A power distribution network cable joint aging prediction method and system

Abstract

The application relates to the technical field of cable joint aging prediction, and particularly provides a power distribution network cable joint aging prediction method and system. The method comprises the following steps: obtaining multi-modal operation data, joint type information and operation condition information of a cable joint to be predicted; performing aging feature extraction on operation data of different modes in the multi-modal operation data based on a preset aging feature extraction rule to obtain joint aging features corresponding to different modes, and then performing feature fusion on all the joint aging features to obtain fused aging features; determining target parameters of an aging degree prediction algorithm according to the joint type information and the operation condition information; and performing aging degree prediction according to the fused aging features by using the aging degree prediction algorithm to obtain an aging prediction result. The method can make the aging degree prediction algorithm have dynamic adaptability to real-time operation conditions and joint types.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cable joint aging prediction, in particular to a power distribution network cable joint aging prediction method and system. BACKGROUND

[0002] The cable joint is a key component of the power network, and its reliability directly affects the safety of the power grid and the stability of power supply. In long-term operation, the internal insulation material of the cable joint will gradually deteriorate, such as a decrease in partial discharge inception voltage, an increase in dielectric loss, and an increase in contact resistance due to oxidation of the metal contact surface. These performance deteriorations can cause partial discharge or overheating problems, ultimately leading to insulation breakdown or connection failure, and thus causing line faults and large-scale power outages.

[0003] In order to effectively manage the cable joint and avoid failure losses, the prior art needs to accurately predict the aging condition of the cable joint. The existing cable joint aging prediction method uses a parameter threshold judgment-based method to predict the aging condition of the cable joint, that is, the existing cable joint aging prediction method first collects the operation data of the cable joint, and then predicts the aging condition of the cable joint by analyzing whether the operation data exceeds the preset threshold or analyzing which preset range the operation data is in.

[0004] However, in actual application, the materials and manufacturing processes of different types of cable joints differ, that is, the aging mechanisms of different types of cable joints and their sensitivity to operating conditions and environmental stresses are different, and the operating conditions of the cable joint are also constantly changing (for example, seasonal temperature and humidity changes, load growth, or power grid topology adjustment). These dynamic changes affect the real-time aging process of the cable joint, that is, the aging of the cable joint is a nonlinear process affected by multiple factors. Since the existing cable joint aging prediction method uses a parameter threshold judgment-based method to predict the aging condition of the cable joint, that is, the existing cable joint aging prediction method lacks dynamic adaptability to real-time operating conditions and joint types, the prior art has the problem that the predicted aging condition cannot accurately reflect the actual aging condition of the cable joint due to the lack of dynamic adaptability of the cable joint aging prediction method to real-time operating conditions and joint types, thereby resulting in low accuracy and reliability of cable joint aging prediction.

[0005] There is currently no effective technical solution to the above problems. It should be noted that the above information disclosed in this section is only for understanding the background of the inventive concept, and therefore can contain information that does not constitute prior art. SUMMARY

[0006] The purpose of the present application is to provide a power distribution network cable joint aging prediction method and system, which can effectively solve the problem that the predicted aging condition cannot accurately reflect the actual aging condition of the cable joint due to the lack of dynamic adaptability of the cable joint aging prediction method to real-time operating conditions and joint types.

[0007] In a first aspect, the present application provides a power distribution network cable joint aging prediction method, comprising the following steps:

[0008] S1, obtaining multi-modal operating data, joint type information and operating condition information of a cable joint to be predicted;

[0009] S2, extracting aging features from operating data of different modalities in the multi-modal operating data based on a preset aging feature extraction rule to obtain joint aging features corresponding to different modalities, and then performing feature fusion on all joint aging features to obtain fused aging features;

[0010] S3, determining target parameters of the aging degree prediction algorithm according to the joint type information and the operating condition information;

[0011] S4, predicting the aging degree according to the fused aging features using the aging degree prediction algorithm to obtain an aging prediction result.

[0012] The power distribution network cable joint aging prediction method provided by the present application can dynamically determine the parameters of the aging degree prediction algorithm according to the joint type information and the operating condition information, so that the aging degree prediction algorithm can adapt to different types of cable joints and dynamically changing operating conditions, i.e., the present application can make the aging degree prediction algorithm have dynamic adaptability to real-time operating conditions and joint types. Therefore, the present application can effectively solve the problem that the predicted aging condition cannot accurately reflect the actual aging condition of the cable joint due to the lack of dynamic adaptability of the cable joint aging prediction method to real-time operating conditions and joint types, thereby effectively improving the accuracy and reliability of cable joint aging prediction.

[0013] In a second aspect, the present application also provides a power distribution network cable joint aging prediction system, comprising:

[0014] A data acquisition module for acquiring multi-modal operating data, joint type information and operating condition information of a cable joint to be predicted;

[0015] A data fusion module for extracting aging features from operating data of different modalities in the multi-modal operating data based on a preset aging feature extraction rule to obtain joint aging features corresponding to different modalities, and then performing feature fusion on all joint aging features to obtain fused aging features;

[0016] The prediction algorithm parameter determination module is configured to determine target parameters of the aging degree prediction algorithm according to the joint type information and the operation condition information.

[0017] The aging degree prediction module is configured to perform aging degree prediction according to the fused aging features by using the aging degree prediction algorithm to obtain an aging prediction result.

[0018] The power distribution network cable joint aging prediction system provided by the application can make the aging degree prediction algorithm adapt to different types of cable joints and dynamically changing operation conditions by dynamically determining the parameters of the aging degree prediction algorithm according to the joint type information and the operation condition information, that is, the application can make the aging degree prediction algorithm have dynamic adaptability to real-time operation conditions and joint types, and therefore the application can effectively solve the problem that the predicted aging condition cannot accurately reflect the actual aging condition of the cable joint due to the lack of dynamic adaptability of the cable joint aging prediction method to real-time operation conditions and joint types, thereby effectively improving the accuracy and reliability of the cable joint aging prediction.

[0019] As can be seen from the above, the power distribution network cable joint aging prediction method and system provided by the application can make the aging degree prediction algorithm adapt to different types of cable joints and dynamically changing operation conditions by dynamically determining the parameters of the aging degree prediction algorithm according to the joint type information and the operation condition information, that is, the application can make the aging degree prediction algorithm have dynamic adaptability to real-time operation conditions and joint types, and therefore the application can effectively solve the problem that the predicted aging condition cannot accurately reflect the actual aging condition of the cable joint due to the lack of dynamic adaptability of the cable joint aging prediction method to real-time operation conditions and joint types, thereby effectively improving the accuracy and reliability of the cable joint aging prediction. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 The flowchart of the power distribution network cable joint aging prediction method provided by the embodiment of the application.

[0021] Figure 2 The structural schematic diagram of the power distribution network cable joint aging prediction system provided by the embodiment of the application.

[0022] Reference signs: 1, data acquisition module; 2, data fusion module; 3, prediction algorithm parameter determination module; 4, aging degree prediction module. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0024] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. Meanwhile, in the description of the present application, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0025] In a first aspect, as shown in the drawings, the present application provides a power distribution network cable joint aging prediction method, comprising the following steps: Figure 1

[0026] S1, obtaining multi-modal operation data, joint type information and operating condition information of a cable joint to be predicted;

[0027] S2, extracting aging features from the operation data of different modes in the multi-modal operation data based on a preset aging feature extraction rule to obtain joint aging features corresponding to different modes, and then performing feature fusion on all joint aging features to obtain fused aging features;

[0028] S3, determining target parameters of an aging degree prediction algorithm according to the joint type information and the operating condition information;

[0029] S4, predicting the aging degree according to the fused aging features by using the aging degree prediction algorithm to obtain an aging prediction result.

[0030] ​The cable joint to be predicted refers to the cable joint that needs to be predicted for the degree of aging. The multi-modal operation data refers to the operation indicators of the cable joint from different sensors. The multi-modal operation data of this embodiment preferably includes temperature data collected by a temperature sensor, current data collected by a current monitor, voltage data collected by a voltage recorder, and vibration data collected by an accelerometer. The multi-modal operation data can comprehensively reflect the operation state and potential aging signs of the cable joint from multiple dimensions. The joint type information refers to the type of the cable joint. In this embodiment, the joint type information can be obtained by equipment file query, database query, nameplate reading, or identifying the label recording the type of the cable joint on the cable joint. Since the materials and manufacturing processes of different types of cable joints differ, and the aging mechanisms and sensitivities to operating conditions and environmental stresses of different types of cable joints are different, this embodiment can understand the inherent characteristics, aging mechanisms of the cable joint to be predicted, and the cable joint to be predicted by obtaining the joint type information. The operating condition information refers to the external conditions in which the cable joint is located during operation. In this embodiment, the operating condition information can be obtained by using sensors to collect data on the external conditions in which the cable joint is located or using a power grid monitoring system to monitor the external conditions in which the cable joint is located. The operating condition information at least includes operating load and environmental parameters. The operating condition information can also include the pollution level of the environment in which the cable joint is located. The operating condition information can reflect dynamic external factors that affect the aging process of the cable joint. The preset aging feature extraction rule refers to a predefined rule set for identifying and quantifying specific patterns or indicators related to the aging of the cable joint from raw multi-modal operation data. The preset aging feature extraction rule can be a signal processing-based method, a statistical analysis method, or a machine learning model. For example, the preset aging feature extraction rule can extract the discharge amount, discharge frequency, and phase distribution of partial discharge based on the current data collected by the current monitor and the voltage data collected by the voltage recorder, and extract the rate of change, maximum value, and volatility of the temperature curve based on the temperature data collected by the temperature sensor. This embodiment can use the preset aging feature extraction rule to convert raw and heterogeneous multi-modal data into structured features that can represent the aging state. Feature fusion refers to the process of integrating aging features extracted from different modal data to form a unified feature representation. This embodiment can use feature splicing, weighted average based on weights, or deep learning network fusion to realize feature fusion. For example, the temperature feature vector and the partial discharge feature vector can be directly spliced, or different weights can be assigned according to the importance of the features for linear combination. This embodiment can use feature fusion of all joint aging features to comprehensively utilize aging information from different data sources to form a more comprehensive and robust fusion feature vector.The target parameter of the aging degree prediction algorithm refers to a parameter used to configure or adjust the aging prediction model to adapt to a specific prediction task. The aging degree prediction algorithm can be an existing aging degree prediction algorithm (such as a finite element analysis algorithm and an Arrhenius equation) or an aging prediction model (such as a support vector machine, a neural network, or a decision tree). It should be understood that this embodiment can obtain an aging prediction model by first training a pre-constructed deep learning model based on a gradient descent algorithm using a pre-labeled training data set (including multiple sets of training aging features and their corresponding training aging degree scores), then optimizing the hyperparameters of the deep learning model using a validation set (including validation aging features and their corresponding validation aging degree scores), and finally testing the performance of the deep learning model using a test set (including test aging features and their corresponding test aging degree scores). The target parameter can be any one or more of the parameters that affect the output results of the aging prediction model, such as the weights, hyperparameters, and rule thresholds of the aging prediction model. Since this embodiment determines the target parameter of the aging degree prediction algorithm according to the joint type information and the operating condition information, i.e., this embodiment is equivalent to configuring the aging degree prediction algorithm individually according to the joint type and the operating condition, this embodiment can make the aging degree prediction algorithm have dynamic adaptability to real-time operating conditions and joint types, thereby effectively improving the adaptability and relevance of the aging degree prediction algorithm.

[0031] The core innovation of the present application is that by dynamically determining the parameters of the aging degree prediction algorithm according to the joint type information and the operating condition information, the aging degree prediction algorithm can adapt to different types of cable joints and dynamically changing operating conditions, i.e., the present application can make the aging degree prediction algorithm have dynamic adaptability to real-time operating conditions and joint types. Therefore, the present application can effectively solve the problem that the predicted aging condition cannot accurately reflect the actual aging condition of the cable joint due to the lack of dynamic adaptability of the cable joint aging prediction method to real-time operating conditions and joint types, thereby effectively improving the accuracy and reliability of the cable joint aging prediction.

[0032] Specifically, the method first acquires multi-modal operation data of a cable joint to be predicted, joint type information and operation condition information, which are the basis for accurate prediction. Then, based on a preset aging feature extraction rule, features are extracted from the data of different modalities in the acquired multi-modal operation data to obtain joint aging features corresponding to different modalities. Since these joint aging features are obtained by abstracting and refining the original multi-modal operation data, they can better reflect the aging state than the original multi-modal operation data. Next, all the extracted joint aging features are fused to form a comprehensive fusion aging feature, which contains aging information from multiple data sources. Subsequently, according to the acquired joint type information and operation condition information, key parameters (target parameters) of the aging degree prediction algorithm for subsequent prediction are dynamically determined, so that the aging degree prediction algorithm can be dynamically adjusted according to the specific joint characteristics and operation conditions. Finally, the fusion aging feature obtained above is input into the aging assessment algorithm for determining the target parameters, so that the aging degree prediction algorithm predicts the aging degree according to the fusion aging feature and outputs the final aging prediction result. The whole process forms a closed loop from multi-source data input to dynamic adjustment of the aging degree prediction algorithm and finally to prediction output, ensuring that the prediction result can more accurately reflect the actual aging state of the cable joint.

[0033] As a preferred embodiment, the scheme of the present application is implemented as follows: First, multi-modal operation data (temperature, partial discharge signal and dielectric loss data of the cable joint), operation condition information (operation load and environmental parameters such as temperature and humidity) and joint type information are acquired through sensors installed near the cable joint, remote monitoring systems and checking equipment records. Then, using signal processing techniques (preset aging feature extraction rules), the temperature change rate and maximum temperature features are extracted from the temperature data, the discharge amount and discharge frequency features are extracted from the partial discharge signal, and the dielectric loss tangent value features are extracted from the dielectric loss data. Then, these extracted temperature features, partial discharge features and dielectric loss features are spliced to form a fusion feature vector. Subsequently, according to the acquired joint type information and the current current load, environmental temperature and humidity, a pre-established mapping table is queried or a rule engine is used to determine the learning rate and regularization coefficient of the neural network model for prediction (equivalent to determining the target parameters of the aging degree prediction algorithm). Finally, the fusion feature vector is input into the neural network model with the determined parameters, so that the model outputs a numerical value or level representing the aging degree of the cable joint (aging prediction result).

[0034] The power distribution network cable joint aging prediction method provided by the application can dynamically determine the parameters of the aging degree prediction algorithm according to the joint type information and the operation condition information, so that the aging degree prediction algorithm can adapt to different types of cable joints and dynamically changing operation conditions, that is, the application can make the aging degree prediction algorithm have dynamic adaptability to real-time operation conditions and joint types. Therefore, the application can effectively solve the problem that the predicted aging condition cannot accurately reflect the actual aging condition of the cable joint due to the lack of dynamic adaptability of the cable joint aging prediction method to real-time operation conditions and joint types, thereby effectively improving the accuracy and reliability of the cable joint aging prediction.

[0035] In some preferred embodiments, step S3 comprises:

[0036] S31, querying a mapping relationship table about joint types, operation conditions and aging degree prediction algorithm parameters pre-constructed according to the joint type information and the operation condition information to obtain preliminary parameters of the aging degree prediction algorithm;

[0037] S32, obtaining the cumulative maintenance number and the historical maintenance content set of the cable joint to be predicted;

[0038] S33, obtaining a first prediction algorithm parameter adjustment strategy according to the cumulative maintenance number and the historical maintenance content set;

[0039] S34, adjusting the preliminary parameters according to the first prediction algorithm parameter adjustment strategy to determine the target parameters of the aging degree prediction algorithm.

[0040] The mapping relationship table of joint type, operating condition and aging prediction algorithm parameter refers to a data structure that stores the corresponding relationship between different joint types, different operating conditions and preliminary parameters of the aging prediction algorithm. The preliminary parameters of the aging degree prediction algorithm refer to the initial values of the parameters of the aging degree prediction algorithm obtained by querying the mapping relationship table based on the joint type and the operating condition. The cumulative maintenance number refers to the total number of times of maintenance of the cable joint to be predicted since it was put into operation. The historical maintenance content set refers to a set of specific information of each maintenance (such as maintenance date, maintenance type, problems found, measures taken, etc.) since the cable joint to be predicted was put into operation. The first prediction algorithm parameter adjustment strategy refers to a rule or method for guiding how to correct the preliminary parameters to make them more consistent with the actual situation of the joint, which is determined according to the cumulative maintenance number and the historical maintenance content set of the cable joint. For example, if the historical maintenance content set shows that the joint has been maintained several times due to overheating, the first prediction algorithm parameter adjustment strategy indicates to increase the weight or acceleration factor of the temperature-related aging parameter. This embodiment can obtain the first prediction algorithm parameter adjustment strategy by querying the pre-constructed mapping relationship table of maintenance number, maintenance content combination and prediction algorithm parameter adjustment strategy according to the cumulative maintenance number and the historical maintenance content set. This embodiment can also obtain the first prediction algorithm parameter adjustment strategy by inputting the cumulative maintenance number and the historical maintenance content set into the pre-trained algorithm parameter adjustment strategy development model based on maintenance records. Since the difference in maintenance conditions will affect the real aging state and future aging trend of the cable joint under the same other conditions, this embodiment can make the parameters of the aging prediction algorithm avoid the situation that the parameters of the aging prediction algorithm are not accurate and reliable enough due to insufficient consideration of the influence of historical maintenance conditions on the aging process and the selection of prediction algorithm parameters by integrating the maintenance history information into the parameter determination process. Therefore, this embodiment can effectively improve the accuracy and reliability of the parameter determination of the aging prediction algorithm, so that the parameters of the aging prediction algorithm can more accurately reflect the individual characteristics and actual aging process of the joint to be predicted, thereby further improving the accuracy and reliability of the cable joint aging prediction.

[0041] In some preferred embodiments, step S34 comprises:

[0042] S341, obtaining quality detection data and installation process information of the cable joint to be predicted;

[0043] S342, obtaining a second prediction algorithm parameter adjustment strategy according to the quality detection data and the installation process information;

[0044] S343, adjusting the preliminary parameters according to the first prediction algorithm parameter adjustment strategy and the second prediction algorithm parameter adjustment strategy to determine the target parameters of the aging degree prediction algorithm.

[0045] The quality detection data refers to data reflecting inherent performance and quality of the cable joint to be predicted at the time of manufacturing or leaving factory. In this embodiment, the quality detection data of the cable joint to be predicted can be obtained based on a factory test report, a material composition analysis report, an insulation withstand voltage test result, and a partial discharge test result. The installation process information refers to data recording relevant details and specifications of the cable joint in the field installation process. In this embodiment, the installation process information of the cable joint to be predicted can be obtained by integrating information from an installation record table, field photos, installation personnel qualifications, environmental condition records, and inspection results of key operations (such as stripping size, crimping force, and sealing treatment). The second prediction algorithm parameter adjustment strategy refers to rules or methods for guiding how to modify the preliminary parameters of the aging degree prediction algorithm based on the quality detection data and the installation process information. For example, when the quality detection data reflects that the insulation resistance is lower than a certain threshold or the partial discharge inception voltage is too low, the second prediction algorithm parameter adjustment strategy is to increase a parameter related to the insulation resistance or the discharge voltage by a certain proportion in the aging rate. For another example, when the installation process information reflects that the installation personnel qualifications are not in conformity or the sealing inspection result is unqualified, the second prediction algorithm parameter adjustment strategy is to increase a parameter related to moisture or partial discharge by a certain proportion. In this embodiment, the second prediction algorithm parameter adjustment strategy can be obtained by querying a mapping relationship table about the quality detection data, the installation process, and the prediction algorithm parameter adjustment strategy in advance according to the quality detection data and the installation process information. The second prediction algorithm parameter adjustment strategy can also be obtained by inputting the quality detection data and the installation process information into a pre-trained algorithm parameter adjustment strategy development model based on the quality detection data and the installation process. Since the initial quality and the installation process of the cable joint are also key factors affecting its long-term performance and aging speed, this embodiment can adjust the parameters of the aging degree prediction algorithm based on the initial quality and the installation process of the cable joint by first obtaining the second prediction algorithm parameter adjustment strategy according to the quality detection data and the installation process information, and then adjusting the preliminary parameters based on the second prediction algorithm parameter adjustment strategy. Therefore, this embodiment can make the aging degree prediction algorithm better adapt to cable joints with different initial qualities and installation conditions, thereby further improving the accuracy and reliability of the cable joint aging prediction.

[0046] In some preferred embodiments, step S343 comprises:

[0047] A1, obtaining historical operating conditions of the cable joint to be predicted;

[0048] A2, obtaining a third prediction algorithm adjustment strategy according to the historical operating conditions;

[0049] A3, adjusting the preliminary parameters according to the first, second and third prediction algorithm parameter adjustment strategies to determine the target parameters of the aging degree prediction algorithm.

[0050] The historical operation condition refers to a record of the operation condition experienced by the cable joint in the past time. In this embodiment, the historical operation condition of the cable joint to be predicted can be obtained by obtaining a data set such as a historical load curve, an environmental temperature record, a humidity record, or a voltage fluctuation record. The historical operation condition can reflect the actual stress accumulation of the cable joint to be predicted in different historical operation stages. The third prediction algorithm adjustment strategy refers to a rule or function obtained according to the historical operation condition for guiding how to adjust the preliminary parameters of the aging degree prediction algorithm. For example, when the cumulative stress is high, the third prediction algorithm adjustment strategy indicates to increase the weight of the parameter related to thermal aging. In this embodiment, the third prediction algorithm parameter adjustment strategy can be obtained by querying a pre-constructed mapping relationship table about the historical operation condition and the prediction algorithm parameter adjustment strategy according to the historical operation condition. The third prediction algorithm parameter adjustment strategy can also be obtained by inputting the historical operation condition into a pre-trained algorithm parameter adjustment strategy development model based on the historical operation condition. Since this embodiment first obtains the third prediction algorithm adjustment strategy according to the historical operation condition, and then adjusts the preliminary parameters based on the third prediction algorithm parameter adjustment strategy, this embodiment can effectively avoid the situation that the aging degree prediction algorithm cannot adapt to the dynamic changes of the cable joint stress accumulation due to not considering the actual stress accumulation of the cable joint to be predicted in different historical operation stages when determining the target parameters of the aging degree prediction algorithm, thereby further improving the accuracy and reliability of the aging degree prediction algorithm parameter determination, and further improving the accuracy and reliability of the cable joint aging prediction.

[0051] In some preferred embodiments, step S2 comprises:

[0052] S21, obtaining the cumulative operation time and historical operation data of the cable joint to be predicted;

[0053] S22, analyzing and obtaining the aging stage information of the cable joint to be predicted according to the cumulative operation time and the historical operation data;

[0054] S23, querying a pre-constructed mapping relationship table about the aging stage and the aging feature extraction rule according to the aging stage information to obtain a preset aging feature extraction rule;

[0055] S24, extracting aging features from the operation data of different modalities in the multi-modal operation data based on the preset aging feature extraction rule to obtain joint aging features corresponding to different modalities;

[0056] S25, performing feature fusion on all joint aging features to obtain fused aging features.

[0057] The cumulative running length refers to the total running time of the to-be-predicted cable joint from the start of operation to the current time, which can reflect the total working time accumulation experienced by the to-be-predicted cable joint. The historical running data refers to a set of various running state parameters (such as temperature data, current data, and voltage data) collected by the to-be-predicted cable joint in the past time, which contains the running state information of the to-be-predicted cable joint in different time periods, which can be stored in the form of time series data, statistical summary data, etc. The aging stage information refers to the aging state stage that the to-be-predicted cable joint may currently be in, which is analyzed and judged according to the cumulative running length and the historical running data of the cable joint. The aging stage information can be represented by discrete stage identifiers (for example, initial, middle, and late) or continuous aging degree indicators. In this embodiment, the aging stage information can be obtained by querying a pre-constructed mapping relationship table about running length, historical running data, and aging stage according to the cumulative running length and the historical running data. The aging stage information can also be obtained by inputting the cumulative running length and the historical running data into a pre-trained aging stage evaluation model. The mapping relationship table about the aging stage and the aging feature extraction rule refers to a data structure pre-established for associating different aging stages with the feature extraction rules applicable to the stage. The preset aging feature extraction rule refers to a specific rule set obtained from the mapping relationship table according to the current aging stage information of the cable joint to guide the feature extraction process. The preset aging feature extraction rule can include specific algorithms, parameter settings, or feature selection standards for different modal data. For example, for the initial aging stage, the preset aging feature extraction rule focuses on the extraction of features such as partial discharge inception voltage and dielectric loss. For the late aging stage, the preset aging feature extraction rule focuses on the extraction of features such as temperature change rate and contact resistance. The aging mechanism and key features of cable joints in different aging stages may be different. Using a fixed aging feature extraction rule may not accurately capture the features that best reflect the aging degree of the current joint in the current aging stage. Since this embodiment can dynamically adjust the feature extraction rule according to the aging stage by first analyzing the aging stage information of the to-be-predicted cable joint according to the cumulative running length and the historical running data, and then determining the preset aging feature extraction rule according to the aging stage information, this embodiment can accurately capture the features that best reflect the aging degree of the current joint in the current aging stage, thereby effectively improving the accuracy and reliability of the aging feature extraction, and further improving the accuracy and reliability of the cable joint aging prediction.

[0058] In some preferred embodiments, step S22 comprises:

[0059] S221, query a pre-constructed mapping relationship table about aging stages and aging feature extraction rules according to the aging stage information to obtain a preliminary aging feature extraction rule;

[0060] S222, query a pre-constructed mapping relationship table about joint types and feature extraction rule adjustment strategies according to the joint type information to obtain an aging feature extraction rule adjustment strategy;

[0061] S223, adjust the preliminary aging feature extraction rule according to the aging feature extraction rule adjustment strategy to obtain a preset aging feature extraction rule.

[0062] The mapping relationship table about joint types and feature extraction rule adjustment strategies is a data structure that stores the feature extraction rule adjustment strategies corresponding to different joint types. The aging feature extraction rule adjustment strategy is an instruction or parameter obtained from the mapping relationship table based on the joint type information, which is used to modify or improve the preliminary aging feature extraction rule. The aging feature extraction rule adjustment strategy can include adding a specific type of feature, adjusting the weight of certain features, or modifying the feature calculation method. For example, when the joint type information is thermal shrinkage type (the predicted cable joint is more sensitive to temperature changes), the aging feature extraction rule adjustment strategy is to increase the weight of temperature-related features and extract temperature gradient features. This embodiment can make the extracted structural aging features more accurately reflect the actual aging of a specific type of cable joint at a specific aging stage by adjusting the aging feature extraction rule based on the joint type of the predicted cable joint, thereby further improving the accuracy and reliability of the aging feature extraction, and further improving the accuracy and reliability of the cable joint aging prediction.

[0063] In some preferred embodiments, step S25 comprises:

[0064] S251, respectively evaluate the quality of each joint aging feature to obtain a feature quality score corresponding to each joint aging feature;

[0065] S252, determine the fusion weight corresponding to each joint aging feature according to all feature quality scores;

[0066] S253, fuse all joint aging features and their corresponding fusion weights to obtain a fused aging feature.

[0067] The feature quality evaluation refers to a process of measuring the reliability, information amount or correlation with the aging degree of the extracted joint aging features. The embodiment can realize feature quality evaluation of each joint aging feature respectively by using statistical methods (for example, calculating the variance and signal-to-noise ratio of the features), rules based on expert knowledge, or a feature quality evaluation learning model (for example, evaluating the correlation of the features with known aging indicators). The feature quality score refers to a numerical representation of the quality evaluation result of each joint aging feature. The feature quality score can reflect the reliability of the corresponding joint aging feature. The fusion weight refers to a coefficient assigned to each joint aging feature in the feature fusion process. The fusion weight determines the contribution of the corresponding joint aging feature to the final fusion result. Specifically, the higher the quality score of the joint aging feature, the greater the corresponding fusion weight. The embodiment can determine the fusion weight corresponding to each joint aging feature by querying a pre-constructed feature quality score combination and fusion weight combination according to all feature quality scores. The embodiment can also determine the fusion weight corresponding to each joint aging feature by normalizing all feature quality scores. Since the embodiment can distinguish the quality and importance of different joint aging features, and the embodiment can highlight the role of high-quality features and suppress the negative impact of low-quality features by weighting and fusing all joint aging features based on all fusion weights, the embodiment can obtain more effective and reliable fused aging features, thereby effectively improving the accuracy and reliability of the fused aging features, and further improving the accuracy and reliability of the cable joint aging prediction.

[0068] In some preferred embodiments, step S1 comprises:

[0069] S11, acquiring multi-modal operation data, joint type information and operation condition information of the cable joint to be predicted;

[0070] S12, pre-processing the multi-modal operation data, which includes data cleaning, data synchronization and format conversion.

[0071] The preprocessing refers to a preparation process of the original multi-modal operation data before data analysis, and the preprocessing includes data cleaning, data synchronization and format conversion. The data cleaning refers to identifying and processing errors, noises, outliers or missing values in the data, which can be achieved by deleting, correcting, filling and the like. The data synchronization refers to aligning data of different sources or different collection frequencies in the time dimension, which can be achieved by time stamp alignment, resampling, interpolation and the like. The format conversion refers to uniformly converting data of different formats into one or several standard formats, which can be achieved by data format parsing, data structure conversion and the like. The embodiment can improve the data quality and availability of the multi-modal operation data by preprocessing the multi-modal operation data, thereby laying a foundation for subsequent feature extraction and aging prediction based on the multi-modal data, and thus improving the accuracy and reliability of the entire aging prediction method.

[0072] In some preferred embodiments, the operation condition information includes operation load and environmental parameters. The operation load refers to electrical parameters such as active power and voltage, which reflects the current electrical stress level and working strength of the cable joint to be predicted. The environmental parameters refer to external environmental conditions of the cable joint, such as environmental temperature, humidity and the like, which can reflect the current environmental stress level of the cable joint to be predicted. The embodiment can measure the environmental parameters by using temperature sensors, humidity sensors and the like.

[0073] As can be seen from the above, the power distribution network cable joint aging prediction method provided by the application can dynamically determine the parameters of the aging degree prediction algorithm according to the joint type information and the operation condition information, so that the aging degree prediction algorithm can adapt to different types of cable joints and dynamically changing operation conditions, i.e., the application can make the aging degree prediction algorithm have dynamic adaptability to real-time operation conditions and joint types. Therefore, the application can effectively solve the problem that the predicted aging condition cannot accurately reflect the actual aging condition of the cable joint due to the lack of dynamic adaptability of the cable joint aging prediction method to real-time operation conditions and joint types, thereby effectively improving the accuracy and reliability of the cable joint aging prediction.

[0074] In a second aspect, as shown in Figure 2 The application further provides a power distribution network cable joint aging prediction system, which comprises:

[0075] A data acquisition module 1 is configured to acquire multi-modal operation data, joint type information and operation condition information of a cable joint to be predicted.

[0076] The data fusion module 2 is configured to extract the aging features of the operation data of different modes in the multi-mode operation data based on preset aging feature extraction rules, to obtain the joint aging features corresponding to different modes, and then to perform feature fusion on all joint aging features to obtain the fused aging features.

[0077] The prediction algorithm parameter determination module 3 is configured to determine the target parameters of the aging degree prediction algorithm according to the joint type information and the operation condition information.

[0078] The aging degree prediction module 4 is configured to perform aging degree prediction according to the fused aging features by using the aging degree prediction algorithm, to obtain the aging prediction result.

[0079] The power distribution network cable joint aging prediction system provided in the embodiment includes the data acquisition module 1, the data fusion module 2, the prediction algorithm parameter determination module 3 and the aging degree prediction module 4. The power distribution network cable joint aging prediction system provided in the embodiment is used to perform the steps in the power distribution network cable joint aging prediction method provided in the first aspect. The principle of the power distribution network cable joint aging prediction system provided in the embodiment is the same as that of the power distribution network cable joint aging prediction method provided in the first aspect, and will not be discussed in detail here.

[0080] As can be seen from the above, the power distribution network cable joint aging prediction method and system provided in the application can make the aging degree prediction algorithm adapt to different types of cable joints and dynamically changing operation conditions by dynamically determining the parameters of the aging degree prediction algorithm according to the joint type information and the operation condition information, that is, the application can make the aging degree prediction algorithm have dynamic adaptability to real-time operation conditions and joint types. Therefore, the application can effectively solve the problem that the predicted aging condition cannot accurately reflect the actual aging condition of the cable joint due to the lack of dynamic adaptability of the cable joint aging prediction method to real-time operation conditions and joint types, thereby effectively improving the accuracy and reliability of the cable joint aging prediction.

[0081] In the embodiments provided in the application, it should be understood that the disclosed devices and methods can be implemented in other manners. The above-described device embodiments are only schematic. For example, the division of the units is only a logical function division. There can be another division manner for the actual implementation. For example, a plurality of units or components can be combined or integrated into another device, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical, mechanical or other forms.

[0082] In addition, each functional module in each embodiment of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0083] In this document, relational terms such as first and second and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.

[0084] The above description is merely illustrative of the application, and not in limitation of the principles of the application. Any modification and change, which can be made by those skilled in the art, within the spirit and principle of the application, should be included in the scope of the application.

Claims

1. A method for predicting aging of cable joints in a distribution network, characterized in that: The distribution network cable joint aging prediction method comprises the following steps: S1. Obtain multimodal operation data, joint type information, and operation condition information of the cable joint to be predicted; S2. Extracting aging features from the operation data of different modes in the multimodal operation data based on a preset aging feature extraction rule to obtain joint aging features corresponding to the different modes, and then fusing all the joint aging features to obtain a fused aging feature. S3, according to the joint type information and the operating condition information to determine the target parameters of the aging degree prediction algorithm; S4. Using the aging degree prediction algorithm to predict the aging degree according to the fused aging features to obtain an aging prediction result; Step S3 includes: S31, querying a pre-built mapping relationship table of joint type, operating condition, and aging degree prediction algorithm parameters according to the joint type information and the operating condition information to obtain preliminary parameters of the aging degree prediction algorithm; S32. Obtaining the cumulative maintenance times and historical maintenance content set of the cable joint to be predicted; S33, obtaining a first prediction algorithm parameter adjustment strategy according to the accumulated maintenance times and the historical maintenance content set; S34, adjusting the preliminary parameters according to the first prediction algorithm parameter adjustment strategy to determine target parameters of the aging degree prediction algorithm; Step S34 includes: S341, obtaining quality inspection data and installation process information of the cable joint to be predicted; S342. Obtain a second prediction algorithm parameter adjustment strategy based on the quality inspection data and the installation process information; S343: Adjust the preliminary parameters according to the first prediction algorithm parameter adjustment strategy and the second prediction algorithm parameter adjustment strategy to determine target parameters of the aging degree prediction algorithm.

2. The method for predicting aging of cable joints in a distribution network according to claim 1, characterized in that: Step S343 includes: A1. Obtaining the historical operating conditions of the cable joint to be predicted; A2. Obtaining a third prediction algorithm parameter adjustment strategy based on the historical operating conditions; A3. Adjust the preliminary parameters according to the first prediction algorithm parameter adjustment strategy, the second prediction algorithm parameter adjustment strategy, and the third prediction algorithm parameter adjustment strategy to determine the target parameters of the aging degree prediction algorithm.

3. The method for predicting aging of cable joints in distribution network according to claim 1, characterized in that: Step S2 includes: S21, obtaining the cumulative operating time and historical operating data of the cable joint to be predicted; S22. Analyzing and obtaining aging stage information of the cable joint to be predicted based on the accumulated operating time and historical operating data; S23. Querying a pre-built mapping relationship table between aging stages and aging feature extraction rules according to the aging stage information to obtain preset aging feature extraction rules; S24, performing aging feature extraction on the operation data of different modes in the multimodal operation data based on the preset aging feature extraction rule to obtain joint aging features corresponding to different modes; S25. Perform feature fusion on all the joint aging features to obtain a fused aging feature.

4. The method for predicting aging of cable joints in a distribution network according to claim 3, characterized in that: Step S22 includes: S221, querying a pre-built mapping relationship table between aging stages and aging feature extraction rules according to the aging stage information to obtain preliminary aging feature extraction rules; S222: querying a pre-built mapping relationship table between connector types and feature extraction rule adjustment strategies based on the connector type information to obtain an aging feature extraction rule adjustment strategy; S223 : Adjust the preliminary aging feature extraction rule according to the aging feature extraction rule adjustment strategy to obtain a preset aging feature extraction rule.

5. The method for predicting aging of cable joints in distribution network according to claim 3, characterized in that: Step S25 includes: S251, performing feature quality assessment on each of the joint aging features to obtain a feature quality score corresponding to each of the joint aging features; S252, determining a fusion weight corresponding to each of the joint aging features according to all the feature quality scores; S253. Perform feature fusion according to all the joint aging features and their corresponding fusion weights to obtain a fused aging feature.

6. The method for predicting aging of cable joints in a distribution network according to claim 1, characterized in that: Step S1 includes: S11, obtaining multimodal operation data, joint type information, and operation condition information of the cable joint to be predicted; S12. Preprocess the multimodal operation data, where the preprocessing includes data cleaning, data synchronization, and format conversion.

7. The method for predicting aging of cable joints in a distribution network according to claim 1, characterized in that: The operating condition information includes operating load and environmental parameters.

8. A distribution network cable joint aging prediction system, characterized in that: The distribution network cable joint aging prediction system is used to perform the steps in the distribution network cable joint aging prediction method according to any one of claims 1 to 7, and the distribution network cable joint aging prediction system includes: A data acquisition module is used to obtain multimodal operating data, joint type information and operating condition information of the cable joint to be predicted; a data fusion module, configured to extract aging features from the operation data of different modes in the multimodal operation data based on preset aging feature extraction rules to obtain joint aging features corresponding to different modes, and then perform feature fusion on all the joint aging features to obtain a fused aging feature; A prediction algorithm parameter determination module for determining target parameters of an aging degree prediction algorithm based on the joint type information and the operating condition information; The aging degree prediction module is used to use the aging degree prediction algorithm to perform aging degree prediction according to the fused aging features to obtain an aging prediction result.

Images