A cable aging fault prediction method and device, electronic equipment and storage medium

Abstract

The application relates to the technical field of cable detection, and discloses a cable aging fault prediction method and device, electronic equipment and a storage medium, which are used to solve the technical problem that the cable prediction accuracy is low, early warning is lagged, and the demand for intelligent operation and maintenance cannot be met. The application comprises the following steps: collecting multi-dimensional data and an operation life of a cable; pre-processing the multi-dimensional data to obtain a plurality of types of pre-processed data; calculating the insulation resistance decay rate and the local discharge growth rate of the cable; calculating an aging coefficient according to the insulation resistance decay rate and the local discharge growth rate; extracting key features according to the pre-processed data and the aging coefficient; performing prediction by using the key features to obtain a predicted characteristic value; performing feature matching on the predicted characteristic value to obtain an aging grade of the cable; calculating a fault probability according to the aging grade, the predicted characteristic value and the operation life; and performing fault grading early warning according to the fault probability.

Description

Technical Field

[0001] This invention relates to the field of cable testing technology, and in particular to a method, device, electronic equipment, and storage medium for predicting cable aging faults. Background Technology

[0002] As the core carrier of power transmission systems, the operational status of power cables directly affects the reliability and stability of the entire power network. With the continuous increase in the service life of cables, and the combined effects of complex external environmental factors (such as temperature, humidity, chemical corrosion, and mechanical stress), the cable body inevitably experiences a series of aging problems, including insulation material degradation, conductor oxidation and corrosion, and shielding layer damage. If these potential defects and hidden dangers are not detected and accurately predicted in a timely manner, once a fault occurs, it can easily lead to cable insulation breakdown, line short circuits, and even large-scale power outages, causing significant economic losses and social impact. Currently, most cable fault prediction methods commonly used in the industry rely on periodic measurement or online monitoring of a single state parameter (e.g., monitoring only the insulation resistance value). This single-dimensional monitoring method has obvious limitations: the acquired feature information is insufficient in dimension, making it difficult to comprehensively and profoundly reflect the overall health status of the cable; at the same time, the constructed prediction models are often structurally simple and lack adaptability. These shortcomings collectively lead to low accuracy in fault prediction, delayed early warning signals, and an inability to make forward-looking and accurate assessments of cable conditions. Consequently, they fail to meet the urgent needs of modern power systems to transform and upgrade to digital and intelligent operation and maintenance models. Summary of the Invention

[0003] This invention provides a method, device, electronic equipment, and storage medium for predicting cable aging faults, which solves the technical problems of low prediction accuracy and delayed early warning, failing to meet the needs of intelligent operation and maintenance.

[0004] This invention provides a method for predicting cable aging faults, comprising:

[0005] Collect multidimensional data and service life of the cable;

[0006] The multidimensional data is preprocessed to obtain multiple types of preprocessed data;

[0007] Calculate the insulation resistance attenuation rate and the partial discharge rate increase of the cable;

[0008] The aging coefficient is calculated based on the insulation resistance attenuation rate and the partial discharge rate increase rate.

[0009] Based on the preprocessed data and the aging coefficient, key features are extracted;

[0010] The key features are used for prediction to obtain predicted feature values;

[0011] The aging level of the cable is obtained by performing feature matching on the predicted feature values;

[0012] The failure probability is calculated based on the aging level, the predicted characteristic value, and the years of operation.

[0013] Fault classification and early warning are performed based on the aforementioned fault probability.

[0014] Optionally, the multidimensional data includes electrical parameters, environmental parameters, and physical state parameters.

[0015] Optionally, the step of preprocessing the multidimensional data to obtain multiple types of preprocessed data includes:

[0016] Outliers in the multidimensional data are removed using the box plot method to obtain outlier-removed data;

[0017] Based on the K-nearest neighbor difference method, missing values ​​are filled in the abnormal data to obtain the completed data;

[0018] The completed data is standardized to obtain multiple types of preprocessed data.

[0019] Optionally, the step of calculating the insulation resistance attenuation rate and the partial discharge rate increase of the cable includes:

[0020] Obtain the initial insulation resistance and initial partial discharge of the new cable;

[0021] Obtain the current insulation resistance and current partial discharge level of the cable;

[0022] Calculate the ratio of the current insulation resistance to the initial insulation resistance of the new cable to obtain the insulation resistance attenuation rate of the cable;

[0023] The ratio between the current partial discharge quantity and the initial partial discharge quantity is calculated to obtain the partial discharge growth rate of the cable.

[0024] Optionally, the step of extracting key features based on the preprocessed data and the aging coefficient includes:

[0025] Based on the preprocessed data, the correlation between each type of preprocessed data and each type of aging coefficient is calculated;

[0026] Preprocessed data corresponding to types with a correlation greater than a preset threshold are identified as key features.

[0027] Optionally, the step of performing feature matching on the predicted feature values ​​to obtain the aging level of the cable includes:

[0028] Obtain a preset aging feature library, which includes multiple levels of aging feature sets;

[0029] Calculate the similarity between the predicted feature values ​​and the aging feature sets of multiple levels, respectively;

[0030] The aging level of the cable is determined by the aging feature set with the highest similarity.

[0031] Optionally, the step of calculating the failure probability based on the aging level, the predicted characteristic value, and the years of operation includes:

[0032] The basic failure probability is obtained by mapping the aging level.

[0033] Obtain the material of the cable and determine the attenuation coefficient based on the material;

[0034] Using the aforementioned operating years and the aforementioned attenuation coefficient, combined with an exponential attenuation model, the probability of attenuation failure is generated.

[0035] Query a preset historical fault database for a preset number of similar cases that have the highest similarity to the predicted feature value and the actual fault probability corresponding to the similar cases;

[0036] The similar failure probability of the cable is calculated based on the actual failure probability of all the similar cases.

[0037] The cable's failure probability is obtained by weighted fusion of the basic failure probability, the attenuation failure probability, and the similarity failure probability.

[0038] The present invention also provides a cable aging fault prediction device, comprising:

[0039] The multi-dimensional data and service life acquisition module is used to collect multi-dimensional data and service life of the cable.

[0040] The preprocessing module is used to preprocess the multidimensional data to obtain multiple types of preprocessed data;

[0041] An insulation resistance attenuation rate and partial discharge growth rate calculation module is used to calculate the insulation resistance attenuation rate and partial discharge growth rate of the cable.

[0042] An aging coefficient calculation module is used to calculate the aging coefficient based on the insulation resistance attenuation rate and the partial discharge growth rate.

[0043] A key feature extraction module is used to extract key features based on the preprocessed data and the aging coefficient;

[0044] A prediction feature value generation module is used to make predictions using the key features to obtain prediction feature values.

[0045] An aging level generation module is used to perform feature matching on the predicted feature values ​​to obtain the aging level of the cable.

[0046] The failure probability calculation module is used to calculate the failure probability based on the aging level, the predicted feature value, and the years of operation.

[0047] The fault classification and early warning module is used to classify and warn of faults based on the fault probability.

[0048] The present invention also provides an electronic device, the device comprising a processor and a memory:

[0049] The memory is used to store program code and transmit the program code to the processor;

[0050] The processor is used to execute the cable aging fault prediction method as described above, according to the instructions in the program code.

[0051] The present invention also provides a computer-readable storage medium for storing program code for executing the cable aging fault prediction method as described in any of the preceding claims.

[0052] As can be seen from the above technical solutions, the present invention has the following advantages: The present invention discloses a method for predicting cable aging faults, and specifically discloses: collecting multidimensional data and operating years of the cable; preprocessing the multidimensional data to obtain multiple types of preprocessed data; calculating the insulation resistance attenuation rate and partial discharge growth rate of the cable; calculating the aging coefficient based on the insulation resistance attenuation rate and partial discharge growth rate; extracting key features based on the preprocessed data and aging coefficient; using the key features for prediction to obtain predicted feature values; performing feature matching on the predicted feature values ​​to obtain the aging level of the cable; calculating the fault probability based on the aging level, predicted feature values, and operating years; and performing fault classification and early warning based on the fault probability.

[0053] This invention acquires basic information about cable operating status through multi-dimensional data acquisition, eliminates interference factors through data preprocessing, extracts key features related to aging, and uses a prediction architecture that fuses time-series prediction and feature matching to assess cable aging and predict the probability of fault occurrence, generating tiered early warning signals. The closed-loop design of each step improves the accuracy and timeliness of prediction, providing a scientific basis for cable operation and maintenance and reducing the risk of outages due to faults. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 This is a flowchart of the steps in a cable aging fault prediction method provided in Embodiment 1 of the present invention;

[0056] Figure 2 This is a flowchart of the steps in a cable aging fault prediction method provided in Embodiment 2 of the present invention;

[0057] Figure 3 This is a flowchart of the steps in a cable aging fault prediction method provided in Embodiment 3 of the present invention;

[0058] Figure 4 This is a flowchart of the steps in a cable aging fault prediction method provided in Embodiment 4 of the present invention;

[0059] Figure 5 This is a structural block diagram of a cable aging fault prediction device provided in an embodiment of the present invention. Detailed Implementation

[0060] As the core carrier of power transmission, the operating status of power cables directly affects the stability of the power system. With increasing service life and environmental factors, cables experience insulation aging and conductor corrosion. Failure to predict faults in a timely manner can lead to serious consequences such as cable breakdown and power outages. Existing cable fault prediction methods mostly rely on single-parameter monitoring (such as monitoring only insulation resistance), which suffers from insufficient feature dimensions and a single prediction model, resulting in low prediction accuracy and delayed early warning, failing to meet the needs of intelligent operation and maintenance. Therefore, this invention provides a cable aging fault prediction method, device, electronic equipment, and storage medium to solve the technical problems of low prediction accuracy and delayed early warning, which fail to meet the needs of intelligent operation and maintenance.

[0061] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0062] Please see Figure 1 , Figure 1 The flowchart illustrates the steps of a cable aging fault prediction method provided in Embodiment 1 of the present invention.

[0063] The present invention provides a method for predicting cable aging faults, which may specifically include the following steps:

[0064] Step 101: Collect multi-dimensional data and service life of the cable;

[0065] In this embodiment of the invention, the multidimensional data for the cable may include electrical parameters, environmental parameters, and physical state parameters.

[0066] Among them, electrical parameters can be obtained through equipment such as insulation resistance testers and partial discharge detectors, directly reflecting changes in cable insulation performance; environmental parameters are collected through temperature and humidity sensors and soil pH sensors to capture the accelerating effect of the external environment on aging; physical state parameters are obtained through infrared thermometers and sheath damage monitoring cameras to reflect the aging of cable appearance and structure.

[0067] In practical implementation, the data collection cycle can be set to 1 hour per collection, and the data is uploaded to the server in real time via a 5G transmission unit to ensure data timeliness. The original multidimensional dataset is defined as follows:

[0068] ;

[0069] Where n is the number of samples collected, m is the number of parameter types, and d i,j This represents the parameter value of the j-th class for the i-th sample.

[0070] Step 102: Preprocess the multidimensional data to obtain multiple types of preprocessed data;

[0071] In real-world scenarios, raw multidimensional data may contain outliers caused by sensor malfunctions, missing values ​​due to transmission interruptions, and inconsistencies in the dimensions of different parameters. Preprocessing is necessary to optimize data quality and obtain preprocessed data of multiple data types.

[0072] Step 103: Calculate the insulation resistance attenuation rate and the partial discharge growth rate of the cable;

[0073] Step 104: Calculate the aging coefficient based on the insulation resistance attenuation rate and the partial discharge growth rate;

[0074] In practice, cable aging is the result of multiple factors, among which the key factors include the insulation resistance attenuation rate and the partial discharge rate.

[0075] The aging coefficient of a cable can be calculated by measuring its insulation resistance attenuation rate and partial discharge growth rate.

[0076] Step 105: Extract key features based on preprocessed data and aging coefficient;

[0077] After calculating the cable aging coefficient, this key parameter can be effectively combined with preprocessed data. By comprehensively analyzing and matching the aging coefficient with the preprocessed data, the core characteristics and key indicators reflecting changes in cable condition can be identified and extracted more accurately and specifically from massive amounts of preprocessed information.

[0078] Step 106: Use key features to make predictions and obtain predicted feature values;

[0079] After collecting key features, a time-series prediction sub-model can be used to preset the predicted feature values ​​of the cable's future cyclic state. This allows for the determination of its aging degree based on the predicted feature values.

[0080] Step 107: Perform feature matching on the predicted feature values ​​to obtain the aging level of the cable;

[0081] After obtaining the predicted characteristic values ​​of the cable condition for future periods, feature matching can be performed on the predicted characteristic values ​​to obtain the aging level of the cable.

[0082] In one example, the predicted feature value can be matched with a preset aging feature library, and the aging level of the cable can be obtained based on the level of each feature in the aging feature library.

[0083] Step 108: Calculate the failure probability based on the aging level, predicted characteristic value, and years of operation;

[0084] In this embodiment of the invention, the failure probability needs to take into account multiple factors. Therefore, after determining the aging level, predicted characteristic value and service life of the cable, a failure probability can be calculated by using the aging level, predicted characteristic value and service life respectively, and the failure probability of the cable in the future period can be calculated by weighting the failure probabilities of each factor.

[0085] Step 109: Perform fault classification and early warning based on fault probability.

[0086] In this embodiment of the invention, based on the predicted probability of cable failure, a graded early warning of the failure can be performed, and corresponding processing strategies can be executed according to different graded early warnings of the failure.

[0087] This invention acquires basic information about cable operating status through multi-dimensional data acquisition, eliminates interference factors through data preprocessing, extracts key features related to aging, and uses a prediction architecture that fuses time-series prediction and feature matching to assess cable aging and predict the probability of fault occurrence, generating tiered early warning signals. The closed-loop design of each step improves the accuracy and timeliness of prediction, providing a scientific basis for cable operation and maintenance and reducing the risk of outages due to faults.

[0088] Please see Figure 2 , Figure 2 This is a flowchart illustrating the steps of a cable aging fault prediction method provided in Embodiment 2 of the present invention. Specifically, it may include the following steps:

[0089] Step 201: Collect multi-dimensional data and service life of the cable;

[0090] In this embodiment of the invention, the multidimensional data for the cable may include electrical parameters, environmental parameters, and physical state parameters.

[0091] Among them, electrical parameters can be obtained through equipment such as insulation resistance testers and partial discharge detectors, directly reflecting changes in cable insulation performance; environmental parameters are collected through temperature and humidity sensors and soil pH sensors to capture the accelerating effect of the external environment on aging; physical state parameters are obtained through infrared thermometers and sheath damage monitoring cameras to reflect the aging of cable appearance and structure.

[0092] In practical implementation, the data collection cycle can be set to 1 hour per collection, and the data is uploaded to the server in real time via a 5G transmission unit to ensure data timeliness. The original multidimensional dataset is defined as follows:

[0093] ;

[0094] Where n is the number of samples collected, m is the number of parameter types, and d i,j This represents the parameter value of the j-th class for the i-th sample.

[0095] Step 202: Use the box plot method to remove outliers from the multidimensional data to obtain outlier-removed data;

[0096] In real-world scenarios, raw multidimensional data may contain outliers caused by sensor malfunctions, missing values ​​due to transmission interruptions, and inconsistencies in the dimensions of different parameters. Preprocessing is necessary to optimize data quality and obtain preprocessed data of multiple data types.

[0097] Box plots are a statistical method based on quartiles to identify outliers in data, commonly used for data cleaning and quality control. They objectively define the boundary between normal data and outliers by quantifying the dispersion of the data.

[0098] First, the box plot method can be adopted to define the quartiles Q1 (lower quartile) and Q3 (upper quartile), with the interquartile range IQR=Q3-Q1, to eliminate abnormal data that satisfies d i,j <Q1-1.5×IQR or d i,j >Q3+1.5×IQR.

[0099] Step 203: Perform missing value imputation on the anomaly-eliminated data based on the K-nearest neighbor difference method to obtain imputed data;

[0100] In a specific implementation, based on the K-nearest neighbor difference method, for a missing value d i,j , select k samples with the closest Euclidean distance to the i-th sample (k is 5), and perform imputation through weighted average:

[0101] ;

[0102] ;

[0103] where, w t is the weight coefficient, and dist(i,t) is the Euclidean distance between sample i and sample t.

[0104] Step 204: Standardize the imputed data to obtain multiple types of preprocessed data;

[0105] In a specific implementation, Z-score standardization can be used to eliminate dimensional differences, with the formula:

[0106] ;

[0107] ;

[0108] ;

[0109] where, is the standardized data; μ j and σ j are respectively the mean value and standard deviation of the j-th type of parameter, and the standardized preprocessed data is:

[0110] .

[0111] Step 205: Calculate the insulation resistance attenuation rate and the partial discharge increment rate of the cable;

[0112] Step 206: Calculate the aging coefficient according to the insulation resistance attenuation rate and the partial discharge increment rate;

[0113] In a specific implementation, cable aging is the result of the combined action of multiple factors, among which the key factors include the insulation resistance attenuation rate and the partial discharge increment rate of the cable.

[0114] The aging coefficient of a cable can be calculated by measuring its insulation resistance attenuation rate and partial discharge growth rate.

[0115] Step 207: Extract key features based on preprocessed data and aging coefficient;

[0116] After calculating the cable aging coefficient, this key parameter can be effectively combined with preprocessed data. By comprehensively analyzing and matching the aging coefficient with the preprocessed data, the core characteristics and key indicators reflecting changes in cable condition can be identified and extracted more accurately and specifically from massive amounts of preprocessed information.

[0117] Step 208: Use key features to make predictions and obtain predicted feature values;

[0118] After collecting key features, a time-series prediction sub-model can be used to preset the predicted feature values ​​of the cable's future cyclic state. This allows for the determination of its aging degree based on the predicted feature values.

[0119] Step 209: Perform feature matching on the predicted feature values ​​to obtain the aging level of the cable;

[0120] After obtaining the predicted characteristic values ​​of the cable condition for future periods, feature matching can be performed on the predicted characteristic values ​​to obtain the aging level of the cable.

[0121] In one example, the predicted feature value can be matched with a preset aging feature library, and the aging level of the cable can be obtained based on the level of each feature in the aging feature library.

[0122] Step 210: Calculate the failure probability based on the aging level, predicted characteristic value, and years of operation;

[0123] In this embodiment of the invention, the failure probability needs to take into account multiple factors. Therefore, after determining the aging level, predicted characteristic value and service life of the cable, a failure probability can be calculated by using the aging level, predicted characteristic value and service life respectively, and the failure probability of the cable in the future period can be calculated by weighting the failure probabilities of each factor.

[0124] Step 211: Perform fault classification and early warning based on fault probability.

[0125] In this embodiment of the invention, based on the predicted probability of cable failure, a graded early warning of failure can be performed, and corresponding processing strategies can be executed according to different graded early warnings of failure.

[0126] This invention acquires basic information about cable operating status through multi-dimensional data acquisition, eliminates interference factors through data preprocessing, extracts key features related to aging, and uses a prediction architecture that fuses time-series prediction and feature matching to assess cable aging and predict the probability of fault occurrence, generating tiered early warning signals. The closed-loop design of each step improves the accuracy and timeliness of prediction, providing a scientific basis for cable operation and maintenance and reducing the risk of outages due to faults.

[0127] Please see Figure 3 , Figure 3 This is a flowchart illustrating the steps of a cable aging fault prediction method provided in Embodiment 3 of the present invention. Specifically, it may include the following steps:

[0128] Step 301: Collect multi-dimensional data and service life of the cable;

[0129] In this embodiment of the invention, the multidimensional data for the cable may include electrical parameters, environmental parameters, and physical state parameters.

[0130] Among them, electrical parameters can be obtained through equipment such as insulation resistance testers and partial discharge detectors, directly reflecting changes in cable insulation performance; environmental parameters are collected through temperature and humidity sensors and soil pH sensors to capture the accelerating effect of the external environment on aging; physical state parameters are obtained through infrared thermometers and sheath damage monitoring cameras to reflect the aging of cable appearance and structure.

[0131] In practical implementation, the data collection cycle can be set to 1 hour per collection, and the data is uploaded to the server in real time via a 5G transmission unit to ensure data timeliness. The original multidimensional dataset is defined as follows:

[0132] ;

[0133] Where n is the number of samples collected, m is the number of parameter types, and d i,j This represents the parameter value of the j-th class for the i-th sample.

[0134] Step 302: Preprocess the multidimensional data to obtain multiple types of preprocessed data;

[0135] In real-world scenarios, raw multidimensional data may contain outliers caused by sensor malfunctions, missing values ​​due to transmission interruptions, and inconsistencies in the dimensions of different parameters. Preprocessing is necessary to optimize data quality and obtain preprocessed data of multiple data types.

[0136] Box plots are a statistical method based on quartiles to identify outliers in data, commonly used for data cleaning and quality control. They objectively define the boundary between normal data and outliers by quantifying the dispersion of the data.

[0137] First, the box plot method may be adopted to define the quartiles Q1 (lower quartile) and Q3 (upper quartile), with the interquartile range IQR=Q3-Q1, to eliminate abnormal data that satisfies d i,j <Q1-1.5×IQR or d i,j >Q3+1.5×IQR.

[0138] Next, based on the K-nearest neighbor difference method, for a missing value d i,j , k samples with the shortest Euclidean distance to the i-th sample are selected (k is 5), and the missing value is completed by weighted average:

[0139] ;

[0140] ;

[0141] where w t is the weight coefficient, and dist(i,t) is the Euclidean distance between sample i and sample t.

[0142] Finally, Z-score normalization may be used to eliminate dimensional differences, with the formula as follows:

[0143] ;

[0144] ;

[0145] ;

[0146] where is the normalized data; μ j and σ j are respectively the mean value and standard deviation of the j-th type of parameter, and the normalized preprocessed data is:

[0147] .

[0148] Step 303: Calculate the insulation resistance decay rate and the partial discharge growth rate of the cable;

[0149] In specific implementation, cable aging is the result of the combined action of multiple factors, among which the key factors include the insulation resistance decay rate and the partial discharge growth rate of the cable.

[0150] In an example, the step of calculating the insulation resistance decay rate and the partial discharge growth rate of the cable may include the following sub-steps:

[0151] S31: Acquire the initial insulation resistance and the initial partial discharge amount of a new cable;

[0152] S32: Acquire the current insulation resistance and the current partial discharge amount of the cable;

[0153] S33, calculate the ratio of the current insulation resistance to the initial insulation resistance of the new cable to obtain the insulation resistance attenuation rate of the cable;

[0154] S34, calculate the ratio between the current partial discharge quantity and the initial partial discharge quantity to obtain the partial discharge growth rate of the cable.

[0155] In a practical implementation, the insulation resistance attenuation rate can be defined as follows:

[0156] ;

[0157] in, The current insulation resistance, The initial insulation resistance of the new cable.

[0158] The partial discharge rate growth rate is defined as follows:

[0159] ;

[0160] in, This represents the current partial discharge level. This represents the initial partial discharge quantity.

[0161] Step 304: Calculate the aging coefficient based on the insulation resistance attenuation rate and the partial discharge growth rate;

[0162] In practical implementation, the aging factor K can be obtained through weighted calculation:

[0163] ;

[0164] The value range is [0,1], where 0 represents a brand new cable and 1 represents a fully aged cable.

[0165] Step 305: Based on the preprocessed data, calculate the correlation between each type of preprocessed data and each type of aging coefficient;

[0166] After calculating the aging coefficient, the correlation between each type of preprocessed data and each type of aging coefficient can be calculated, thereby determining key features based on the correlation.

[0167] In practice, the correlation between various types of preprocessed data and the aging coefficient can be calculated using the Pearson correlation coefficient. The specific formula is as follows:

[0168] ;

[0169] in, Let K be the mean of the j-th type of preprocessed data. i Let be the aging coefficient of the i-th sample. This represents the average aging coefficient. The expressions are as follows:

[0170] ;

[0171] ;

[0172] Step 306: The preprocessed data corresponding to the types with a correlation greater than a preset threshold are identified as key features;

[0173] In this embodiment of the invention, preprocessed data corresponding to types with a correlation greater than a preset threshold can be identified as key features.

[0174] For example, Preprocessed data with a value ≥0.7 were identified as key features, forming a key feature set F:

[0175] ;

[0176] Where p is the number of key features, which is usually 6-8.

[0177] Step 307: Use key features to make predictions and obtain predicted feature values;

[0178] In this embodiment of the invention, a dual-model architecture can be used to improve the predictive reliability of cable aging levels.

[0179] First, key features are predicted using a time-series prediction sub-model (LSTM) to obtain the predicted feature values ​​of the cable for future cycles.

[0180] The input to the time-series prediction sub-model (LSTM) is a sequence of key features from the past 12 periods:

[0181] ;

[0182] Among them, F t Given the feature vector of the t-th period, the LSTM model captures temporal dependencies through a gating mechanism, and the hidden layer output formula is:

[0183] Forgotten Gate:

[0184] ;

[0185] Input Gate:

[0186] ;

[0187] Cell state:

[0188] ;

[0189] Output gate:

[0190] ;

[0191] Among them, f t Output for the forget gate; i t For input gate output; Candidate cell state; In cellular state; This represents the cell state at the previous moment; Output gate output; It is in a hidden state; This is the previously hidden state; The input vector at the current time step; Here is the forget gate weight matrix; The input gate weight matrix; This is the candidate cell state weight matrix; This is the output gate weight matrix; Forget gate bias term; For input gate bias terms; This refers to the candidate cell state bias term; σ is the output gate bias term; σ is the sigmoid activation function; tanh is the hyperbolic tangent function; · is the matrix multiplication operation; ⊙ is the element-wise product operation; the model output is the predicted feature values ​​for the next 3 periods. :

[0192] .

[0193] Step 308: Perform feature matching on the predicted feature values ​​to obtain the aging level of the cable;

[0194] After obtaining the predicted characteristic values ​​of the cable condition for future periods, feature matching can be performed on the predicted characteristic values ​​to obtain the aging level of the cable.

[0195] In one example, the predicted feature value can be matched with a preset aging feature library, and the aging level of the cable can be obtained based on the level of each feature in the aging feature library.

[0196] In one example, the steps of performing feature matching on the predicted feature values ​​to obtain the aging level of the cable include:

[0197] S81, Obtain the preset aging feature library, which includes multiple levels of aging feature sets;

[0198] S82, calculate the similarity between the predicted feature values ​​and the aging feature sets of multiple levels respectively;

[0199] S83 determines the aging level of the cable based on the aging feature set with the highest similarity.

[0200] In the specific implementation, an aging feature library L = {L1, L2, L3} is preset (L1 is the feature set for mild aging, L2 is the feature set for moderate aging, and L3 is the feature set for severe aging), and the RBF kernel function is used to calculate the predicted features. Similarity with feature sets at each level:

[0201] ;

[0202] Where x represents the predicted feature vector y represents the standard feature vector L of the preset aging level, and γ is set to 0.1. The SVM decision function outputs the aging level with the highest matching degree:

[0203] .

[0204] Step 309: Calculate the failure probability based on the aging level, predicted characteristic value, and years of operation;

[0205] In this embodiment of the invention, the failure probability needs to take into account multiple factors. Therefore, after determining the aging level, predicted characteristic value and service life of the cable, a failure probability can be calculated by using the aging level, predicted characteristic value and service life respectively, and the failure probability of the cable in the future period can be calculated by weighting the failure probabilities of each factor.

[0206] Step 310: Perform fault classification and early warning based on fault probability.

[0207] In this embodiment of the invention, based on the predicted probability of cable failure, a graded early warning of failure can be performed, and corresponding processing strategies can be executed according to different graded early warnings of failure.

[0208] This invention acquires basic information about cable operating status through multi-dimensional data acquisition, eliminates interference factors through data preprocessing, extracts key features related to aging, and uses a prediction architecture that fuses time-series prediction and feature matching to assess cable aging and predict the probability of fault occurrence, generating tiered early warning signals. The closed-loop design of each step improves the accuracy and timeliness of prediction, providing a scientific basis for cable operation and maintenance and reducing the risk of outages due to faults.

[0209] Please see Figure 4 , Figure 4 This is a flowchart illustrating the steps of a cable aging fault prediction method provided in Embodiment 4 of the present invention, which specifically includes the following steps:

[0210] Step 401: Collect multi-dimensional data and service life of the cable;

[0211] In this embodiment of the invention, the multidimensional data for the cable may include electrical parameters, environmental parameters, and physical state parameters.

[0212] Wherein, the electrical parameters can be obtained through devices such as insulation resistance testers and partial discharge detectors, and directly reflect changes in cable insulation performance; the environmental parameters are collected through temperature and humidity sensors and soil pH sensors to capture the accelerating effect of the external environment on aging; the physical state parameters are obtained through infrared thermometers and sheath damage monitoring cameras to reflect the aging condition of the cable's appearance and structure.

[0213] In a specific implementation, the collection period can be set to 1 hour / time, and the data is uploaded to the server in real time via the 5G transmission unit to ensure data timeliness. The original multi-dimensional dataset is defined as:

[0214] ;

[0215] Wherein, n is the number of collected samples, m is the number of parameter types, d i,j represents the j-th parameter value of the i-th sample.

[0216] Step 402: preprocess the multi-dimensional data to obtain multiple types of preprocessed data;

[0217] In actual scenarios, the original multi-dimensional data may have outliers caused by sensor failures, missing values caused by transmission interruptions, and inconsistent dimensions of different parameters. Preprocessing is required to optimize data quality and obtain preprocessed data of multiple data types.

[0218] The box plot method is a statistical method for identifying outliers in data based on quartiles, and is often used in data cleaning and quality control. It objectively defines the boundary between normal data and outliers by quantifying the degree of data dispersion.

[0219] First, the box plot method can be adopted to define the quartiles Q1 (lower quartile) and Q3 (upper quartile), the interquartile range IQR=Q3-Q1, and eliminate the abnormal data satisfying d i,j <Q1-1.5×IQR or d i,j >Q3+1.5×IQR.

[0220] Second, based on the K-nearest neighbor difference method, for the missing value d i,j , select k samples (k is 5) that are closest to the i-th sample in Euclidean distance, and complete the value through weighted average:

[0221] ;

[0222] ;

[0223] Wherein, w t is the weight coefficient, and dist(i,t) is the Euclidean distance between sample i and sample t.

[0224] Finally, Z-score standardization can be used to eliminate dimensional differences, as shown in the formula:

[0225] ;

[0226] ;

[0227] ;

[0228] in, For standardized data; μ j and σ j Let be the mean and standard deviation of the j-th parameter, respectively. The standardized preprocessed data are as follows:

[0229] .

[0230] Step 403: Calculate the insulation resistance attenuation rate and the partial discharge growth rate of the cable;

[0231] In a practical implementation, the insulation resistance attenuation rate can be defined as follows:

[0232] ;

[0233] in, The current insulation resistance, The initial insulation resistance of the new cable.

[0234] The partial discharge rate growth rate is defined as follows:

[0235] ;

[0236] in, This represents the current partial discharge level. This represents the initial partial discharge quantity.

[0237] Step 404: Calculate the aging coefficient based on the insulation resistance attenuation rate and the partial discharge growth rate;

[0238] In practical implementation, the aging factor K can be obtained through weighted calculation:

[0239] ;

[0240] The value range is [0,1], where 0 represents a brand new cable and 1 represents a fully aged cable.

[0241] Step 405: Extract key features based on preprocessed data and aging coefficient;

[0242] After calculating the aging coefficient, the correlation between each type of preprocessed data and each type of aging coefficient can be calculated, thereby determining key features based on the correlation.

[0243] In practice, the correlation between various types of preprocessed data and the aging coefficient can be calculated using the Pearson correlation coefficient. The specific formula is as follows:

[0244] ;

[0245] in, Let be the mean of the j-th type of preprocessed data, and Ki be the aging coefficient of the i-th sample. This represents the average aging coefficient. The expressions are as follows:

[0246] ;

[0247] ;

[0248] In this embodiment of the invention, preprocessed data corresponding to types with a correlation greater than a preset threshold can be identified as key features.

[0249] For example, Preprocessed data with a value ≥0.7 were identified as key features, forming a key feature set F:

[0250] ;

[0251] Where p is the number of key features, which is usually 6-8.

[0252] Step 406: Use key features to make predictions and obtain predicted feature values;

[0253] In this embodiment of the invention, a dual-model architecture can be used to improve the predictive reliability of cable aging levels.

[0254] First, key features are predicted using a time-series prediction sub-model (LSTM) to obtain the predicted feature values ​​of the cable for future cycles.

[0255] The input to the time-series prediction sub-model (LSTM) is a sequence of key features from the past 12 periods:

[0256] ;

[0257] Among them, F t Given the feature vector of the t-th period, the LSTM model captures temporal dependencies through a gating mechanism, and the hidden layer output formula is:

[0258] Forgotten Gate:

[0259] ;

[0260] Input Gate:

[0261] ;

[0262] Cell state:

[0263] ;

[0264] Output gate:

[0265] ;

[0266] Among them, f t Output for the forget gate; i t For input gate output; Candidate cell state; In cellular state; This represents the cell state at the previous moment; Output gate output; It is in a hidden state; This is the previously hidden state; The input vector at the current time step; Here is the forget gate weight matrix; The input gate weight matrix; This is the candidate cell state weight matrix; This is the output gate weight matrix; Forget gate bias term; For input gate bias terms; This refers to the candidate cell state bias term; σ is the output gate bias term; σ is the sigmoid activation function; tanh is the hyperbolic tangent function; · is the matrix multiplication operation; ⊙ is the element-wise product operation; the model output is the predicted feature values ​​for the next 3 periods. :

[0267] .

[0268] Step 407: Perform feature matching on the predicted feature values ​​to obtain the aging level of the cable;

[0269] After obtaining the predicted characteristic values ​​of the cable condition for future periods, feature matching can be performed on the predicted characteristic values ​​to obtain the aging level of the cable.

[0270] In the specific implementation, an aging feature library L = {L1, L2, L3} is preset (L1 is the feature set for mild aging, L2 is the feature set for moderate aging, and L3 is the feature set for severe aging), and the RBF kernel function is used to calculate the predicted features. Similarity with feature sets at each level:

[0271] ;

[0272] Where x represents the predicted feature vector y represents the standard feature vector L of the preset aging level, and γ is set to 0.1. The SVM decision function outputs the aging level with the highest matching degree:

[0273] .

[0274] Step 408: Obtain the basic failure probability based on the aging level mapping;

[0275] In practical implementation, the failure probability needs to be calculated using a weighted algorithm that considers multiple factors, including the base failure probability, the decaying failure probability, and the similar failure probability.

[0276] First, the basic failure probability P1 can be obtained by mapping the aging level, setting P1(1)=0.05 (mild aging), P1(2)=0.3 (moderate aging), and P1(3)=0.6 (severe aging).

[0277] Step 409: Obtain the material of the cable and determine the attenuation coefficient based on the material;

[0278] Step 410: Using the operating years and attenuation coefficient, combined with the exponential attenuation model, generate the attenuation failure probability;

[0279] The attenuation failure probability P2 can be based on the exponential attenuation model, taking into account the cable's service life t, and the formula is:

[0280] ;

[0281] The attenuation coefficient k is determined based on the cable material (k=0.08 for cross-linked polyethylene cables and k=0.12 for polyvinyl fluoride cables).

[0282] Step 411: Query the preset number of similar cases with the highest similarity to the predicted feature value and the actual failure probability corresponding to the similar cases from the preset historical fault database;

[0283] Step 412: Calculate the similar failure probability of the cable based on the actual failure probabilities of all similar cases;

[0284] The similarity probability P3 can be obtained by querying the historical fault database using the K-nearest neighbor algorithm and selecting faults with the current features. The average failure probability of the 10 most similar cases is calculated using the following formula:

[0285] ;

[0286] Among them, P sLet P be the actual failure probability of the s-th similar case. s The specific calculation method is divided into two scenarios:

[0287] Scenario 1: The actual failure probability P of a single historical case s :

[0288] If the historical cases are complete single-cycle cases (i.e., after recording features, only one clear result appears within the prediction time window), then P s This is a direct mapping of the fault state in this case:

[0289] ;

[0290] Scenario 2: Aggregated actual failure probability P of similar cases s :

[0291] If there are multiple historical cases with "highly consistent characteristics", then P s The probability of failure occurring in this type of case:

[0292] ;

[0293] in, The total number of similar historical cases with highly consistent characteristics; This represents the number of cases where a failure occurred within 3 hours of being recorded, among similar cases.

[0294] Step 413: Weighted fusion of basic fault probability, attenuation fault probability and similar fault probability to obtain the cable fault probability;

[0295] Finally, by weighted fusion of the basic fault probability, attenuation fault probability, and similar fault probability, the cable fault probability P can be obtained as follows:

[0296] ;

[0297] In this context, α, β, and θ are weighting coefficients (e.g., α=0.5, β=0.3, θ=0.2), and satisfy α+β+θ=1.

[0298] Step 414: Perform fault classification and early warning based on fault probability.

[0299] In real-world scenarios, three levels of early warning can be established based on the probability of failure:

[0300] 1) Level 1 warning: P ≥ 0.8 (fault probability ≥ 80%, high risk, immediate repair required);

[0301] 2) Level II warning: 0.5≤P<0.8 (fault probability 50%~80%, medium to high risk, inspection within 24 hours);

[0302] 3) Level 3 warning: 0.3≤P<0.5 (failure probability 30%~50%, medium risk, regular tracking and monitoring).

[0303] This invention acquires basic information about cable operating status through multi-dimensional data acquisition, eliminates interference factors after data output, extracts key features related to aging, and uses a predictive architecture that fuses time-series prediction and feature matching to assess cable aging and predict fault probability, generating tiered early warning signals. The closed-loop design of each step improves the accuracy and timeliness of prediction, providing a scientific basis for cable operation and maintenance and reducing the risk of outages due to faults.

[0304] Please see Figure 5 , Figure 5 This is a structural block diagram of a cable aging fault prediction device provided in an embodiment of the present invention.

[0305] This invention provides a cable aging fault prediction device, comprising:

[0306] The multidimensional data and service life acquisition module 501 is used to acquire multidimensional data and service life of the cable.

[0307] Preprocessing module 502 is used to preprocess multidimensional data to obtain multiple types of preprocessed data;

[0308] Insulation resistance attenuation rate and partial discharge growth rate calculation module 503 is used to calculate the insulation resistance attenuation rate and partial discharge growth rate of the cable.

[0309] The aging coefficient calculation module 504 is used to calculate the aging coefficient based on the insulation resistance attenuation rate and the partial discharge growth rate.

[0310] The key feature extraction module 505 is used to extract key features based on the preprocessed data and aging coefficient;

[0311] The predictive feature value generation module 506 is used to make predictions using key features to obtain predicted feature values.

[0312] The aging level generation module 507 is used to perform feature matching on the predicted feature values ​​to obtain the aging level of the cable.

[0313] The failure probability calculation module 508 is used to calculate the failure probability based on the aging level, predicted characteristic value and years of operation.

[0314] The fault classification and early warning module 509 is used to classify and warn of faults based on the probability of faults.

[0315] In this embodiment of the invention, the multidimensional data includes electrical parameters, environmental parameters, and physical state parameters.

[0316] In this embodiment of the invention, the preprocessing module 502 includes:

[0317] The outlier removal submodule is used to remove outliers from multidimensional data using the box plot method, resulting in outlier-removed data.

[0318] The missing value completion submodule is used to complete the missing values ​​of the outlier data based on the K-nearest neighbor difference method to obtain the completed data.

[0319] The standardization submodule is used to standardize the completed data to obtain multiple types of preprocessed data.

[0320] In this embodiment of the invention, the insulation resistance attenuation rate and partial discharge rate increase calculation module 503 includes:

[0321] The new cable initial insulation resistance and initial partial discharge quantity acquisition submodule is used to acquire the new cable initial insulation resistance and initial partial discharge quantity.

[0322] The current insulation resistance and current partial discharge quantity acquisition submodule is used to acquire the current insulation resistance and current partial discharge quantity of the cable;

[0323] The insulation resistance attenuation rate calculation submodule is used to calculate the ratio of the current insulation resistance to the initial insulation resistance of the new cable, and obtain the insulation resistance attenuation rate of the cable.

[0324] The partial discharge growth rate calculation submodule is used to calculate the ratio between the current partial discharge quantity and the initial partial discharge quantity to obtain the partial discharge growth rate of the cable.

[0325] In this embodiment of the invention, the key feature extraction module 505 includes:

[0326] The correlation calculation submodule is used to calculate the correlation between various types of preprocessed data and various types of aging coefficients based on the preprocessed data.

[0327] The key feature extraction submodule is used to identify preprocessed data with a correlation greater than a preset threshold as key features.

[0328] In this embodiment of the invention, the aging level generation module 507 includes:

[0329] The preset aging feature library acquisition submodule is used to acquire the preset aging feature library, which includes aging feature sets at multiple levels.

[0330] The similarity calculation submodule is used to calculate the similarity between the predicted feature value and the aging feature set at multiple levels, respectively.

[0331] The aging level determination submodule is used to determine the aging level of the cable based on the aging feature set with the highest similarity.

[0332] In this embodiment of the invention, the fault probability calculation module 508 includes:

[0333] The basic failure probability calculation submodule is used to obtain the basic failure probability based on the aging level mapping.

[0334] The attenuation coefficient determination submodule is used to obtain the cable material and determine the attenuation coefficient based on the material.

[0335] The decay failure probability calculation submodule is used to generate decay failure probability by taking the years of operation and decay coefficient and combining them with the exponential decay model.

[0336] The actual failure probability acquisition submodule is used to query a preset number of similar cases with the highest similarity to the predicted feature value from a preset historical failure database, as well as the actual failure probability of the similar cases.

[0337] The similar fault probability calculation submodule is used to calculate the similar fault probability of the cable based on the actual fault probability of all similar cases.

[0338] The fault probability calculation submodule is used to weight and fuse the basic fault probability, attenuation fault probability, and similar fault probability to obtain the fault probability of the cable.

[0339] This invention also provides an electronic device, which includes a processor and a memory:

[0340] The memory is used to store program code and transfer the program code to the processor;

[0341] The processor is used to execute the cable aging fault prediction method of this invention according to the instructions in the program code.

[0342] This invention also provides a computer-readable storage medium for storing program code for executing the cable aging fault prediction method of this invention.

[0343] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0344] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0345] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, embodiments of the present invention can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of the present invention can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0346] Embodiments of the present invention are described with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0347] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0348] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0349] Although preferred embodiments of the present invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present invention.

[0350] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0351] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0352] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. 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. Such 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 predicting cable aging faults, characterized in that, include: Collect multidimensional data and service life of the cable; The multidimensional data is preprocessed to obtain multiple types of preprocessed data; Calculate the insulation resistance attenuation rate and the partial discharge rate increase of the cable; The aging coefficient is calculated based on the insulation resistance attenuation rate and the partial discharge rate increase rate. Based on the preprocessed data and the aging coefficient, key features are extracted; The key features are used for prediction to obtain predicted feature values; The aging level of the cable is obtained by performing feature matching on the predicted feature values; The failure probability is calculated based on the aging level, the predicted characteristic value, and the years of operation. Fault classification and early warning are performed based on the aforementioned fault probability.

2. The method according to claim 1, characterized in that, The multidimensional data includes electrical parameters, environmental parameters, and physical state parameters.

3. The method according to claim 1, characterized in that, The step of preprocessing the multidimensional data to obtain multiple types of preprocessed data includes: Outliers in the multidimensional data are removed using the box plot method to obtain outlier-removed data; Based on the K-nearest neighbor difference method, missing values ​​are filled in the abnormal data to obtain the completed data; The completed data is standardized to obtain multiple types of preprocessed data.

4. The method according to claim 1, characterized in that, The steps for calculating the insulation resistance attenuation rate and the partial discharge rate increase of the cable include: Obtain the initial insulation resistance and initial partial discharge of the new cable; Obtain the current insulation resistance and current partial discharge level of the cable; Calculate the ratio of the current insulation resistance to the initial insulation resistance of the new cable to obtain the insulation resistance attenuation rate of the cable; The ratio between the current partial discharge quantity and the initial partial discharge quantity is calculated to obtain the partial discharge growth rate of the cable.

5. The method according to claim 1, characterized in that, The step of extracting key features based on the preprocessed data and the aging coefficient includes: Based on the preprocessed data, the correlation between each type of preprocessed data and each type of aging coefficient is calculated; Preprocessed data corresponding to types with a correlation greater than a preset threshold are identified as key features.

6. The method according to claim 1, characterized in that, The step of performing feature matching on the predicted feature values ​​to obtain the aging level of the cable includes: Obtain a preset aging feature library, which includes multiple levels of aging feature sets; Calculate the similarity between the predicted feature values ​​and the aging feature sets of multiple levels, respectively; The aging level of the cable is determined by the aging feature set with the highest similarity.

7. The method according to claim 1, characterized in that, The step of calculating the failure probability based on the aging level, the predicted characteristic value, and the years of operation includes: The basic failure probability is obtained by mapping the aging level. Obtain the material of the cable and determine the attenuation coefficient based on the material; Using the aforementioned operating years and the aforementioned attenuation coefficient, combined with an exponential attenuation model, the probability of attenuation failure is generated. Query a preset historical fault database for a preset number of similar cases that have the highest similarity to the predicted feature value and the actual fault probability corresponding to the similar cases; The similar failure probability of the cable is calculated based on the actual failure probability of all the similar cases. The cable's failure probability is obtained by weighted fusion of the basic failure probability, the attenuation failure probability, and the similarity failure probability.

8. A cable aging fault prediction device, characterized in that, include: The multi-dimensional data and service life acquisition module is used to collect multi-dimensional data and service life of the cable. The preprocessing module is used to preprocess the multidimensional data to obtain multiple types of preprocessed data; An insulation resistance attenuation rate and partial discharge growth rate calculation module is used to calculate the insulation resistance attenuation rate and partial discharge growth rate of the cable. An aging coefficient calculation module is used to calculate the aging coefficient based on the insulation resistance attenuation rate and the partial discharge growth rate. A key feature extraction module is used to extract key features based on the preprocessed data and the aging coefficient; A prediction feature value generation module is used to make predictions using the key features to obtain prediction feature values. An aging level generation module is used to perform feature matching on the predicted feature values ​​to obtain the aging level of the cable. The failure probability calculation module is used to calculate the failure probability based on the aging level, the predicted feature value, and the years of operation. The fault classification and early warning module is used to classify and warn of faults based on the fault probability.

9. An electronic device, characterized in that, The device includes a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the cable aging fault prediction method according to any one of claims 1-7 according to the instructions in the program code.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store program code for executing the cable aging fault prediction method according to any one of claims 1-7.

Images