Online monitoring method for cable insulation layer

Abstract

The invention relates to the related technical field of power cables, in particular to a cable insulation layer on-line monitoring method which comprises the following steps: a cable monitoring system carries out harmonic monitoring through a plurality of monitoring circuits, constructs an edge topology network to analyze circuit load, collects harmonic component characteristics, and carries out on-line monitoring on a cable insulation layer; the cloud center node trains a load harmonic processing model and issues the load harmonic processing model to the edge nodes to process harmonic signals, cable insulation defects are evaluated, indexes are output, and the technical problems that harmonic interference caused by cable circuit loads is propagated through a cable, adverse effects are generated on an insulation layer, and the precision of cable insulation defect point positioning is limited are solved. An external monitoring circuit is arranged at a cable interface, the circuit load of each edge node is meticulously analyzed by adopting a node division strategy, harmonic waves caused by non-insulating layer defects are effectively eliminated, multi-scale feature extraction is carried out for harmonic wave changes related to the insulating defects, the influence of load changes on harmonic wave signals is carefully captured, and the harmonic wave detection accuracy is improved. And the identification capability of the defects of the insulating layer is enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field related to power cables, and in particular to an online monitoring method for a cable insulation layer. Background Art

[0002] With the continuous expansion and complexity of the power system, cables are an important part of power transmission, and the safe and stable operation of cables is of vital importance. The aging and damage of cable insulation layer is one of the main causes of power system failures. If not discovered and handled in time, it may cause power outages.

[0003] Conventional cable insulation detection methods, such as judging the operating status of the cable insulation layer by the changes in the current and voltage flowing through the cable, are often used. However, in the case of long-distance cables, due to the complexity of the cable lines and the uncertainty of interference factors, the changes in the current and voltage flowing through the cable are interfered with, and the detection accuracy is severely limited.

[0004] In summary, in the prior art, there is a technical problem that harmonic interference caused by cable circuit load is transmitted through the cable, which has an adverse effect on the insulation layer and limits the accuracy of locating cable insulation defect points. Summary of the invention

[0005] The present application provides an online monitoring method for cable insulation layer, aiming to solve the technical problem in the prior art that harmonic interference caused by cable circuit load is transmitted through the cable, has an adverse effect on the insulation layer, and limits the accuracy of locating cable insulation defect points.

[0006] In view of the above problems, the technical solution to implement this application is: The present application provides a method for online monitoring of cable insulation layer, wherein the method comprises: obtaining a plurality of monitoring circuits connected to a cable monitoring system, each monitoring circuit being used for harmonic monitoring of the cable and outputting a plurality of harmonic signals; Using the multiple monitoring circuits as edge nodes and the cable monitoring system as a cloud center node to generate an edge topology network; Using multiple edge nodes in the edge topology network to perform load analysis on the multiple monitoring circuits, and obtain multiple circuit load information, including the number of circuit elements and the circuit load sizes corresponding to the multiple circuit elements; Collect harmonic component characteristics according to the multiple circuit load information, and output load-harmonic signal characteristics; The cloud center node in the edge topology network performs federated training on the load-harmonic signal features and outputs a load harmonic processing model; Downloading the load harmonic processing model to multiple edge nodes, and performing harmonic processing on the multiple harmonic signals using the load harmonic processing model to obtain processed harmonic signals; The processed harmonic signal is used to evaluate the cable insulation defects and output the cable insulation defect evaluation index.

[0007] In summary, one or more technical solutions provided in the present application solve the technical problem that harmonic interference caused by cable circuit load propagates through the cable, has an adverse effect on the insulation layer, and limits the accuracy of locating cable insulation defect points. It achieves the technical effect of setting an external monitoring circuit at the cable interface, adopting a node-dividing strategy, comprehensively and synchronously monitoring the widely distributed cable system, building an edge topology network, and performing detailed analysis of the circuit load through each edge node, effectively eliminating harmonics caused by non-insulation layer defects, performing multi-scale feature extraction on harmonic changes related to insulation defects, accurately capturing the impact of load changes on harmonic signals, and enhancing the ability to identify insulation layer defects. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 A schematic flow chart of a method for online monitoring of cable insulation layer is provided for this application; Figure 2 A schematic flow chart of obtaining a load harmonic processing model in an online monitoring method for a cable insulation layer is provided for the present application. DETAILED DESCRIPTION

[0009] The present application is described in detail below with reference to the accompanying drawings. Figure 1 As shown, the present application provides a method for online monitoring of cable insulation layer, wherein the method comprises: S1: Acquire multiple monitoring circuits connected to the cable monitoring system, each monitoring circuit is used to perform harmonic monitoring on the cable and output multiple harmonic signals; S2: Using the multiple monitoring circuits as edge nodes and the cable monitoring system as a cloud center node to generate an edge topology network; The operating status of the cable insulation layer is judged by the change of the current and voltage flowing through the cable. Specifically, the megohmmeter method, that is, judging the insulation quality of the cable by measuring the insulation resistance value of the cable, is usually carried out in a power outage state and is suitable for detecting the overall insulation condition of the cable; the high-voltage flashover method, that is, applying a high voltage higher than the working voltage to the cable to induce and locate the weakness or existing fault point of the insulation layer, is suitable for detecting the withstand voltage strength of the insulation layer; the bridge method, that is, detecting the defects of the insulation layer by comparing the resistance value or capacitance value of the cable, is suitable for finding local damage inside the cable; In actual use, it is found that the harmonic interference caused by the cable circuit load does limit the accuracy of locating the cable insulation defect points. Specifically, when the cable circuit load is heavy, high-order harmonics may be generated. High-order harmonics not only affect the normal operation of the power system, but also propagate through the cable and have an adverse effect on the insulation layer. The presence of high-order harmonics will cause the local temperature of the cable insulation layer to rise, accelerate its aging process, and then cause insulation failure; Long-term monitoring and research have found that, on the one hand, long-distance cable lines pass through different geological and climatic environments, and changes in environmental factors will cause changes in the physical and chemical properties of the cable insulation layer, which in turn affects the measurement results of current and voltage. For example, temperature changes will cause the cable material to expand and contract, resulting in changes in the resistivity of the insulation layer; humidity changes will cause the insulation layer to become damp, reducing its insulation performance; on the other hand, there are many interference factors in long-distance cable lines, such as electromagnetic interference and noise interference. Interference factors will affect the transmission and measurement of current and voltage signals, resulting in inaccurate detection results. Especially in a complex power grid environment, electromagnetic interference from adjacent lines may have a serious impact on the insulation detection of long-distance cables. Based on this, the present application sets an external monitoring circuit at the cable interface and adopts a node-dividing strategy to ensure comprehensive and synchronous monitoring of the widely distributed cable system. The constructed edge topology network improves the efficiency of data processing, and conducts a detailed analysis of the circuit load through the edge node, effectively eliminating the harmonics caused by non-insulating layer defects, and improving the accuracy and real-time performance of the monitoring data. Specifically, for different sections of the cable, multiple monitoring circuits are designed and installed, and the multiple monitoring circuits are designed to be able to access the cable interface to ensure real-time monitoring of the electrical performance of the cable, especially the capture of harmonics, wherein each monitoring circuit contains necessary sensors, signal conditioning modules and preliminary data processing units, so as to directly collect harmonic signals on the cable; Build an edge computing architecture, in which each installed monitoring circuit is defined as an edge node. The edge node has a certain data processing capability and can preliminarily analyze the collected harmonic signals, reduce the amount of data transmitted to the cloud center, and improve the response speed. Furthermore, set one or more centralized processing cloud center nodes to coordinate the edge nodes, conduct advanced analysis, model training and optimization, and manage the overall monitoring system; then, establish an edge topology network. Specifically, ensure that each edge node can maintain stable data communication with the cloud center node through wired or wireless means, involving the use of industrial-grade routers, 4G / 5G cellular networks or private network communication technologies. Furthermore, configure network communication protocols, including data transmission protocols, security protocols, etc., to ensure that data is exchanged safely and efficiently between edge nodes and cloud centers; realize online real-time monitoring of cable insulation layer, and provide support for safe operation of power system.

[0010] S3: using multiple edge nodes in the edge topology network to perform load analysis on the multiple monitoring circuits to obtain multiple circuit load information, including the number of circuit elements and the circuit load sizes corresponding to the multiple circuit elements; S4: collecting harmonic component characteristics according to the plurality of circuit load information, and outputting load-harmonic signal characteristics; Utilize multiple edge nodes in the edge topology network to perform load analysis on the multiple monitoring circuits to obtain multiple circuit load information. Specifically, ensure that all monitoring circuits have been correctly installed and connected to the cable, and can continuously collect current and voltage signals in the cable, especially harmonic components. In addition, establish a reliable data communication protocol between the edge node and the cloud center to ensure that the monitoring data can be uploaded to the edge node in real time or regularly for processing. Furthermore, the edge node receives real-time current and voltage data from each monitoring circuit. The real-time current and voltage data received by each monitoring circuit contains the load status information of the circuit. Then, by analyzing the circuit design diagram or historical data, determine the type and quantity of circuit components at each monitoring point, such as transformers, capacitors, resistors, etc.; calculate the real-time load size of each circuit component based on the current and voltage signals and the impedance information of the components; Harmonic component characteristics are collected according to the multiple circuit load information, and load-harmonic signal characteristics are output. Specifically, digital signal processing technology (such as fast Fourier transform FFT) is used to separate the harmonic components from the original signal, eliminate the fundamental wave and other non-harmonic components, and only retain the harmonic signal related to the load. Further, according to the circuit load information, the harmonic signal of each monitoring point is converted into a feature vector, and the feature vector includes the amplitude, phase, frequency component, etc. of a specific harmonic order; the harmonic characteristics are marked in combination with the number of circuit elements and the load size to form a load-harmonic signal feature, that is, each feature vector is associated with a specific circuit load state, wherein each edge node independently processes the monitoring circuit data to which it is connected, and after completing the load analysis and feature extraction, all load-harmonic signal characteristics of the corresponding node are summarized, so as to dynamically and in real time grasp the load state of the cable and the harmonic changes caused by it, and provide a scientific basis for the evaluation of the health status of the cable insulation layer.

[0011] S5: The cloud center node in the edge topology network performs federated training on the load-harmonic signal feature and outputs a load harmonic processing model; S6: downloading the load harmonic processing model to multiple edge nodes, and performing harmonic processing on the multiple harmonic signals using the load harmonic processing model to obtain processed harmonic signals; S7: Evaluate the cable insulation defects on the processed harmonic signal and output a cable insulation defect evaluation index.

[0012] At the cloud center node, a global load harmonic processing model is initialized according to a predefined algorithm framework (such as deep learning, machine learning, etc.). The load harmonic processing model is used to learn the load-harmonic signal features provided by different edge nodes to distinguish and process the harmonic interference caused by the circuit load. Further, the initialized global model is downloaded to each edge node, and each edge node starts local training according to the load-harmonic signal features collected by itself. Among them, each edge node performs model training based on its own circuit load information and corresponding harmonic signal features, including feature extraction, classification, anomaly detection and other operations on the signal; after a certain period of local training, the edge node updates the candidate model obtained by training and uploads it to the cloud center; The cloud center node receives model updates from all edge nodes, and uses a federated learning algorithm (such as FedAvg, federated averaging method) to perform weighted averaging or fusion to generate a new global model, taking into account the diversity of edge nodes and data privacy protection; repeat multiple rounds of federated training until the model converges or reaches the preset stop condition, and then download the final optimized load harmonic processing model to each edge node again. The edge node uses this model to process the received harmonic signal, eliminate the harmonic caused by the circuit load, and only retain the harmonic signal reflecting the insulation condition of the cable itself; The processed harmonic signal is deeply analyzed to extract features related to insulation layer defects, such as the increase of specific harmonic components and frequency changes. Based on the extracted features, the preset evaluation algorithm is applied to quantitatively evaluate the insulation defects of the cable, and generate cable insulation defect evaluation indicators, such as insulation aging degree, damage probability, etc. Furthermore, the cable insulation defect evaluation indicators are sent to the integrated automation management system so that corresponding maintenance or preventive measures can be taken. Furthermore, the model parameters or training strategies are adjusted according to the feedback in actual applications, and the monitoring system performance is continuously optimized to ensure long-term and effective monitoring of the health status of the cable insulation layer, realizing a closed-loop process from model training to cable insulation defect evaluation, and improving the intelligence level of cable maintenance and the accuracy of fault prediction.

[0013] Furthermore, the load-harmonic signal features are federated trained to output a load harmonic processing model, such as Figure 2 As shown, the present application method includes: Training and initializing a global model in a cloud center node of the edge topology network; Downloading the initialized global model to multiple edge nodes, and obtaining the multiple circuit load information corresponding to the multiple edge nodes; Performing training in corresponding edge nodes according to the plurality of circuit load information to obtain a plurality of converged candidate harmonic processing models; The multiple candidate harmonic models are then uploaded to the cloud center node, and the cloud center node performs weighted aggregation on the multiple candidate harmonic models to obtain a load harmonic processing model.

[0014] The specific steps of implementing the federated training process to output a load harmonic processing model are as follows: The load-harmonic signal feature is federated trained to output a load harmonic processing model. Specifically, a basic load harmonic processing model architecture is designed on the cloud center node (the load harmonic processing model architecture can be a convolutional neural network (CNN), a recurrent neural network (RNN) or other deep learning structures suitable for time series signal processing). The load harmonic processing model architecture usually includes multiple levels to identify and separate the harmonic signal caused by the circuit load and the harmonic characteristics of the cable itself. Then, initial weights and bias values ​​are assigned to the model, which is the starting point of the federated training. The initialization method can be random or a preset value based on some prior knowledge. The cloud center node pushes the initialized global model to each edge node. Each edge node is responsible for monitoring a specific part of the cable network, so the received model copies are the same. Furthermore, each edge node collects the corresponding circuit load information and harmonic signal data according to the monitoring circuit to which it is connected. Data preprocessing includes filtering, standardization, normalization and other steps to ensure that the data quality meets the model training requirements. Based on the circuit load information collected by each node, each edge node trains the model locally. During the training process, the model parameters are fine-tuned according to the data set unique to the edge node, so that the model can better adapt to the harmonic characteristics in the local environment. Furthermore, each edge node determines whether its model has reached a convergence state by monitoring indicators such as loss function or accuracy. The convergence standard can be reaching a predetermined number of iterations, the loss value is lower than a threshold, or the performance on the verification set is no longer improved. Then, when the model is trained locally and is considered to have converged, the edge node uploads the candidate model obtained by its training to the cloud center node. The candidate model contains updated model parameters, reflecting the harmonic characteristics of the cable segment covered by the corresponding edge node; After receiving the candidate models from all edge nodes, the cloud center node performs a weighted aggregation operation to integrate the model updates of each edge node. Taking into account the differences in data volume and quality of different nodes, the model parameters are fused through weighted averaging or other strategies to form a more generalized load harmonic processing model. Furthermore, the aggregated model will be tested on a small part of the validation set to ensure that the model performance is improved or at least remains stable. Then, the verified load harmonic processing model is downloaded to the edge node again for real-time processing of the monitored harmonic signals, eliminating the impact of circuit loads, and evaluating the health of the cable insulation layer. Based on the actual application effect and newly collected data, the cloud center node decides to conduct a new round of federated training to continuously iterate and optimize the model performance. The federated training mechanism enhances the generalization ability of the model, protects data privacy, and realizes efficient and distributed online monitoring of the cable insulation layer.

[0015] Furthermore, the present application method also includes: An objective function is constructed. When the convergence condition of the objective function is met, multiple candidate harmonic processing models after convergence are obtained. The expression of the objective function is as follows: ; in, As the objective function, Iteratively update the target to output multiple candidate harmonic models. is the number of edge nodes, For the The edge nodes correspond to the weight parameters of the model, Is a parameter No. Edge nodes for harmonic signals The resulting signal loss function.

[0016] Construct an objective function. Specifically, the expression of the objective function is as follows: ,in, is the objective function corresponding to the entire federated learning system, which aims to minimize the sum of signal losses of all edge nodes. Iteratively update the target to output multiple candidate harmonic models. is the number of edge nodes, For the The edge nodes correspond to the weight parameters of the model, reflecting the The importance or contribution of each edge node, Is a parameter No. Edge nodes for harmonic signals The resulting signal loss function; Local training and loss calculation, further, at each edge node, using the respective harmonic signal data and initialized model parameters Train and optimize , that is, reduce signal loss; for each edge node, calculate the signal loss under the current model parameters , according to the calculated loss function value, the model parameters Perform local gradient descent or other optimization algorithm iterative updates to make Approaching the minimum value, where each edge node uploads the updated model parameters or gradient information to the cloud center node; Furthermore, the cloud center node receives the information uploaded by all edge nodes, and updates the global model parameters by combining the updates of all nodes with the weighted average method (or other aggregation rules); calculates the objective function value under the new global model parameters and compares it with the value of the previous round. If the change is less than the preset threshold or reaches the preset number of iterations, the model is considered to have converged; if the convergence condition is not met, the iterative optimization continues until the convergence condition is met. Then, once the model converges, the cloud center node distributes the final global model parameters back to each edge node. The final global model parameters represent the converged candidate harmonic processing model. The federated learning process based on the objective function effectively integrates the data and computing power of multiple edge nodes for harmonic signal processing under different circuit load conditions.

[0017] Furthermore, to obtain multiple circuit load information, the method of the present application also includes: Judging the plurality of edge nodes according to the plurality of circuit load information, and obtaining component similarity and load similarity; Perform weighting according to the component similarity and the load similarity to obtain a similarity weighted result; Clustering the plurality of edge nodes according to the similarity weighted result, and outputting a plurality of types of edge nodes; Federated learning is performed on the multiple types of edge nodes respectively to obtain multiple types of load harmonic processing models.

[0018] Acquire multiple circuit load information, specifically, collect real-time load information of the cable segments connected to each edge node through monitoring circuits, including the number of circuit components, the load size of each component, etc.; analyze the types and configurations of circuit components monitored by each edge node, and evaluate the similarity between each edge node, for example, compare based on the model, specification, function and other attributes of the component; calculate the load level of each edge node based on the collected load size data, and evaluate the similarity of the load mode, involving comparing the load fluctuation range, peak value, mean value and other indicators, and further, assign appropriate weight parameters to component similarity and load similarity, depending on which aspect is more important in actual applications, for example, if the component configuration has a greater impact on harmonics, its weight can be set higher; Combined with the scores of component similarity and load similarity, the comprehensive similarity score of each edge node is calculated through a weighted formula. Further, an appropriate clustering algorithm, such as K-means, hierarchical clustering or DBSCAN, is selected to group the edge nodes according to the comprehensive similarity score. The selected clustering algorithm is run to assign the edge nodes to different categories according to the similarity weighted results to form multiple categories of edge nodes. Then, an initial harmonic processing model framework is prepared for each category of edge nodes, or a global initial model is used as a starting point. On the edge nodes of each category, the data of this category is used to perform local training of the model, reflecting the decentralized characteristics of federated learning, that is, the model is optimized on local data; the edge nodes of each category upload the trained model parameters or gradients to the cloud center node, and the cloud center node performs aggregation operations on the model parameters, such as weighted averaging, to generate a global model for the category; multiple rounds of local training and cloud aggregation are performed to further improve the model performance until the model converges or reaches the predetermined stop condition; After completing federated learning, the obtained multi-category load harmonic processing models are deployed back to the corresponding edge nodes. Each category of edge nodes applies its own model to process harmonic signals, thereby more accurately evaluating cable insulation defects and realizing effective utilization of circuit load information in online monitoring of cable insulation layers, improving the pertinence and accuracy of the model, and thereby enhancing the flexibility and adaptability of the system.

[0019] Furthermore, the present application method also includes: After obtaining multiple types of load harmonic processing models, the multiple types of load harmonic processing models are used to perform harmonic processing on multiple harmonic signals output by the multiple types of edge nodes respectively; Each type of edge node among the multiple types of edge nodes stores a corresponding load harmonic processing model, and each type of edge node is connected to a monitoring circuit terminal of the type to which it belongs.

[0020] Ensure that the corresponding load harmonic processing model has been loaded into the local storage or memory of each type of edge node, which is obtained by optimizing the characteristics of each category in the previous federated learning process; before formal application, perform functional verification on the model on each type of edge node to confirm that the model is loaded correctly and can respond and process data normally. Furthermore, the monitoring circuit on each edge node continuously monitors the cable harmonics and captures the harmonic signals in the cable in real time; the monitoring circuit terminal transmits the captured original harmonic signal to the edge node of the corresponding category for processing.

[0021] When the edge node receives the harmonic signal, it immediately processes the signal using the deployed load harmonic processing model, including using the model for filtering, noise reduction, feature extraction, etc., to remove unnecessary harmonic components introduced by the circuit load. Furthermore, according to the specific design of the model, the processing parameters need to be dynamically adjusted to adapt to the actual working status and environmental changes of the current cable. The processed harmonic signal is further analyzed to evaluate the potential defects of the cable insulation layer. This usually involves analyzing the changing trend of the harmonic signal, the abnormality of the frequency component, etc. In actual application, if there are signs of insulation defects, the edge node should generate an alarm signal in time and report to the monitoring center through the network so that preventive measures can be taken or maintenance can be arranged; monitor the model processing effect regularly or as needed, evaluate the model accuracy and stability, and collect feedback data on the model performance; according to the monitoring results and new monitoring data, the model is fine-tuned or updated in time to maintain the optimal performance of the model. The updated model is redeployed to the edge node to ensure the efficiency and accuracy of online monitoring of the cable insulation layer. At the same time, the multi-type load harmonic processing model obtained by federated learning is used to provide customized processing strategies for different types of circuit load environments to enhance the reliability of the system.

[0022] Furthermore, the harmonic component characteristics are collected according to the plurality of circuit load information, and load-harmonic signal characteristics are output. The method of the present application includes: According to the plurality of circuit load information, setting a circuit load sample; Acquire harmonic component samples based on the circuit load samples; Inputting the harmonic component samples into a feature convolution network for multi-scale feature convolution, and outputting multi-scale harmonic signal features; The change characteristics of the multi-scale harmonic signal characteristics under the circuit load sample are analyzed, and a load-harmonic signal characteristic is output, wherein the load-harmonic signal characteristic is a change characteristic of the harmonic component signal affected by the load change.

[0023] Harmonic component characteristics are collected based on the multiple circuit load information, and load-harmonic signal characteristics are output. Specifically, circuit operation data under different time periods or working conditions are collected from the monitoring system, including key parameters such as current, voltage, and power; based on the collected data, different circuit load states are defined, such as light load, full load, transient load, etc., and circuit load sample sets are created based on these states. Each sample contains electrical parameters of the circuit under specific load conditions; For each circuit load sample, a fast Fourier transform (FFT) or other spectrum analysis method is applied to convert the time domain signal to the frequency domain to separate the harmonic components; then, each harmonic component is associated with its corresponding circuit load state to form a labeled harmonic component sample set to facilitate subsequent feature learning and analysis; a feature convolution network is constructed, wherein the feature convolution network includes multiple convolution layers, each convolution layer uses convolution kernels of different scales to capture the characteristics of the harmonic signal at different time scales or frequency bandwidths; the harmonic component samples are input into the feature convolution network, wherein each sample is processed by multiple layers of convolution of the network, and each layer focuses on different feature scales of the signal; The output of the network is a multi-scale harmonic signal feature, which reflects the intensity, distribution and changing trend of the harmonic signal in different frequency bands and time scales. By comparing the changes in harmonic features under different load states, the specific pattern of the impact of load changes on harmonic signals is identified, such as the phenomenon that some harmonic components are significantly enhanced as the load increases. Further, the analysis results are integrated into load-harmonic signal features, including determining which harmonic components are most sensitive to load changes and the laws of change with changes in load states. Then, the load-harmonic signal features are determined to provide a basis for subsequent health assessment of the cable insulation layer, and a detailed analysis of harmonic signals in complex circuit environments is achieved. The harmonic components most affected by the load and their changing characteristics are described in detail, as well as the characteristics indicating potential insulation problems, providing data support for online monitoring of the cable insulation layer.

[0024] Furthermore, the harmonic component samples are input into a feature convolution network for multi-scale feature convolution, and multi-scale harmonic signal features are output. The method of the present application includes: Setting a multi-scale convolution kernel, and using the multi-scale convolution kernel to train a feature convolution network; Inputting the harmonic component samples into a trained feature convolutional network, wherein the convolutional features of the feature convolutional network include frequency characteristics, amplitude characteristics and wavelength characteristics; The harmonic component samples are subjected to multi-scale feature convolution according to the frequency characteristics, amplitude characteristics and wavelength characteristics, and multi-scale harmonic signal characteristics are output.

[0025] The harmonic component samples are input into a feature convolution network for multi-scale feature convolution, and multi-scale harmonic signal features are output. Specifically, circuit operation data under different time periods or working conditions are collected from the monitoring system, including key parameters such as current, voltage, and power; different circuit load states are defined based on the collected data, such as light load, full load, transient load, etc., and a circuit load sample set is created based on the circuit load state, wherein each sample contains electrical parameters of the circuit under specific load conditions; For each circuit load sample, fast Fourier transform (FFT) or other spectrum analysis methods are applied to convert the time domain signal into the frequency domain and separate the harmonic components. Then, each harmonic component is associated with its corresponding circuit load state to form a labeled harmonic component sample set for subsequent feature learning and analysis. Further, a feature convolution network is constructed. The feature convolution network contains multiple convolution layers, each of which uses convolution kernels of different scales to capture the characteristics of harmonic signals at different time scales or frequency bandwidths. The harmonic component samples are input into the feature convolution network, where each sample is processed by multiple layers of convolution of the network, and each layer focuses on different characteristic scales of the signal. The output is multi-scale harmonic signal features, which reflect the intensity, distribution and change trend of the harmonic signal at different frequency bands and time scales. By comparing the changes in harmonic features under different load states, the specific pattern of the impact of load changes on harmonic signals is identified, such as the phenomenon that some harmonic components are significantly enhanced as the load increases.

[0026] The analysis results are integrated into load-harmonic signal characteristics, including determining which harmonic components are most sensitive to load changes and how the harmonic components change with load status; the load-harmonic signal characteristics are sorted and reported to determine the harmonic components with the greatest load impact and their changing characteristics as well as the characteristics indicating potential insulation problems, providing a basis for subsequent health assessment of the cable insulation layer, achieving detailed analysis of harmonic signals in complex circuit environments, and providing support for online monitoring of the cable insulation layer.

[0027] Furthermore, the load harmonic processing model is used to perform harmonic processing on the multiple harmonic signals. The method of the present application includes: Connecting the load harmonic processing model to a low-pass filter; The harmonic processing parameters output by the load harmonic processing model are obtained, and the low-pass filter performs harmonic processing on the multiple harmonic signals according to the harmonic processing parameters to obtain processed harmonic signals.

[0028] Performing harmonic processing on the multiple harmonic signals using the load harmonic processing model, specifically, ensuring that the load harmonic processing model has been federally trained and downloaded to the edge node or the centralized processing unit, the load harmonic processing model being used to identify and process non-essential harmonics caused by the circuit load; configuring a digital low-pass filter, the digital low-pass filter will work in conjunction with the load harmonic processing model, the design of the low-pass filter needs to determine the cut-off frequency based on the frequency characteristics of the cable insulation layer monitoring, to ensure that only signals below the frequency are allowed to pass, thereby effectively removing high-frequency noise and load-related harmonics; Extracting the required harmonic processing parameters from the load harmonic processing model, the harmonic processing parameters include cutoff frequency, attenuation coefficient, etc., which are crucial for guiding the low-pass filter on how to accurately filter out the circuit load harmonics. Further, applying the extracted parameters to the low-pass filter, through software programming or hardware configuration, ensuring that the performance of the filter matches the output parameters of the model to achieve the best harmonic rejection effect, and then sending multiple harmonic signals output by the monitoring circuit into the integrated system. The multiple harmonic signals contain all types of harmonics, including those generated by cable insulation defects and those generated by circuit loads; The input harmonic signal is preliminarily analyzed and preprocessed using the load harmonic processing model, which involves feature extraction, classification or pre-filtering of the signal, providing more precise guidance for the subsequent filtering steps. Furthermore, based on the parameters provided by the model, the low-pass filter processes the signal to remove the harmonic components related to the circuit load, and only retains the harmonic signal related to the cable insulation defect; the processed harmonic signal is output, and most of the load influence in the signal has been eliminated, which is conducive to the subsequent analysis step to more accurately evaluate the cable insulation condition, and then, the processed harmonic signal is used to evaluate the cable insulation defect, which involves analyzing the amplitude and frequency changes of the harmonics, or comparing with known defect patterns; according to the evaluation results and actual monitoring effects, the model parameters and filter configuration are continuously adjusted to improve the monitoring accuracy and efficiency. A complete process from model parameter application to signal processing to the final insulation defect evaluation is realized to ensure the effectiveness and accuracy of online monitoring of the cable insulation layer.

[0029] In summary, the beneficial effects of the embodiments of the present application are: By adopting distributed node monitoring technology, combined with edge cloud computing architecture, and using edge computing nodes to perform load analysis and feature extraction on harmonic signals, combined with federated learning technology, it effectively eliminates the interference caused by load and environmental factors, and focuses on analyzing harmonic signals directly related to the health status of the insulation layer, realizing real-time and synchronous monitoring of long-distance cables, and improving the accuracy and efficiency of monitoring.

[0030] Through the multi-scale feature convolutional network, the frequency, amplitude and wavelength characteristics of the harmonic signal are deeply analyzed, and the edge nodes are clustered and customized model training is performed, so that the monitoring system can adapt to different types of loads and circuit configurations, and improve the sensitivity and accuracy of insulation layer defect identification.

[0031] 3. Since the circuit load samples are set according to multiple circuit load information; the harmonic component samples based on the circuit load samples are obtained; the harmonic component samples are input into the feature convolution network for multi-scale feature convolution, and the multi-scale harmonic signal features are output; the change characteristics of the multi-scale harmonic signal characteristics under the circuit load samples are analyzed, and the load-harmonic signal characteristics are output, where the load-harmonic signal characteristics are the change characteristics of the harmonic component signal affected by the load change. The fine analysis of the harmonic signal in the complex circuit environment is realized, and the harmonic components most affected by the load and their change characteristics are explained in detail, as well as the potential insulation problems indicated by the characteristics, providing data support for the online monitoring of the cable insulation layer.

[0032] In summary, any step can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor, and no unnecessary restrictions are made here.

[0033] Furthermore, the above-mentioned technical scheme only reflects the preferred technical scheme of the technical scheme of the embodiment of the present application. Some changes that may be made to some parts thereof by technicians in this technical field all reflect the novel principles of the embodiment of the present application. Obviously, technicians in this field can make various changes and modifications to the present application without departing from the scope of the present application.

Claims

1. A method for online monitoring of cable insulation layer, characterized in that: The method comprises: Acquire multiple monitoring circuits connected to the cable monitoring system, each monitoring circuit is used to perform harmonic monitoring on the cable and output multiple harmonic signals; Using the multiple monitoring circuits as edge nodes and the cable monitoring system as a cloud center node to generate an edge topology network; Using multiple edge nodes in the edge topology network to perform load analysis on the multiple monitoring circuits, and obtain multiple circuit load information, including the number of circuit elements and the circuit load sizes corresponding to the multiple circuit elements; Collect harmonic component characteristics according to the multiple circuit load information, and output load-harmonic signal characteristics; The cloud center node in the edge topology network performs federated training on the load-harmonic signal features and outputs a load harmonic processing model; Downloading the load harmonic processing model to multiple edge nodes, and performing harmonic processing on the multiple harmonic signals using the load harmonic processing model to obtain processed harmonic signals; The processed harmonic signal is used to evaluate the cable insulation defects and output the cable insulation defect evaluation index.

2. The method according to claim 1, characterized in that The load-harmonic signal feature is federated trained to output a load harmonic processing model, the method comprising: Training and initializing a global model in a cloud center node of the edge topology network; Downloading the initialized global model to multiple edge nodes, and obtaining the multiple circuit load information corresponding to the multiple edge nodes; Performing training in corresponding edge nodes according to the plurality of circuit load information to obtain a plurality of converged candidate harmonic processing models; The multiple candidate harmonic models are then uploaded to the cloud center node, and the cloud center node performs weighted aggregation on the multiple candidate harmonic models to obtain a load harmonic processing model.

3. The method according to claim 2, characterized in that An objective function is constructed. When the convergence condition of the objective function is met, multiple candidate harmonic processing models after convergence are obtained. The expression of the objective function is as follows: ； in, As the objective function, Iteratively update the target to output multiple candidate harmonic models. is the number of edge nodes, For the The edge nodes correspond to the weight parameters of the model, Is a parameter No. Edge nodes for harmonic signals The resulting signal loss function.

4. The method according to claim 1, characterized in that Acquiring multiple circuit load information, the method further includes: Judging the plurality of edge nodes according to the plurality of circuit load information, and obtaining component similarity and load similarity; Perform weighting according to the component similarity and the load similarity to obtain a similarity weighted result; Clustering the plurality of edge nodes according to the similarity weighted result, and outputting a plurality of types of edge nodes; Federated learning is performed on the multiple types of edge nodes respectively to obtain multiple types of load harmonic processing models.

5. The method according to claim 4, characterized in that After obtaining multiple types of load harmonic processing models, the multiple types of load harmonic processing models are used to perform harmonic processing on multiple harmonic signals output by the multiple types of edge nodes respectively; Each type of edge node among the multiple types of edge nodes stores a corresponding load harmonic processing model, and each type of edge node is connected to a monitoring circuit terminal of the type to which it belongs.

6. The method according to claim 1, characterized in that Harmonic component characteristics are collected according to the plurality of circuit load information, and load-harmonic signal characteristics are output, the method comprising: According to the plurality of circuit load information, setting a circuit load sample; Acquire harmonic component samples based on the circuit load samples; Inputting the harmonic component samples into a feature convolution network for multi-scale feature convolution, and outputting multi-scale harmonic signal features; The change characteristics of the multi-scale harmonic signal characteristics under the circuit load sample are analyzed, and a load-harmonic signal characteristic is output, wherein the load-harmonic signal characteristic is a change characteristic of the harmonic component signal affected by the load change.

7. The method according to claim 6, characterized in that Inputting the harmonic component samples into a feature convolution network for multi-scale feature convolution, and outputting multi-scale harmonic signal features, the method includes: Setting a multi-scale convolution kernel, and using the multi-scale convolution kernel to train a feature convolution network; Inputting the harmonic component samples into a trained feature convolutional network, wherein the convolutional features of the feature convolutional network include frequency characteristics, amplitude characteristics and wavelength characteristics; The harmonic component samples are subjected to multi-scale feature convolution according to the frequency characteristics, amplitude characteristics and wavelength characteristics, and multi-scale harmonic signal characteristics are output.

8. The method according to claim 1, characterized in that The load harmonic processing model is used to perform harmonic processing on the multiple harmonic signals, and the method includes: Connecting the load harmonic processing model to a low-pass filter; The harmonic processing parameters output by the load harmonic processing model are obtained, and the low-pass filter performs harmonic processing on the multiple harmonic signals according to the harmonic processing parameters to obtain processed harmonic signals.

Images