data and introduces a compensation coefficient. This minimizes positioning errors caused by environmental factors.

[0133] Equivalent circuit model construction and parameter initialization: In the fault location algorithm module, the cable resistance is first accurately measured based on the cable's detailed electrical parameter manual and actual measurement data. ,inductance ,capacitance Parameters such as these are used to build an equivalent circuit model of the cable in the software, and circuit simulation algorithms are employed to simulate the electrical characteristics of the cable under normal and fault conditions. For example, for long-distance cable lines, considering the influence of distributed parameters, a distributed parameter circuit model is used for more accurate simulation. During model building, parameters are initialized and their rationality is verified to ensure that the model accurately reflects the electrical behavior of the cable.

[0134] Fault location calculation based on the two-terminal traveling wave method: The circulating current at the monitoring points when a fault occurs at both ends of the line is collected. , and circulating current during normal operation , Import the fault location algorithm. Based on the formula... = Calculations are performed. This requires determining the time interval. , Make a selection;

[0135] Specifically, through analysis of a large amount of historical fault data and simulated fault experiments, it was determined that within a specific time window before and after the fault occurred, the circulating current signal contained rich fault traveling wave information. For most cable short-circuit faults, this information can be selected after the fault occurred. to time interval as to wave speed The characteristics of the cable's material and structure are determined using empirical formulas or actual measurement data. When calculating the integral term, numerical integration algorithms (such as the trapezoidal rule or Simpson's rule) are employed for precise calculation to ensure the accuracy of the results.

[0136] Compensation calculations considering the influence of electromagnetic fields: To reduce the impact of changes in the electromagnetic field around the cable on the circulating current data, high-precision electromagnetic field sensors installed around the cable are used to monitor electromagnetic intensity parameters in real time. Based on electromagnetic field theory and extensive experimental data, a mathematical model between electromagnetic field intensity and circulating current error is established and converted into compensation coefficients. The calculation function;

[0137] The software calculates the compensation coefficient using a built-in function based on real-time monitored electromagnetic field data. And substitute it into the revised formula = This enables precise calculation of the fault location, effectively improving the accuracy of fault location. Among other things... The positioning error is calculated in real time based on the electromagnetic intensity parameters monitored in real time through a built-in function, minimizing the positioning error caused by environmental factors.

[0138] (iv) Execution process of the model module function

[0139] The operating model module has an intelligent scene switching function, which can automatically switch to the corresponding typical working condition simulation based on the real-time collected cable operating parameters, and can generate a comparison report between the simulated working condition and the actual working condition to assist maintenance personnel in analysis and judgment.

[0140] The operational model module constructs a high-fidelity digital twin, combines machine learning algorithms to update model parameters in real time, and continuously corrects model predictions based on Bayesian inference formulas and real-time monitoring data to simulate the entire life cycle operation of the cable, thereby improving the fit with actual operational data.

[0141] Model Initialization and Parameter Setting: Upon startup, the model module builds a basic operating model based on the cable's fundamental design parameters (such as cable length, conductor cross-sectional area, and insulation material properties) and historical operating data (such as circulating current data, temperature variation curves, and load conditions under different operating conditions). Statistical methods and machine learning algorithms are used to analyze and process the historical data, extracting key features and patterns as initial parameter settings for the model. Specifically, cluster analysis algorithms are used to classify historical operating condition data, determining the typical value ranges and trends of cable operating parameters under different operating conditions, providing initial state estimates for the model.

[0142] Real-time data-driven operating condition simulation and model updates: During system operation, the operating parameters of the cable, such as temperature, voltage, current, and circulating current, are acquired in real time through data acquisition devices. Using this real-time data, and based on preset operating condition discrimination logic and threshold judgment conditions, the system automatically switches to the corresponding typical operating condition simulation.

[0143] Specifically, when the cable current exceeds 80% of the rated value and lasts for more than 10 minutes, the cable is determined to be operating under peak load conditions, and the model switches to the corresponding simulation mode. Simultaneously, machine learning algorithms, such as online learning neural networks or incremental support vector machines, are used to continuously update the model parameters based on newly collected data. This is achieved by utilizing Bayesian inference formulas. Real-time monitoring data is used as evidence when updating the model. Continuously refine the model's predictions of cable operating conditions under different operating circumstances. Through continuous iterative updates, the model can closely track changes in the actual operating status of the cable, simulating its operation throughout its entire lifecycle. For posterior probability, Let be the likelihood function. For prior probability, As evidence.

[0144] Comparison Report Generation and Auxiliary Analysis: During simulation operation, the operating model module periodically generates comparison reports between simulated and actual operating conditions. The reports include comparison curves of key operating parameters (such as circulating current amplitude, phase, and temperature changes), deviation statistical analysis, and trend prediction comparisons. Data visualization technology and statistical analysis methods are used to present complex data information to maintenance personnel in an intuitive and easy-to-understand manner. Specifically, line graphs are used to compare the changes in simulated and actual circulating current amplitude over time. The degree of deviation between the two is quantified by calculating indicators such as the root mean square error (RMSE) and correlation coefficient, and trend line fitting is used to analyze the future trends of both. Maintenance personnel can gain a deeper understanding of the cable's operating status based on the comparison reports, promptly identify potential problems, and receive comprehensive and powerful reference information for fault diagnosis and location.

[0145] (V) Working Mechanism and Implementation of Intelligent Early Warning Module

[0146] The intelligent early warning module learns and predicts cable operation data sequences based on a long short-term memory network, and evaluates the data by comparing the predicted value with a threshold through a risk assessment function.

[0147] Data Preprocessing and Model Training: The intelligent early warning module first performs comprehensive data preprocessing on a large amount of historical cable operation data. This includes data cleaning to remove outliers and noise; and data normalization to unify operating parameters with different dimensions to the same numerical range, facilitating model training and computation. For example, the Z-score normalization method is used to convert temperature, current, and other data into standard normal distribution data with a mean of 0.5 and a standard deviation of 0.5. Then, the processed historical data is divided into training and validation sample sets according to a specific time series, and input into a prediction model based on a Long Short-Term Memory (LSTM) network for training. During training, the weights and bias parameters of the LSTM network are adjusted based on the error between the predicted and actual values, and the Adam optimizer is used to minimize the loss function, improving the model's prediction accuracy and generalization ability.

[0148] Real-time prediction and risk assessment: In actual operation, the intelligent early warning module continuously receives new cable operation data sequences and inputs them into a trained LSTM model for real-time prediction. Based on the comparison between the predicted value and a preset threshold, a risk assessment function is used to determine the risk.

[0149]

[0150] An assessment was conducted, in which... This is the risk value. For feature weights, For input features, This is the feature mapping function. Feature weights. The feature mapping function was determined through analysis of historical fault data and a joint principal component analysis-linear discriminant analysis (PCA-LDA) algorithm, reflecting the contribution of different input features to fault risk. Based on the cable's operating characteristics and fault mechanisms, a nonlinear mapping function is used to convert the circulating current amplitude characteristics into an index more relevant to fault risk. This is applied in calculating the risk value. During the process, calculations were performed strictly according to the formula to ensure the accuracy of the evaluation results.

[0151] Early warning triggering and information output: When the calculated risk value When the set safety threshold is exceeded, the intelligent early warning module immediately triggers the early warning mechanism, sending warning information to maintenance personnel through various means, such as displaying a prominent pop-up notification on the operation display interface and simultaneously emitting an audible alarm at a specific frequency. The warning information displays detailed data related to the current risk assessment, including each input feature value, predicted value, and risk value. The calculation process and results, along with the analysis of possible fault causes and suggested measures based on historical fault cases and expert experience, are presented. Specifically, if the risk assessment indicates a short-circuit risk in the cable, the warning message can indicate the possible location and range of the fault point, suggest checking whether the cable joints are loose or the insulation is damaged, etc., providing strong support for maintenance personnel to take timely and effective maintenance measures and ensuring the safe and stable operation of the cable system.

[0152] Embodiment 3 of the present invention provides a terminal, including a processor and a storage medium; the storage medium is used to store instructions; the processor is used to perform operations according to the instructions to execute the steps of the method.

[0153] Embodiment 4 of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method.

[0154] Compared with the prior art, the beneficial effects of the present invention include at least the following:

[0155] 1. The fault location module of this invention, based on the improved double-ended traveling wave method and precise electrical parameter modeling, combined with an advanced wavelet analysis fault type identification module, can accurately determine the fault location and accurately identify the fault type, effectively shortening the fault investigation time, improving the efficiency of power cable operation and maintenance, and reducing power outage losses. By incorporating the spatial attenuation characteristics of electromagnetic interference into compensation, the interference effects at different locations are dynamically corrected, reducing the location error, making it particularly suitable for areas with strong electromagnetic interference near substations.

[0156] 2. When assessing risk values ​​based on predicted data, this invention overcomes the limitations of fixed weights by dynamically adapting weights to the fault type, effectively improving the accuracy of risk assessment and preventing key features from being masked by secondary features. Feature mapping is achieved based on sensitivity adjustment coefficients; the greater the deviation, the more the sensitivity increases exponentially, strengthening the response to serious deviations from normal conditions, avoiding over-warning of minor fluctuations, and ensuring rapid identification of major anomalies.

[0157] 3. The LSTM model of this invention is a fusion structure that introduces an attention mechanism and residual connections, and adopts a dynamic phased training strategy. It is trained using a weighted hybrid loss function. It combines the attention mechanism to focus on key features, dynamically trains to adapt to the entire life cycle, and the hybrid loss function balances prediction accuracy and robustness. It can improve the sensitivity to early weak fault signals and enhance the model's generalization ability under complex working conditions.

[0158] 4. Low-power continuous monitoring ensures operation: The cable sheath circulating current fault information acquisition device has a built-in micro energy acquisition device that uses the electromagnetic field energy around the cable to convert energy into power, achieving low-power continuous operation, reducing dependence on external power supply, reducing maintenance costs, and enhancing the system's long-term stable monitoring capability in complex environments.

[0159] 5. The multilingual human-machine interface and 3D dynamic graph display function of the operation display module facilitate operation by personnel in different regions, make cable circulation and fault information intuitive and visible, and enable maintenance personnel to quickly grasp the cable operating status, thereby improving the convenience of operation and the efficiency of information acquisition.

[0160] 6. Through intelligent scene switching of the running model module, Bayesian inference update, and long and short time memory network prediction and risk assessment of the intelligent early warning module, it can adapt to changes in cable operating conditions, correct the model in real time, provide accurate early warning, reduce environmental interference, and significantly improve the accuracy and reliability of system fault diagnosis and location.

[0161] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0162] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0163] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0164] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0165] 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 it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A loop-in-the-loop simulation method for locating cable sheath circulating current faults based on LabVIEW, characterized in that, include: Step 1: Under the control of LabVIEW, the current sensor collects the circulating current data of the cable sheath monitoring point in real time based on micro energy, and displays it digitally and graphically on LabVIEW; Step 2: Decompose the circulation data using wavelet transform to obtain the corresponding wavelet coefficients, and calculate the wavelet coefficient features. When the wavelet coefficient features are abnormal, input them into the pre-trained classification model to obtain the fault type. Step 3: Calculate the compensation coefficient that takes into account the influence of electromagnetic field based on the real-time monitored electromagnetic field data, and calculate the location of the fault point using the two-end traveling wave method. Step 4: The operating model uses real-time data of fault type, fault location and cable operating parameters, and automatically switches to the corresponding typical operating condition simulation mode through preset operating condition discrimination logic and threshold judgment conditions. It also continuously corrects the operating model's prediction of cable operating status under different simulated operating conditions through Bayesian inference, simulates the cable's full life cycle operating status, and provides risk assessment reference. Step 5: Input the real-time operating parameters of the cable into the trained LSTM model for real-time prediction, evaluate the risk value based on the prediction data, and trigger the fault warning mechanism when the risk value exceeds the safety threshold.

2. The loop-in-the-loop simulation method for locating cable sheath circulating current faults based on LabVIEW according to claim 1, characterized in that: Step 1, which describes digital display and graphical presentation on LabVIEW, includes: The LabVIEW data display area updates the circulating current values, temperature, and voltage of the cable sheath monitoring points in real time in digital form; the LabVIEW graphical display area uses intuitive charts and 3D dynamic models to present the changing trends of the circulating current data and the internal structure and circulation direction of the cable.

3. The loop-in-the-loop simulation method for locating cable sheath circulating current faults based on LabVIEW according to claim 1, characterized in that: Step 2 involves using wavelet transform to decompose the circulating current data of the cable sheath, as shown in the following formula: in, and Decomposition layers The k-th scaling factor and wavelet coefficient; This is the nth circulation data; and Decompose respectively Time-decomposition layer Next, the k-th parent wavelet and the mother wavelet; This represents the number of decomposition levels in the wavelet transform.

4. The loop-in-the-loop simulation method for locating cable sheath circulating current faults based on LabVIEW according to claim 1, characterized in that: In step 3, the formula for calculating the compensation coefficient considering the influence of the electromagnetic field based on the real-time monitored electromagnetic field data is as follows: in This is the compensation coefficient; The deviation between the real-time monitored electromagnetic field strength and the historical average; This represents the historical average electromagnetic field strength. This indicates the spatial attenuation characteristics of electromagnetic interference. The distance between the historical fault point and the electromagnetic interference source; , This is the calibration coefficient.

5. The loop-in-the-loop simulation method for locating cable sheath circulating current faults based on LabVIEW according to claim 1, characterized in that: In step 3, the compensation coefficient is used in conjunction with the two-terminal traveling wave method to calculate the location of the fault point, as shown in the following formula: = in, Location of the fault; , The circulating current at monitoring point t is the current at time t when there is a fault at ends A and B of the cable sheath. , This represents the circulating current at monitoring point t at points A and B of the cable sheath during normal operation. , The start and end times of the selected time interval; This is the compensation coefficient; This refers to the length of the cable sheath. This represents the propagation speed of the traveling wave.

6. The loop-in-the-loop simulation method for locating cable sheath circulating current faults based on LabVIEW according to claim 1, characterized in that: In step 5, the formula for assessing the risk value based on the predicted data is as follows: in, Risk value; For feature weights; The basic weights are determined through the analysis of historical fault data and the joint principal component analysis-linear discriminant analysis (PCA-LDA) algorithm. The degree of matching between the feature and the current fault type; For adjustment coefficients; It is a feature mapping function; Let be the i-th input feature obtained from the predicted data; n is the total number of features.

7. The loop-in-the-loop simulation method for locating cable sheath circulating current faults based on LabVIEW according to claim 6, characterized in that: Feature mapping function for: in, These are normal reference values ​​for the input features; This is the sensitivity adjustment coefficient. As a baseline sensitivity, This is the magnification factor.

8. The loop-in-the-loop simulation method for locating cable sheath circulating current faults based on LabVIEW according to claim 1, characterized in that: The LSTM model is a fusion structure that incorporates an attention mechanism and residual connections. The attention mechanism assigns dynamic weights to the parameter sequences of different operating stages of the cable, and the residual connections add skip connections between the hidden layers of the LSTM.

9. The loop-in-the-loop simulation method for locating cable sheath circulating current faults based on LabVIEW according to claim 1, characterized in that: The LSTM model adopts a dynamic phased training strategy. In the first phase, it is pre-trained using historical normal operation data. In the second phase, it is fine-tuned by introducing fault simulation data containing fault samples under different operating conditions. In the third phase, it absorbs newly collected operation and maintenance data in real time through an online learning module and updates the model parameters through a sliding window.

10. The loop-in-the-loop simulation method for locating cable sheath circulating current faults based on LabVIEW according to claim 1, characterized in that: The loss function of the LSTM model is a weighted mixed loss function: in Mean square error, For cross-entropy, The mean absolute error, , , This is a dynamically adjusted coefficient.

11. A LabVIEW-based loop-in-the-loop simulation system for locating cable sheath circulating current faults, used to implement the method described in any one of claims 1-10, characterized in that, The system comprises hardware and software components. The hardware component includes a current sensor and its built-in miniature energy harvesting device, along with a matching adaptive anti-interference circuit, for real-time and accurate acquisition of circulating current data at cable sheath monitoring points based on miniature energy. The software component includes: The operation display module is built on the LABVIEW development platform. The human-computer interaction interface of the operation display module includes a system control button area, a real-time data display area, and a graphical result display area. It is used to control the current sensor to collect the circulating current data of the cable sheath monitoring point in real time based on micro energy, and to display it digitally and graphically. The fault type identification module is used to decompose the circulation data using wavelet transform to obtain the corresponding wavelet coefficients and calculate the wavelet coefficient features. When the wavelet coefficient features are abnormal, they are input into a pre-trained classification model to obtain the fault type. The fault location module is used to calculate the compensation coefficient that takes into account the influence of electromagnetic field based on real-time monitored electromagnetic field data, and to calculate the fault location by combining the two-end traveling wave method. The running model module is used to run the model by utilizing real-time data of the cable's operating parameters, automatically switching to the corresponding typical operating condition simulation mode through preset operating condition discrimination logic and threshold judgment conditions, and continuously correcting the running model's prediction of the cable's operating status under different simulated operating conditions through Bayesian inference, simulating the cable's operating status throughout its entire life cycle to provide a risk assessment reference. The intelligent early warning module is used to input real-time data of the cable's operating parameters into a trained LSTM model for real-time prediction. Based on the predicted data, the risk value is evaluated, and when the risk value exceeds the safety threshold, a fault early warning mechanism is triggered.

12. A terminal, comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1-10.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1-10.

Images