Digital testing method, system and equipment for electrical equipment

Abstract

The invention discloses a digital testing method, system and device for electrical equipment, and relates to the related field of fault testing, and the method comprises the steps: carrying out the test demand analysis of target equipment, and constructing a multi-modal test sensing network to collect multi-modal operation data; selecting a communication mode according to the application scene and the equipment characteristics, and transmitting the operation data to a test cloud; the cloud calls multiple fault prediction channels and carries out channel matching analysis, and a target fault prediction channel is determined; and carrying out fault identification based on the prediction channel, outputting a test result and carrying out feedback response. The technical problems that due to the fact that a traditional electrical equipment testing method is single in data collection mode and limited in communication means, the multi-modal operation state of the equipment cannot be comprehensively reflected, the testing result is low in accuracy, and fault prediction is not timely are solved, and the purpose of improving the testing efficiency through multi-modal sensing and cloud multi-channel analysis is achieved. The technical effects of high-efficiency transmission and accurate fault prediction of the operation data of the electrical equipment, and improvement of the intelligent level and reliability of the test are achieved.

Description

Technical Field

[0001] This application relates to the field of fault testing, and in particular to a digital testing method, system and equipment for electrical equipment. Background Technology

[0002] Electrical equipment is widely used in key sectors such as modern industry, energy, and transportation. Equipment includes transformers, switchgear, cables, GIS (Gas Insulation System), converter valves, energy storage converters, and wind-solar grid-connected units. The safety and stability of their operation directly affect production efficiency and public safety. However, with the development of new power systems and the integration of high-proportion renewable energy sources into the grid, the complexity of equipment operating conditions has significantly increased, exhibiting characteristics such as multi-source disturbances, large load fluctuations, frequent start-ups / switches, and environmental variations. Traditional testing techniques generally employ periodic inspections, offline sampling, or threshold alarms based on single measurements. However, these techniques still have limitations in engineering applications. First, various types of fault information are distributed across signals with different physical quantities and sampling frequencies, making it difficult for single-mode sensing to fully reflect the equipment status, thus limiting fault identification accuracy. Second, heterogeneous acquisition protocols and variable network conditions make data aggregation and timing alignment difficult. Third, in complex and variable operating environments, equipment status and fault characteristics exhibit significant time-varying and nonlinear characteristics, making traditional diagnostic methods based on empirical rules or single models unsuitable for identifying multiple operating conditions and types of faults. Fourth, there is a lack of efficient collaboration mechanisms between the equipment field, edge computing, and cloud analysis. Data transmission, real-time assurance, and multi-channel parallel management of cloud models are still imperfect, making it difficult to achieve closed-loop early warning and response from end to cloud, thus limiting overall testing efficiency and response speed. Summary of the Invention

[0003] This application provides a digital testing method, system, and equipment for electrical equipment, which solves the technical problems of traditional electrical equipment testing methods, such as the inability to fully reflect the multimodal operating status of equipment, low accuracy of test results, and untimely fault prediction due to the single data acquisition method and limited communication means. It achieves the technical effect of realizing efficient transmission of electrical equipment operating data and accurate fault prediction through multimodal perception and cloud multi-channel analysis, thereby improving the level of intelligent testing and reliability.

[0004] This application provides a digital testing method for electrical equipment, comprising: analyzing the testing requirements of the target electrical equipment, constructing a multimodal testing perception network, and collecting multimodal operating data of the electrical equipment through the multimodal testing perception network; selecting a target communication method according to the application scenario and equipment characteristics of the target electrical equipment, and transmitting the multimodal operating data of the electrical equipment to a digital testing cloud using the target communication method; calling a multi-channel fault prediction system for the electrical equipment through the digital testing cloud, performing channel matching analysis on the multimodal operating data of the electrical equipment based on the multi-channel fault prediction system, and determining the target electrical equipment fault prediction channel; performing fault prediction and identification on the multimodal operating data of the electrical equipment based on the target electrical equipment fault prediction channel, outputting electrical equipment fault test results, and providing feedback test response based on the electrical equipment fault test results.

[0005] This application also provides a digital testing system for electrical equipment, comprising: a data acquisition module: performing test requirement analysis on the target electrical equipment, constructing a multimodal test sensing network, and acquiring multimodal operating data of the electrical equipment through the multimodal test sensing network; a data transmission module: selecting a target communication method according to the application scenario and equipment characteristics of the target electrical equipment, and transmitting the multimodal operating data of the electrical equipment to a digital testing cloud using the target communication method; a channel matching module: calling the electrical equipment fault prediction multi-channel through the digital testing cloud, performing channel matching analysis on the multimodal operating data of the electrical equipment based on the electrical equipment fault prediction multi-channel, and determining the target electrical equipment fault prediction channel; and a fault identification module: performing fault prediction and identification on the multimodal operating data of the electrical equipment based on the target electrical equipment fault prediction channel, outputting electrical equipment fault test results, and providing feedback test response based on the electrical equipment fault test results.

[0006] This application also provides an electronic device, including: a memory for storing executable instructions; and a processor for implementing a digital testing method for electrical equipment when executing the executable instructions stored in the memory.

[0007] This application proposes a digital testing method, system, and equipment for electrical equipment. First, it analyzes the testing requirements of the electrical equipment and establishes a testing network capable of sensing multiple types of information to collect multimodal operational data. Then, based on the actual application scenario and characteristics of the equipment, a suitable communication method is selected to transmit this operational data to a digital testing cloud. In the cloud, a multi-channel fault prediction model is invoked to match and analyze the collected data, determining the applicable prediction channel. Finally, this prediction channel is used to identify faults in the data, generate test results, and feed the results back to the equipment, forming a closed loop of testing and response. This achieves efficient transmission of electrical equipment operational data and accurate fault prediction, improving the intelligence and reliability of the testing process. Attached Figure Description

[0008] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0009] Figure 1 This is a flowchart illustrating a digital testing method for electrical equipment provided in an embodiment of this application.

[0010] Figure 2 This is a schematic diagram of the structure of a digital testing system for electrical equipment provided in an embodiment of this application.

[0011] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0012] Explanation of reference numerals in the attached drawings: Data acquisition module 11, data transmission module 12, channel matching module 13, fault identification module 14, processor 21, memory 22, input device 23, output device 24. Detailed Implementation

[0013] The above description is merely an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below.

[0014] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application will be provided in conjunction with the accompanying drawings. The described embodiments should not be considered as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0015] In the following description, references to "some embodiments" describe a subset of all possible embodiments. However, it is understood that "some embodiments" can be the same or different subsets of all possible embodiments and can be combined with each other without conflict. The terms "first" and "second" are used merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only.

[0016] This application provides a digital testing method for electrical equipment, such as... Figure 1 As shown, the method includes: A test requirement analysis is performed on the target electrical equipment, a multimodal test sensing network is constructed, and multimodal operation data of the electrical equipment is collected through the multimodal test sensing network.

[0017] Specifically, the testing requirements of the target electrical equipment are first analyzed by considering its operating environment, structural characteristics, and potential failure modes. This identifies the testing targets for the target electrical equipment in practical applications, such as key operating parameters like voltage, current, temperature, transmission rate, transmission delay, and packet loss rate. After completing the testing requirements analysis, based on the identified testing targets, suitable multimodal sensor types are selected, and the sensor deployment and installation locations are designed to construct a multimodal testing and sensing network covering the key operating components of the electrical equipment. This multimodal testing and sensing network allows for comprehensive data collection of the target electrical equipment's operating status under different conditions, generating multimodal operating data. This provides a reliable data foundation for subsequent communication transmission, cloud modeling, and fault prediction.

[0018] In one possible implementation, a multimodal test-aware network is constructed, including: A test requirement analysis is performed on the target electrical equipment to determine the test requirement parameters. Based on the test requirement parameters, multimodal sensors are selected to obtain the multimodal sensor specifications. Fault impact analysis and deployment location identification are performed on the target electrical equipment according to the multimodal sensor specifications to obtain the multimodal sensor deployment parameters. Based on the multimodal sensor specifications and deployment parameters, a sensor network topology is designed to construct the multimodal test sensing network.

[0019] Specifically, the testing requirements of the target electrical equipment are first analyzed by considering its operating environment, structural characteristics, and historical fault modes. This analysis determines the key testing parameters that need to be monitored during operation. For example, for transformers, voltage, current, temperature, partial discharge signals, data transmission rate, transmission delay, and packet loss rate can be considered key monitoring indicators. For circuit breakers, mechanical action characteristics, contact temperature rise, arcing signals, data transmission rate, transmission delay, and bit error rate can be used as testing parameters. After obtaining the testing requirements parameters, multimodal sensors are selected based on these parameters. For example, for voltage and current, high-precision current transformers and voltage sensors can be selected; for temperature, infrared temperature sensors or fiber optic temperature sensors can be selected; for communication parameters such as data transmission rate, transmission delay, packet loss rate, and bit error rate, digital testing instruments for electrical equipment can be selected. These instruments not only have the function of receiving commands and outputting data in message form, but also can automatically output detection / test data digitally, supporting communication methods such as WLAN, Bluetooth, RS232, and USB. By summarizing the relevant parameters of the selected multimodal sensors, the multimodal sensor specifications can be obtained. Then, based on these specifications, and using the historical fault dataset of the target electrical equipment as input, the correspondence between various fault characteristics and sensors is calculated, resulting in the multimodal equipment fault influence coefficient. Based on these coefficients, the key location set of the equipment can be identified, such as high-voltage outgoing terminals, winding hotspots, moving contacts, and weak insulation points. Combined with physical installation feasibility, the deployment parameters of the multimodal sensors are determined. Next, based on the multimodal sensor specifications and deployment parameters, the sensor network topology is designed, including determining the number and location of nodes, planning communication links between sensors, designing data aggregation paths, and implementing redundancy backup. This ensures that the multimodal sensors can work collaboratively under different operating conditions, guaranteeing the comprehensiveness, real-time performance, and stability of data acquisition. Ultimately, a multimodal test and sensing network covering the target electrical equipment is constructed, providing complete support for subsequent data acquisition and fault prediction.

[0020] In one possible implementation, the deployment parameters for the multimodal sensor are obtained, including: Historical fault data mining is performed on the target electrical equipment to obtain a historical fault dataset. The type of multimodal sensor is determined according to the multimodal sensor specifications. Fault impact analysis is performed on the historical fault dataset of the electrical equipment according to the multimodal sensor type to obtain a set of multimodal equipment fault impact coefficients. Key locations of the target electrical equipment are identified based on the set of multimodal equipment fault impact coefficients to obtain a set of multimodal equipment key locations. Finally, multimodal sensor deployment parameters are determined based on the set of multimodal equipment key locations.

[0021] Specifically, firstly, by mining and organizing the operation and maintenance records, inspection logs, and online monitoring data of the target electrical equipment at different operating stages, a historical fault dataset of the target electrical equipment is constructed. This historical fault dataset can include current, voltage, temperature, data transmission rate, latency, packet loss rate, etc., and is labeled according to time series and fault labels. The fault label is usually 1 or 0, where 1 indicates a fault has occurred and 0 indicates normal operation. Subsequently, based on the determined multimodal sensor specifications, the required multimodal sensor types are identified, such as voltage sensors, infrared temperature sensors, and digital testing instruments for electrical equipment. Then, according to the determined multimodal sensor types, the data that can be collected is extracted from the historical fault dataset of the electrical equipment. For each data type, the covariance between the data and the fault label is calculated, and then this covariance is divided by the product of the standard deviation of the data and the standard deviation of the fault label to obtain the Pearson correlation coefficient of each data type as the multimodal equipment fault influence coefficient, which is used to characterize the degree of fault influence of each data type. For sensors that can collect multiple data types, the maximum Pearson correlation coefficient of these data types is used as the fault influence coefficient. Subsequently, based on this set of multimodal equipment fault impact coefficients, the maximum value or weighted fusion of the multimodal equipment fault impact coefficients at each location is selected to obtain the impact degree of each location. Then, the impact degree of each location is weighted with the historical fault frequency of that location to obtain an importance score for each location. Based on this importance score, key locations of the target electrical equipment are identified to determine the set of key locations requiring focused monitoring. For example, the high-voltage output terminal and winding hotspots of transformers, and the moving contacts, operating mechanisms, and insulation parts of circuit breakers may all be identified as key locations. Then, based on the set of key locations of multimodal equipment and the multimodal equipment fault impact coefficients, the types of multimodal sensors required to be deployed at each key location of multimodal equipment are determined through impact threshold screening. Based on the required sensor types, installation conditions, and communication accessibility, the deployment parameters of the multimodal sensors are determined, including sensor model, installation location, number of sensors, communication method, and sampling frequency, thus providing a complete deployment scheme for the subsequent construction of a multimodal test sensing network.

[0022] Based on the application scenario and equipment characteristics of the target electrical equipment, a target communication method is selected, and the multimodal operation data of the electrical equipment is transmitted to the digital test cloud using the target communication method.

[0023] Specifically, based on the actual application scenario and characteristics of the target electrical equipment, the communication requirements are first analyzed. If the target electrical equipment is located in a high electromagnetic interference environment such as a substation, a wired communication method with strong anti-interference capabilities, such as RS232 bus, can be prioritized. If the equipment is in a distributed or mobile scenario, a wireless communication method, such as WLAN or Bluetooth, is preferred. After selecting the target communication method, the message format, data packaging rules, and encrypted transmission protocol are designed based on the characteristics of the collected multimodal operating data to ensure that the data can be automatically output in message form and is compatible with the test cloud. For example, data and communication consistency testing tools can be used to verify the bandwidth, transmission rate, latency, packet loss rate, bit error rate, and connection stability of the selected communication method to ensure stable and reliable data uplink under different operating conditions. For continuous large-capacity data streams, caching and breakpoint resumption mechanisms can be configured to ensure that data transmission can still be restored in the event of network interruption or abnormality. Finally, the target communication method transmits the collected multimodal operating data of the electrical equipment to the digital test cloud in real time and securely, providing data support for subsequent channel matching analysis and fault prediction and identification.

[0024] The digital testing cloud accesses multiple channels for electrical equipment fault prediction, and performs channel matching analysis on the multi-modal operation data of the electrical equipment based on these multiple channels to determine the target electrical equipment fault prediction channel.

[0025] Specifically, after the multimodal operation data of electrical equipment is transmitted to the digital testing cloud via communication, the digital testing cloud first calls the pre-built and stored electrical equipment fault prediction multi-channel. This electrical equipment fault prediction multi-channel consists of multiple predictors trained for different operating conditions and different types of equipment faults. Each prediction channel has the ability to identify specific operating modes or specific fault characteristics. Subsequently, the digital testing cloud compares the received multimodal operation data with the electrical equipment operating conditions to obtain the matching result of the current equipment operating conditions. It then establishes a mapping relationship between this result and each predictor in the fault prediction multi-channel, selecting the prediction channel that best matches the current equipment operating state and using it as the target electrical equipment fault prediction channel. This avoids identification bias caused by differences in operating conditions between different channels, providing a more targeted and accurate analysis path for subsequent fault prediction and identification.

[0026] In one possible implementation, the digital testing cloud is used to invoke multiple channels for electrical equipment fault prediction, including: The historical fault dataset of electrical equipment is subjected to equipment operating condition clustering analysis to obtain the electrical equipment operating condition clustering results; the historical fault dataset of electrical equipment is classified and integrated according to the electrical equipment operating condition clustering results to obtain a multi-operating condition electrical equipment fault dataset; prediction training and optimization are performed based on the multi-operating condition electrical equipment fault dataset to construct a multi-operating condition electrical equipment fault predictor set; the multi-operating condition electrical equipment fault predictor set is parallelized and identified to obtain multiple channels for electrical equipment fault prediction and store them in the digital test cloud.

[0027] Specifically, the process begins by extracting operating condition data from the historical fault dataset of electrical equipment, such as current, voltage, temperature, vibration, data transmission rate, latency, packet loss rate, and bit error rate, to form an operating condition dataset. Multidimensional feature extraction is then performed, and clustering algorithms, such as K-means and DBSCAN, are used to divide the historical operating condition data into different operating condition groups. Taking K-means clustering as an example, the number of clusters k is first set based on common operating modes of electrical equipment or by using contour coefficients. For example, for transformers, k=4 can be set based on "normal operation, overload operation, minor fault, and severe fault." Then, k operating condition data points are randomly selected as initial cluster centers. The Euclidean distance between each historical operating condition data point and each cluster center is calculated, and the data is assigned to the cluster of the nearest cluster center. After all samples are divided, the center point of each cluster is updated to the mean of all operating condition data within that cluster. The above process is repeated until the cluster centers converge or the preset number of iterations is reached, thereby obtaining the electrical equipment operating condition clustering result. This clustering result reflects the typical characteristic distribution of equipment under different operating modes. Subsequently, the historical fault dataset of electrical equipment is classified and integrated according to the clustering result. The original historical fault data is re-divided according to the operating condition category label to form a multi-operating condition electrical equipment fault dataset. This multi-operating condition electrical equipment fault dataset not only retains fault samples under different operating conditions but also maintains the time series relationship and multimodal correspondence between various types of data, providing structured input for subsequent model training. Then, based on each type of operating condition data in this multi-operating condition electrical equipment fault dataset, machine learning methods such as deep neural networks, support vector machines, or random forests are used to train the corresponding operating condition fault predictor, resulting in a multi-operating condition electrical equipment fault predictor set. Finally, the set of fault predictors for multi-condition electrical equipment is identified in parallel. That is, the fault predictors under different conditions are organized in parallel according to the condition labels to form a set of predictors that supports multi-path selection. Each predictor is assigned a unique channel identifier, thereby constructing the final multi-channel electrical equipment fault prediction system. This multi-channel electrical equipment fault prediction system can be called and switched in the digital testing cloud. When newly collected multi-modal operating data is input, the system can automatically select the corresponding prediction channel for analysis based on the condition matching results, thereby ensuring that the fault prediction results are highly matched with the current operating conditions of the equipment and improving the accuracy and real-time performance of fault identification.

[0028] In one possible implementation, a set of multi-condition electrical equipment fault predictors is constructed, including: Fault features are extracted from the multi-condition electrical equipment fault dataset to obtain a multi-condition electrical equipment fault feature fingerprint set; a fault test task list for the electrical equipment is constructed; fault samples are identified in the multi-condition electrical equipment fault feature fingerprint set according to the fault test task list to obtain a multi-condition equipment task fault sample set; prediction training and optimization are performed based on the multi-condition equipment task fault sample set to construct a multi-condition electrical equipment fault predictor set.

[0029] Specifically, firstly, fault feature analysis is performed on the voltage, current, temperature, and data transmission rate data under different operating conditions in the multi-condition electrical equipment fault dataset. Key features that can characterize the occurrence and development of faults are extracted, such as current surge, temperature rise rate, and communication anomaly rate. Then, the extracted multi-dimensional features are encoded and normalized to obtain a multi-condition electrical equipment fault feature fingerprint set. Each feature fingerprint can correspond to the operating state mode under a specific operating condition. Subsequently, a fault test task list for electrical equipment is constructed. This fault test task list is designed according to the operating characteristics of the equipment and common fault types. For example, transformers may include "winding overheat fault detection," "partial discharge fault detection," and "insulation aging identification"; circuit breakers may include "failure to operate fault detection," "contact overheat fault detection," and "opening and closing anomaly detection"; communication links may include "transmission delay anomaly detection" and "packet loss rate anomaly detection," etc. Each task corresponds to a fault scenario that needs to be verified or predicted. After the task list is constructed, fault samples are labeled according to the fault test task list for the multi-condition electrical equipment fault feature fingerprint set. Feature fingerprints under different conditions are bound to fault labels, and fault levels from the historical fault dataset of electrical equipment are used for labeling, forming a multi-condition equipment task fault sample set. Then, a deep neural network (DNN) is used to train each type of task fault sample set in the multi-condition equipment task fault sample set. During training, cross-validation and hyperparameter optimization strategies are introduced to reduce overfitting and improve the model's generalization performance. Simultaneously, sample balancing and data augmentation methods are used to address the sample imbalance problem. Finally, a multi-condition electrical equipment fault predictor set is obtained. Each predictor can effectively identify the target fault under its corresponding operating condition, providing a fundamental support for building a multi-channel electrical equipment fault prediction system.

[0030] In one possible implementation, a set of multi-condition electrical equipment fault predictors is constructed by performing prediction training and optimization based on the multi-condition equipment task fault sample set, including: A deep neural network structure is used to train fault prediction on the fault sample set of the multi-condition equipment task, respectively, to obtain a multi-condition equipment task fault predictor set; cross-validation and iterative optimization are performed on the multi-condition equipment task fault predictor set to obtain a usable multi-condition equipment task fault predictor set; the usable multi-condition equipment task fault predictor set is classified and integrated according to the equipment operating condition type to construct the multi-condition electrical equipment fault predictor set.

[0031] Specifically, after obtaining a multi-condition equipment task fault sample set, this sample set is used as training input to a deep neural network. This deep neural network includes an input layer, several hidden layers, and an output layer. The hidden layers adopt a fully connected structure, and the activation function can be ReLU or Sigmoid to enhance the model's non-linear expressive ability. The output layer corresponds to different fault type labels. During training, iterative training is performed using the backpropagation algorithm and a gradient descent optimizer, such as the Adam optimizer, so that the deep neural network can gradually learn the mapping relationship between fault features and labels under different conditions, thereby obtaining a multi-condition equipment task fault predictor set. Subsequently, the multi-condition equipment task fault sample set is divided into a training set, a validation set, and a test set. Then, a k-fold cross-validation method is used to further divide the training set into k subsets. Each time, k-1 subsets are selected for training, and the remaining subset is used as the validation set. This process is repeated k times to ensure that each subset is used as the validation set once. By calculating performance indicators such as accuracy, recall, and F1 score at different folds, the comprehensive performance evaluation result of the predictor is obtained. Subsequently, based on the cross-validation results, hyperparameters such as the learning rate, batch size, number of layers, and number of nodes of the deep neural network were adjusted, and regularization methods, such as L2 regularization and Dropout, were introduced to improve the model's generalization ability. After optimizing the hyperparameters, the predictor underwent multiple rounds of iterative training. In each round, the weight parameters were updated using the backpropagation algorithm, and performance was evaluated on the validation set. If the performance metrics did not reach the preset threshold, the hyperparameters or training strategies were further adjusted until the model converged. After completing the iterative optimization, the performance of all trained predictors was uniformly evaluated using the test set. Models that met the requirements in terms of accuracy, recall, and overall performance metrics were selected to form a set of usable multi-condition equipment task fault predictors. Finally, based on the category labels of the previous working condition clustering, the available multi-working condition equipment task fault predictor set is classified and integrated according to the equipment working condition type. Under the same working condition category, predictors of corresponding fault types are organized to construct a multi-working condition electrical equipment fault predictor set. This multi-working condition electrical equipment fault predictor set can not only cover common faults of electrical equipment under various typical operating conditions, but also quickly call the corresponding predictor when the equipment operating mode is switched, realizing the working condition adaptive fault prediction capability, and providing complete model support for the subsequent construction of multi-channel fault prediction.

[0032] In one possible implementation, determining the target electrical equipment fault prediction path includes: The multimodal operation data of the electrical equipment is matched with the clustering results of the electrical equipment operating conditions to obtain the matching electrical equipment operating conditions; based on the matching electrical equipment operating conditions, the fault prediction multiple channels of the electrical equipment are associated and mapped to determine the fault prediction channel of the target electrical equipment.

[0033] Specifically, during channel matching, the real-time collected multimodal operating data of electrical equipment is first matched with the pre-obtained clustering results of electrical equipment operating conditions. Euclidean distance is used to measure the similarity between the multimodal operating data and each cluster center, and the cluster with the closest distance is selected as the matching operating condition for the current equipment, thus obtaining the matched electrical equipment operating condition. Next, in the multi-channel electrical equipment fault prediction, an operating condition label retrieval method is used to associate and map the multi-channel based on the matched electrical equipment operating condition, quickly locating and calling the corresponding prediction channel. Finally, the determined target electrical equipment fault prediction channel is connected with the multimodal operating data of electrical equipment for subsequent fault prediction and identification. This method ensures a high degree of matching between the selected prediction channel and the current equipment operating state, thereby improving the accuracy and real-time performance of fault prediction and avoiding misjudgments or omissions caused by inconsistent operating conditions.

[0034] Based on the target electrical equipment fault prediction channel, fault prediction and identification are performed on the multimodal operation data of the electrical equipment, outputting the electrical equipment fault test results, and feedback test response is performed based on the electrical equipment fault test results.

[0035] Specifically, once the fault prediction channel for the target electrical equipment is determined, the real-time collected multimodal operating data of the electrical equipment is input into the fault predictor corresponding to this fault prediction channel for analysis. This feature data is compared with the learned knowledge, and the prediction result of the current equipment operating status is output. This prediction result is used as the electrical equipment fault test result. This electrical equipment fault test result can not only be stored and traced in the digital testing cloud, but can also be fed back to the equipment end or the terminal of the operation and maintenance personnel in real time. The electrical equipment fault test result is then verified, and the corresponding early warning strategy is matched according to the verification result to provide feedback test response to the target electrical equipment, thereby improving the safety and reliability of the electrical equipment operation.

[0036] In one possible implementation, the feedback test response based on the electrical equipment fault test results includes: The target electrical equipment is tested and verified using the electrical equipment fault test results to obtain the electrical equipment fault occurrence verification results; based on the electrical equipment fault occurrence verification results, an early warning strategy analysis is performed to determine the electrical equipment fault classification early warning strategy, and a fault early warning response is performed using the electrical equipment fault classification early warning strategy.

[0037] Specifically, when the fault test results output by the digital testing cloud indicate that the equipment may be faulty, a secondary verification mechanism is triggered. This involves using redundant detection methods, such as backup sensors, historical data comparison, or manual inspection data, for verification. Taking backup sensors as an example, after the secondary verification mechanism is triggered, verification data is collected through the backup sensors to obtain multimodal verification operation data. This multimodal verification operation data is then subjected to the same analysis process as described above to obtain the verification fault test results. By comparing the verification fault test results with the electrical equipment fault test results, it is determined whether the identified fault type and fault level are the same, thus forming a verification result of the electrical equipment fault occurrence. When an electrical equipment fault occurs and the verification result is successful, it indicates that the prediction result is accurate. At this point, an early warning strategy analysis is performed based on the fault level. That is, the fault level is matched with a preset classification strategy to determine the electrical equipment fault classification early warning strategy. For example, in the first-level early warning state, an alarm message is pushed through the information platform, only alerting maintenance personnel. In the second-level early warning state, not only is an alarm issued, but a maintenance work order is also generated and pushed to the maintenance management system to arrange maintenance or load adjustment. In the third-level early warning state, in addition to issuing audible and visual alarms and remote notifications, protective measures can be directly executed through the control interface, such as disconnecting the circuit, activating the backup power supply, or emergency shutdown. Through the above process, a closed-loop control from prediction and verification to early warning response is achieved, improving the operational safety and fault prevention capabilities of electrical equipment.

[0038] In the above text, refer to Figure 1 A digital testing method for electrical equipment according to an embodiment of the present invention is described in detail. Next, reference will be made to... Figure 2 A digital testing system for electrical equipment according to an embodiment of the present invention is described.

[0039] According to an embodiment of the present invention, a digital testing system for electrical equipment is used to solve the technical problems that traditional electrical equipment testing methods, due to their single data acquisition method and limited communication means, cannot fully reflect the multimodal operating status of the equipment, resulting in low accuracy of test results and untimely fault prediction. The system achieves the technical effect of efficient transmission of electrical equipment operating data and accurate fault prediction through multimodal perception and cloud-based multi-channel analysis, thereby improving the level of testing intelligence and reliability.

[0040] A digital testing system for electrical equipment includes: a data acquisition module 11, a data transmission module 12, a channel matching module 13, and a fault identification module 14.

[0041] Data acquisition module 11: Analyzes test requirements for the target electrical equipment, constructs a multimodal test sensing network, and collects multimodal operating data of the electrical equipment through the multimodal test sensing network; Data transmission module 12: Selects a target communication method based on the application scenario and equipment characteristics of the target electrical equipment, and transmits the multimodal operating data of the electrical equipment to the digital test cloud using the target communication method; Channel matching module 13: Calls the electrical equipment fault prediction multi-channel through the digital test cloud, performs channel matching analysis on the multimodal operating data of the electrical equipment based on the electrical equipment fault prediction multi-channel, and determines the fault prediction channel of the target electrical equipment; Fault identification module 14: Performs fault prediction and identification on the multimodal operating data of the electrical equipment based on the fault prediction channel of the target electrical equipment, outputs the electrical equipment fault test result, and provides feedback test response based on the electrical equipment fault test result.

[0042] The specific configuration of the data acquisition module 11 will be described in detail below. As mentioned above, to construct a multimodal test perception network, the data acquisition module 11 may further include: A test requirement analysis is performed on the target electrical equipment to determine the test requirement parameters. Based on the test requirement parameters, multimodal sensors are selected to obtain the multimodal sensor specifications. Fault impact analysis and deployment location identification are performed on the target electrical equipment according to the multimodal sensor specifications to obtain the multimodal sensor deployment parameters. Based on the multimodal sensor specifications and deployment parameters, a sensor network topology is designed to construct the multimodal test sensing network.

[0043] The data acquisition module 11 may further include: Historical fault data mining is performed on the target electrical equipment to obtain a historical fault dataset. The type of multimodal sensor is determined according to the multimodal sensor specifications. Fault impact analysis is performed on the historical fault dataset of the electrical equipment according to the multimodal sensor type to obtain a set of multimodal equipment fault impact coefficients. Key locations of the target electrical equipment are identified based on the set of multimodal equipment fault impact coefficients to obtain a set of multimodal equipment key locations. Finally, multimodal sensor deployment parameters are determined based on the set of multimodal equipment key locations.

[0044] The specific configuration of the channel matching module 13 will be described in detail below. As mentioned above, by calling the electrical equipment fault prediction multi-channel through the digital test cloud, the channel matching module 13 may further include: The historical fault dataset of electrical equipment is subjected to equipment operating condition clustering analysis to obtain the electrical equipment operating condition clustering results; the historical fault dataset of electrical equipment is classified and integrated according to the electrical equipment operating condition clustering results to obtain a multi-operating condition electrical equipment fault dataset; prediction training and optimization are performed based on the multi-operating condition electrical equipment fault dataset to construct a multi-operating condition electrical equipment fault predictor set; the multi-operating condition electrical equipment fault predictor set is parallelized and identified to obtain multiple channels for electrical equipment fault prediction and store them in the digital test cloud.

[0045] Among them, the channel matching module 13, which constructs a multi-condition electrical equipment fault predictor set, may further include: Fault features are extracted from the multi-condition electrical equipment fault dataset to obtain a multi-condition electrical equipment fault feature fingerprint set; a fault test task list for the electrical equipment is constructed; fault samples are identified in the multi-condition electrical equipment fault feature fingerprint set according to the fault test task list to obtain a multi-condition equipment task fault sample set; prediction training and optimization are performed based on the multi-condition equipment task fault sample set to construct a multi-condition electrical equipment fault predictor set.

[0046] Among them, based on the multi-condition equipment task fault sample set, prediction training and optimization are performed to construct a multi-condition electrical equipment fault predictor set, and the channel matching module 13 may further include: A deep neural network structure is used to train fault prediction on the fault sample set of the multi-condition equipment task, respectively, to obtain a multi-condition equipment task fault predictor set; cross-validation and iterative optimization are performed on the multi-condition equipment task fault predictor set to obtain a usable multi-condition equipment task fault predictor set; the usable multi-condition equipment task fault predictor set is classified and integrated according to the equipment operating condition type to construct the multi-condition electrical equipment fault predictor set.

[0047] Among them, the channel matching module 13, which determines the fault prediction channel of the target electrical equipment, may further include: The multimodal operation data of the electrical equipment is matched with the clustering results of the electrical equipment operating conditions to obtain the matching electrical equipment operating conditions; based on the matching electrical equipment operating conditions, the fault prediction multiple channels of the electrical equipment are associated and mapped to determine the fault prediction channel of the target electrical equipment.

[0048] The specific configuration of the fault identification module 14 will be described in detail below. As mentioned above, based on the feedback test response of the electrical equipment fault test results, the fault identification module 14 may further include: The target electrical equipment is tested and verified using the electrical equipment fault test results to obtain the electrical equipment fault occurrence verification results; based on the electrical equipment fault occurrence verification results, an early warning strategy analysis is performed to determine the electrical equipment fault classification early warning strategy, and a fault early warning response is performed using the electrical equipment fault classification early warning strategy.

[0049] The electrical equipment digital testing system provided in this embodiment of the invention can execute the electrical equipment digital testing method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method execution.

[0050] Although this application makes various references to certain modules in the system according to the embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy distinction between each other and are not used to limit the scope of protection of this invention.

[0051] Based on the foregoing embodiments, this application also provides an electronic device. Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention, showing a block diagram of an exemplary electronic device suitable for implementing the embodiments of the present invention. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of the present invention. This electronic device is in the form of a general-purpose computing device, and its components may include, but are not limited to, a processor 21, a memory 22, an input device 23, and an output device 24. The processor 21 may be one or more; the memory 22 may include a computer-readable medium and at least one program product having a set (at least one) of program modules configured to perform the functions of the embodiments of this application.

[0052] The memory 22 shown in the embodiments of the present invention can be any combination of one or more computer-readable media; the computer-readable storage media can be, but is not limited to, infrared, semiconductor systems, devices or components, or any combination thereof, for storing software programs, computer-executable programs and modules, such as the program instructions / modules corresponding to a digital testing method for electrical equipment in an embodiment of the present invention. The processor 21 executes various functional applications and data processing of the computer equipment by running the software programs, instructions and modules stored in the memory 22, thereby realizing the aforementioned digital testing method for electrical equipment.

[0053] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application. In some cases, the actions or steps described in this application can be performed in a different order than that shown in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A method of digitizing testing of electrical equipment, characterized by, The method comprises: Performing test requirement analysis on the target electrical equipment, constructing a multi-modal test perception network, and collecting multi-modal operation data of the electrical equipment through the multi-modal test perception network; According to the application scene and equipment characteristics of the target electrical equipment, a target communication mode is selected, and the multi-modal operation data of the electrical equipment is transmitted to a digital test cloud through the target communication mode; Through the digital test cloud, an electrical equipment fault prediction multi-channel is called, the multi-modal operation data of the electrical equipment is analyzed through channel matching based on the electrical equipment fault prediction multi-channel, and a target electrical equipment fault prediction channel is determined; Based on the target electrical equipment fault prediction channel, the multi-modal operation data of the electrical equipment is subjected to fault prediction identification, and an electrical equipment fault test result is output, and a feedback test response is performed through the electrical equipment fault test result.

2. A method of digitally testing an electrical device as recited in claim 1, wherein, Constructing a multi-modal test perception network comprises: Performing test requirement analysis on the target electrical equipment to determine electrical equipment test requirement parameters; Based on the electrical equipment test requirement parameters, multi-modal sensor selection is performed to obtain multi-modal sensor specification parameters; According to the multi-modal sensor specification parameters, fault influence analysis and deployment location identification are performed on the target electrical equipment to obtain multi-modal sensor deployment parameters; Based on the multi-modal sensor specification parameters and the multi-modal sensor deployment parameters, sensor network topology design is performed to construct the multi-modal test perception network.

3. A method of digitally testing an electrical device as recited in claim 2, wherein, Obtaining multi-modal sensor deployment parameters comprises: Based on the target electrical equipment, historical fault data of the electrical equipment is mined to obtain an electrical equipment historical fault data set; According to the multi-modal sensor specification parameters, the type of multi-modal sensor is determined; According to the type of multi-modal sensor, fault influence analysis is performed on the electrical equipment historical fault data set respectively to obtain a multi-modal equipment fault influence coefficient set; Based on the multi-modal equipment fault influence coefficient set, key position identification is performed on the target electrical equipment to obtain a multi-modal equipment key position set, and according to the multi-modal equipment key position set, multi-modal sensor deployment parameters are determined.

4. A method of digitally testing an electrical device as recited in claim 3, wherein, Through the digital test cloud, an electrical equipment fault prediction multi-channel is called, comprising: Performing equipment working condition clustering analysis on the electrical equipment historical fault data set to obtain an electrical equipment working condition clustering result; According to the electrical equipment working condition clustering result, the electrical equipment historical fault data set is classified and integrated to obtain a multi-working-condition electrical equipment fault data set; Based on the multi-working-condition electrical equipment fault data set, prediction training optimization is performed to construct a multi-working-condition electrical equipment fault predictor set; The multi-working-condition electrical equipment fault predictor set is identified in parallel to obtain an electrical equipment fault prediction multi-channel and store it in the digital test cloud.

5. A method of digitally testing an electrical device as recited in claim 4, wherein, Constructing a multi-working-condition electrical equipment fault predictor set comprises: Respectively performing fault feature extraction on the multi-working-condition electrical equipment fault data set to obtain a multi-working-condition electrical equipment fault feature fingerprint set; Constructing a fault test task list of the electrical equipment; According to the fault test task list, the multi-working-condition electrical equipment fault feature fingerprint set is subjected to fault sample identification, and a multi-working-condition equipment task fault sample set is obtained; Based on the multi-working-condition equipment task fault sample set, prediction training optimization is performed, and a multi-working-condition electrical equipment fault predictor set is constructed.

6. A method of digitally testing an electrical device as recited in claim 5, wherein, Based on the multi-working-condition equipment task fault sample set, prediction training optimization is performed, and a multi-working-condition electrical equipment fault predictor set is constructed, including: Deep neural network structure is used to respectively perform fault prediction training on the multi-working-condition equipment task fault sample set, and a multi-working-condition equipment task fault predictor set is obtained; The multi-working-condition equipment task fault predictor set is subjected to cross-validation and iterative optimization, and an available multi-working-condition equipment task fault predictor set is obtained; The available multi-working-condition equipment task fault predictor set is classified and integrated according to the equipment working condition type, and the multi-working-condition electrical equipment fault predictor set is constructed.

7. A method of digitally testing an electrical device as recited in claim 4, wherein, The target electrical equipment fault prediction channel is determined, including: The electrical equipment multi-modal operation data is subjected to similarity matching with the electrical equipment working condition clustering result, and a matching electrical equipment working condition is obtained; Based on the matching electrical equipment working condition, the electrical equipment fault prediction multi-channel is associated and mapped, and the target electrical equipment fault prediction channel is determined.

8. A method of digitally testing an electrical device as recited in claim 1, wherein, Feedback test response is performed through the electrical equipment fault test result, including: The target electrical equipment is subjected to feedback test verification through the electrical equipment fault test result, and an electrical equipment fault occurrence verification result is obtained; Based on the electrical equipment fault occurrence verification result, early warning strategy analysis is performed, an electrical equipment fault grading early warning strategy is determined, and fault early warning response is performed through the electrical equipment fault grading early warning strategy.

9. An electrical device digitizing test system, characterized by, The system is used to implement the electrical equipment digitized test method of any one of claims 1-8, and the system comprises: A data acquisition module: performing test demand analysis on a target electrical equipment, constructing a multi-modal test perception network, and acquiring electrical equipment multi-modal operation data through the multi-modal test perception network; A data transmission module: selecting a target communication mode according to the application scene and equipment characteristics of the target electrical equipment, and transmitting the electrical equipment multi-modal operation data to a digitized test cloud through the target communication mode; A channel matching module: calling an electrical equipment fault prediction multi-channel through the digitized test cloud, performing channel matching analysis on the electrical equipment multi-modal operation data based on the electrical equipment fault prediction multi-channel, and determining a target electrical equipment fault prediction channel; A fault identification module: performing fault prediction identification on the electrical equipment multi-modal operation data based on the target electrical equipment fault prediction channel, outputting an electrical equipment fault test result, and performing feedback test response through the electrical equipment fault test result.

10. An electronic device, comprising: The electronic device comprises: A memory for storing executable instructions; A processor for executing the executable instructions stored in the memory, and implementing the electrical equipment digitized test method of any one of claims 1-8.

Images