Fault diagnosis method for flexible load regulation and control terminal

Abstract

The invention relates to the technical field of power equipment operation and maintenance, and provides a fault diagnosis method for a flexible load regulation and control terminal, which comprises the following steps of: firstly, acquiring operation data of the terminal in real time, carrying out data preprocessing and feature extraction, and then, carrying out preliminary diagnosis on the data by utilizing a deep learning model; the model is continuously optimized by combining user feedback, if a preliminary diagnosis result does not conform to expectation, new training data are generated based on the user feedback and abnormal feature simulation, the model is retrained, and finally, the model is finely adjusted through a genetic algorithm until user requirements are met, so that the diagnosis accuracy is ensured. By introducing a real-time feedback mechanism and a simulation training technology, the timeliness and accuracy of fault diagnosis are improved, and the maintenance problem caused by diagnosis errors is effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of power equipment operation and maintenance, and in particular to a flexible load control terminal fault diagnosis method. Background Art

[0002] Flexible load control is a technology that flexibly adjusts the load demand in the power system by changing the load consumption pattern to achieve balance and stability of the power system. Flexible load control technology can make the power system more controllable and improve the reliability and stability of the system.

[0003] The core of flexible load regulation is load management and dispatching strategy, among which load priority dispatching refers to optimizing the dispatching load consumption according to the priority of the load to meet the needs of the power system. By flexibly adjusting the load consumption demand, flexible load regulation can improve the dispatchability of the load, enable the system to better cope with sudden changes in the load, and thus improve the load transfer capacity of the power system. At the same time, since the load consumption demand can be flexibly adjusted, the system can also return to normal operation faster, thereby reducing the impact of faults on the system and reducing the load recovery time.

[0004] Fault diagnosis of flexible load control terminals includes the following fault diagnosis process: First, collect specific information about the fault from the user or operator, including the time when the fault occurred, environmental conditions such as temperature and humidity, symptoms of the fault such as terminal unresponsiveness, data abnormalities, and any possible related details.

[0005] Physically inspect the flexible load control terminal to see if there is any obvious physical damage or abnormal phenomena, such as indicator light status, display information, abnormal sound or vibration, etc. Based on the collected information and observed phenomena, conduct a preliminary analysis to determine the possible causes of the fault, such as power supply problems, communication failures, software errors, or hardware damage.

[0006] After inspection, determine the specific cause of the flexible load control terminal failure, and take appropriate solutions based on the confirmed cause of the failure. This may include replacing damaged parts, fixing software errors, adjusting configuration parameters, etc. After resolving the failure, conduct a comprehensive test on the flexible load control terminal to ensure that it can work normally and the failure has been completely resolved.

[0007] In the process of performing fault diagnosis on the flexible load control terminal, there are the following technical pain points; the diagnosis of the flexible load control terminal is obtained by collecting data and analyzing and processing the data to obtain a fault diagnosis result. If the fault diagnosis result is wrong, the maintenance method corresponding to the fault diagnosis will not be able to complete the repair of the corresponding fault, resulting in the fault diagnosis result. The maintenance of the flexible load control terminal cannot play a helpful role. For this reason, the present invention provides a flexible load control terminal fault diagnosis method for secondary optimization of the fault diagnosis result of the flexible load control terminal, thereby improving the accuracy of the flexible load control terminal fault diagnosis. Summary of the invention

[0008] In view of the shortcomings of the prior art, the present invention provides a flexible load control terminal fault diagnosis method to solve the problem that if the fault diagnosis result is wrong, the maintenance method corresponding to the fault diagnosis will not be able to complete the repair of the corresponding fault, resulting in the fault diagnosis result The repair of the flexible load control terminal cannot play a helpful role.

[0009] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0010] The present invention provides a flexible load control terminal fault diagnosis method.

[0011] Step S101, obtaining operation data of the flexible load control terminal, the operation data of the flexible load control terminal including current data, voltage data, power factor data, temperature data and humidity data;

[0012] Step S102, preprocessing the operation data of the flexible load control terminal and performing feature extraction to obtain abnormal current waveform features and abnormal voltage waveform features;

[0013] Step S103, substituting the operation data of the real-time flexible load control terminal into the original flexible load control terminal fault diagnosis model to obtain the original flexible load control terminal fault diagnosis result, matching the original flexible load control terminal fault diagnosis result in the fault solution knowledge base to obtain the original fault diagnosis result solution, receiving the user's feedback information on the original fault diagnosis result solution, and if the user's feedback information on the original fault diagnosis result solution is that the solution is invalid, optimizing the original flexible load control terminal fault diagnosis model;

[0014] Step S104, collecting user feedback information on the original fault diagnosis result solution, re-acquiring the flexible load control terminal operation data, generating simulation training data according to the data simulation under the fault condition, using the collected user feedback information on the original fault diagnosis result solution, the abnormal current waveform characteristics and the abnormal voltage waveform characteristics, re-acquiring the flexible load control terminal operation data and the simulation training data to re-train the model, and obtain an updated flexible load control terminal fault diagnosis model;

[0015] Step S105, use the updated flexible load control terminal fault diagnosis model to analyze the operating data of the real-time flexible load control terminal, obtain the fault diagnosis result after the model is updated, and send the updated fault diagnosis result to the user. If the user feedbacks that the solution corresponding to the fault diagnosis result after the model is updated is invalid, the updated flexible load control terminal fault diagnosis model is optimized using a genetic algorithm until the user feedbacks that the solution is effective, and the optimization of the model is stopped, and the optimized flexible load control terminal fault diagnosis model is deployed and used.

[0016] Furthermore, the flexible load control terminal fault diagnosis method of the present invention, step S101 includes:

[0017] Setting a time period for obtaining the operation data of the flexible load control terminal, obtaining data according to the time period, and classifying and storing the obtained operation data of the flexible load control terminal;

[0018] Create data labels for the classified and stored data and build a flexible load control terminal database;

[0019] The data in the flexible load control terminal database is randomly screened to obtain a randomly sampled data group, and the data in the randomly sampled data group is compared with the corresponding historical data of the flexible load control terminal database. If the comparison results are consistent, the data in the flexible load control terminal database has not been tampered with, and the data in the flexible load control terminal database can be used.

[0020] Furthermore, in the flexible load control terminal fault diagnosis method of the present invention, step S102 includes:

[0021] Standardize the operation data of the flexible load control terminal to remove invalid data and duplicate data in the operation data of the flexible load control terminal;

[0022] The operating data features of the flexible load control terminal are extracted after preprocessing to obtain current features and voltage features.

[0023] Furthermore, in the flexible load control terminal fault diagnosis method of the present invention, step S102 includes:

[0024] The peak value, valley value and average value of the current characteristic and voltage characteristic waveforms are calculated through waveform analysis, and the waveform anomalies of the current characteristic and voltage characteristic are analyzed to obtain the abnormal current waveform characteristics and the abnormal voltage waveform characteristics.

[0025] Furthermore, the flexible load control terminal fault diagnosis method of the present invention, step S103 includes:

[0026] Obtain fault solutions, perform cluster analysis on the fault solutions, obtain cluster analysis results of the fault solutions, and establish corresponding fault types according to the cluster analysis results of the fault solutions;

[0027] Establish a knowledge base for fault types and fault solutions, and establish a retrieval mechanism for the knowledge base;

[0028] After receiving the fault diagnosis result of the original flexible load control terminal, a search algorithm is used to find a matching solution in the fault solution knowledge base.

[0029] Furthermore, the flexible load control terminal fault diagnosis method of the present invention, step S103 includes:

[0030] After receiving the feedback information of the user's original fault diagnosis result solution, the received feedback information of the user's original fault diagnosis result solution is parsed to extract keywords in the feedback information of the user's original fault diagnosis result solution;

[0031] If the keywords in the feedback information of the user's original fault diagnosis result solution include invalid, unusable or cannot be used, the feedback information of the user's original fault diagnosis result solution is marked as invalid fault solution.

[0032] Furthermore, the flexible load control terminal fault diagnosis method of the present invention, step S104 includes:

[0033] Perform feature analysis on the data under the fault condition, and perform cluster analysis on the data under the fault condition, to obtain the data features under the fault condition and the data cluster analysis results under the fault condition;

[0034] Determine the adjustment range of data values ​​of data features under fault conditions and the adjustment range of data cluster analysis results under fault conditions;

[0035] The data is adjusted according to the data value adjustment range of the data feature under the fault condition and the data cluster analysis result adjustment range under the fault condition to obtain the adjusted data under the fault condition and the data set under the fault condition;

[0036] The adjusted data under the fault condition and the data set under the fault condition are used as simulation training data.

[0037] Furthermore, the flexible load control terminal fault diagnosis method of the present invention, step S104 includes:

[0038] Integrate the user's feedback on the original fault diagnosis result solution, the abnormal current waveform characteristics and voltage waveform characteristics, the re-acquired flexible load control terminal operation data and the simulation training data into the data set;

[0039] After data cleaning and feature extraction of the data in the dataset, a deep learning algorithm is selected, and the integrated dataset is used to retrain the flexible load control terminal fault diagnosis model to obtain an updated flexible load control terminal fault diagnosis model.

[0040] Furthermore, the flexible load control terminal fault diagnosis method of the present invention, step S105 includes:

[0041] Set the initial genetic algorithm parameters, use the parameters of the flexible load control terminal fault diagnosis model as individuals of the genetic algorithm, and generate the initial population for the individuals of the genetic algorithm based on the parameters of the flexible load control terminal fault diagnosis model:

[0042] The fitness function is used to evaluate the fitness of the parameters of each flexible load control terminal fault diagnosis model, and excellent individuals are selected to enter the next generation according to the fitness of the parameters of each flexible load control terminal fault diagnosis model;

[0043] The parameters of the selected flexible load control terminal fault diagnosis model are randomly paired and cross-operated to generate new parameters of the flexible load control terminal fault diagnosis model.

[0044] Furthermore, the flexible load control terminal fault diagnosis method of the present invention, step S105 includes:

[0045] After each iterative optimization, the flexible load control terminal fault diagnosis model is reconfigured using model parameters corresponding to the parameters of the current optimal flexible load control terminal fault diagnosis model;

[0046] Diagnose the operation data of the real-time flexible load control terminal and send the diagnosis results and corresponding solutions to the user;

[0047] Collect user feedback on the solution. If the feedback is invalid, continue to optimize. If the feedback is effective, stop optimizing.

[0048] When the user feedback solution is effective or the termination condition is reached, the optimization process of the genetic algorithm is stopped;

[0049] The model parameters finally optimized are configured into the flexible load control terminal fault diagnosis model.

[0050] Beneficial effects of the present invention:

[0051] First, by introducing a real-time user feedback mechanism and utilizing deep learning and genetic algorithms, the present invention can significantly improve the accuracy of fault diagnosis. The model can self-adjust and learn to adapt to changing fault conditions, thereby ensuring the accuracy of the diagnosis results.

[0052] Secondly, the present invention collects the operating data of the flexible load control terminal under fault conditions and simulates and generates simulation training data close to the actual fault conditions, thereby enhancing the adaptability and generalization ability of the model. At the same time, through feature extraction and waveform analysis, the abnormal waveform characteristics of current and voltage are accurately captured, further improving the model's ability to identify fault types.

[0053] In addition, the present invention provides a complete set of flexible load control terminal fault diagnosis process, and the systematic diagnosis method can significantly improve the operation and maintenance efficiency and reduce the fault resolution time. At the same time, accurate fault diagnosis and effective solutions can reduce the extra cost and time waste caused by misdiagnosis or wrong repair, and reduce the overall operation and maintenance cost.

[0054] Finally, through real-time monitoring and accurate diagnosis of flexible load control terminals, the present invention can timely discover and solve potential faults, thereby enhancing the reliability and stability of the power system. The construction of a fault solution knowledge base enables operation and maintenance personnel to quickly find and apply effective solutions.

[0055] In summary, the present invention has shown significant beneficial effects in improving fault diagnosis accuracy, enhancing model adaptability, improving operation and maintenance efficiency, reducing operation and maintenance costs, and enhancing system reliability and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0057] Figure 1 A schematic diagram of a flexible load control terminal fault diagnosis method provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. The technical solutions provided by the embodiments of the present invention are described in detail below in conjunction with the drawings.

[0059] In order to better understand the purpose of the present invention, the present invention is described in further detail below.

[0060] The present invention provides a flexible load control terminal fault diagnosis method.

[0061] Step S101, obtaining operation data of the flexible load control terminal, the operation data of the flexible load control terminal including current data, voltage data, power factor data, temperature data and humidity data;

[0062] Current data and voltage data:

[0063] Collection method: Use the built-in sensors of the flexible load control terminal or external measuring devices to monitor and record the changes of current and voltage in real time.

[0064] Note: Ensure the accuracy and stability of the sensor and calibrate it regularly to avoid data deviation. At the same time, consider the impact of factors such as electromagnetic interference on data collection.

[0065] Application in fault diagnosis: Abnormal fluctuations in current and voltage are often early signs of equipment failure. By analyzing these data, potential problems can be discovered in a timely manner.

[0066] Power factor data:

[0067] Collection method: Use a professional power quality analyzer or related equipment to measure the power factor, which reflects the efficiency of electric energy use.

[0068] Note: Power factor measurement requires accurate time synchronization to ensure data comparability.

[0069] Application in fault diagnosis: Abnormal power factor may indicate abnormal operation of equipment or reduced energy efficiency.

[0070] Temperature data and humidity data:

[0071] Collection method: Use temperature and humidity sensors to monitor the temperature and humidity of the environment where the flexible load control terminal is located.

[0072] Note: The sensor should be placed away from direct sunlight or other heat sources to ensure data accuracy. At the same time, the working status of the sensor should be checked regularly.

[0073] Application in fault diagnosis: Changes in temperature and humidity may affect the performance and life of equipment. Analysis of these data can help prevent failures caused by environmental factors.

[0074] These data are not only crucial for fault diagnosis, but also the basis for subsequent model training. By analyzing these data, we can extract features related to the fault, and then build and optimize the fault diagnosis model. At the same time, the richness of these data also provides more information for model training, which helps to improve the accuracy of diagnosis.

[0075] Step S102, preprocessing the operation data of the flexible load control terminal and performing feature extraction to obtain abnormal current waveform features and abnormal voltage waveform features;

[0076] First, the operation data of the flexible load control terminal is processed to remove invalid and duplicate data. The purpose of this step is to clean the data and ensure the data quality for subsequent analysis.

[0077] The pre-processed data is deeply analyzed to extract key current and voltage features, which will serve as important inputs for the fault diagnosis model.

[0078] The peak value, valley value and average value of the current and voltage characteristic waveforms are calculated through waveform analysis technology. These parameters help to more fully understand the characteristics of the waveform.

[0079] According to the results of waveform analysis, the abnormal characteristics of the current waveform and voltage waveform are identified. The abnormal characteristics of the current waveform and voltage waveform will play a role in the subsequent fault diagnosis process and provide strong support for accurate fault diagnosis.

[0080] Step S103, substituting the operation data of the real-time flexible load control terminal into the original flexible load control terminal fault diagnosis model to obtain the original flexible load control terminal fault diagnosis result, matching the original flexible load control terminal fault diagnosis result in the fault solution knowledge base to obtain the original fault diagnosis result solution, receiving the user's feedback information on the original fault diagnosis result solution, and if the user's feedback information on the original fault diagnosis result solution is that the solution is invalid, optimizing the original flexible load control terminal fault diagnosis model;

[0081] First, the operation data of the real-time flexible load control terminal is input into the original fault diagnosis model. These data include key parameters such as current, voltage, power factor, temperature and humidity, which provide the model with the information needed for fault diagnosis.

[0082] The original fault diagnosis model analyzes the input real-time data and gives a preliminary fault diagnosis result, which points out the possible fault type or problem.

[0083] This preliminary diagnosis result is matched in the fault solution knowledge base, which is a database containing various fault types and their corresponding solutions. Through matching, the solution corresponding to the diagnosis result can be found. The found solution is provided to the user and the user's feedback is awaited. User feedback is crucial to evaluating the effectiveness of the solution and is also the key basis for subsequent optimization of the model.

[0084] If user feedback indicates that the provided solution is ineffective, the system will recognize the shortcomings of the current fault diagnosis model and trigger the model optimization process. Optimization may include adjusting model parameters, improving algorithms, or adding more training data to improve the accuracy and reliability of the model. User feedback information is also parsed, especially extracting keywords to determine whether the user thinks the solution is effective. If the keywords indicate that the solution is ineffective, the system will automatically optimize the model.

[0085] The fault solution knowledge base is the core component of this step. The fault solution knowledge base establishes different types of faults and their corresponding solutions through cluster analysis, and has a retrieval mechanism to quickly find matching solutions.

[0086] Step S104, collecting user feedback information on the original fault diagnosis result solution, re-acquiring the flexible load control terminal operation data, generating simulation training data according to the data simulation under the fault condition, using the collected user feedback information on the original fault diagnosis result solution, the abnormal current waveform characteristics and the abnormal voltage waveform characteristics, re-acquiring the flexible load control terminal operation data and the simulation training data to re-train the model, and obtain an updated flexible load control terminal fault diagnosis model;

[0087] First, the system collects user feedback on the original fault diagnosis result solution. Feedback is an important reference for optimizing the model and can help us understand the performance of the model in actual applications. After collecting user feedback, the latest operating data is obtained from the flexible load control terminal again. The data reflects the latest status of the terminal and is crucial for model training.

[0088] In order to enhance the generalization ability of the model, the system will generate training data based on data simulation under fault conditions. These training data are generated through methods such as feature analysis and cluster analysis, which can help the model better learn the characteristics of fault conditions.

[0089] The collected user feedback information, abnormal current waveform characteristics, abnormal voltage waveform characteristics, re-acquired operation data and simulation training data are integrated. The model can take into account actual fault data, user feedback and simulation data at the same time during the training process, thereby improving the diagnostic capabilities more comprehensively.

[0090] After integrating the data, the system will perform preprocessing such as data cleaning and feature extraction to ensure the accuracy and validity of the data. Using the new integrated data set, the system will retrain the flexible load control terminal fault diagnosis model. Through deep learning algorithms, the model will self-adjust and optimize based on the new data features to improve the accuracy of diagnosis. After training is completed, the system will evaluate the updated model to ensure that its performance meets expectations. This usually includes using a validation data set to test the accuracy and reliability of the model.

[0091] Step S105, use the updated flexible load control terminal fault diagnosis model to analyze the operating data of the real-time flexible load control terminal, obtain the fault diagnosis result after the model is updated, and send the updated fault diagnosis result to the user. If the user feedbacks that the solution corresponding to the fault diagnosis result after the model is updated is invalid, the updated flexible load control terminal fault diagnosis model is optimized using a genetic algorithm until the user feedbacks that the solution is effective, and the optimization of the model is stopped, and the optimized flexible load control terminal fault diagnosis model is deployed and used.

[0092] Initialize genetic algorithm parameters: This step is to set the starting point of the genetic algorithm, involving the setting of parameters such as population size and number of iterations, and using the parameters of the fault diagnosis model as individuals of the genetic algorithm.

[0093] Generate initial population: Generate an initial population based on the previously set parameters. The "population" here is composed of multiple "individuals", each of which represents a possible combination of a set of model parameters.

[0094] Evaluate fitness: Next, you need to evaluate the fitness of each "individual". This is usually done by observing the performance of the model on the validation dataset, such as how high the model's accuracy is.

[0095] Selection operation: Based on the results of fitness evaluation, select the "individuals" with good performance to enter the next generation. This is like in natural selection, organisms that adapt to the environment are more likely to survive.

[0096] Crossover operation: The selected “individuals” will be paired and some parameters will be exchanged to generate new “individuals”. This helps the model explore more possible parameter combinations.

[0097] Mutation operation: In order to increase the diversity of the population, the parameters of the "individual" are adjusted randomly in small amounts. This is like the genetic mutation that may occur in the process of reproduction of organisms.

[0098] Iterative optimization: By repeating the process of selection, crossover, and mutation, the model parameters are continuously optimized until the stopping condition is met.

[0099] Model evaluation and feedback: After each iteration, the current optimal model parameters are used to diagnose the real-time data and send the results to the user. At the same time, user feedback on the solution is collected to evaluate the effectiveness of the current model.

[0100] Stop condition: If the user feedback solution is effective, or the number of iterations reaches the preset upper limit, the optimization process is stopped.

[0101] Model deployment: Finally, the optimized model parameters are configured into the flexible load control terminal fault diagnosis model and formally deployed for use.

[0102] Specifically, the flexible load control terminal fault diagnosis method of the present invention, step S101 includes:

[0103] Setting a time period for obtaining the operation data of the flexible load control terminal, obtaining data according to the time period, and classifying and storing the obtained operation data of the flexible load control terminal;

[0104] Create data labels for the classified and stored data and build a flexible load control terminal database;

[0105] The data in the flexible load control terminal database is randomly screened to obtain a randomly sampled data group, and the data in the randomly sampled data group is compared with the corresponding historical data of the flexible load control terminal database. If the comparison results are consistent, the data in the flexible load control terminal database has not been tampered with, and the data in the flexible load control terminal database can be used.

[0106] Specifically, the flexible load control terminal fault diagnosis method of the present invention, step S102 includes:

[0107] Standardize the operation data of the flexible load control terminal to remove invalid data and duplicate data in the operation data of the flexible load control terminal;

[0108] The operating data features of the flexible load control terminal are extracted after preprocessing to obtain current features and voltage features.

[0109] Specifically, the flexible load control terminal fault diagnosis method of the present invention, step S102 includes:

[0110] The peak value, valley value and average value of the current characteristic and voltage characteristic waveforms are calculated through waveform analysis, and the waveform anomalies of the current characteristic and voltage characteristic are analyzed to obtain the abnormal current waveform characteristics and the abnormal voltage waveform characteristics.

[0111] Specifically, the flexible load control terminal fault diagnosis method of the present invention, step S103 includes:

[0112] Obtain fault solutions, perform cluster analysis on the fault solutions, obtain cluster analysis results of the fault solutions, and establish corresponding fault types according to the cluster analysis results of the fault solutions;

[0113] Establish a knowledge base for fault types and fault solutions, and establish a retrieval mechanism for the knowledge base;

[0114] After receiving the fault diagnosis result of the original flexible load control terminal, a search algorithm is used to find a matching solution in the fault solution knowledge base.

[0115] Specifically, the flexible load control terminal fault diagnosis method of the present invention, step S103 includes:

[0116] After receiving the feedback information of the user's original fault diagnosis result solution, the received feedback information of the user's original fault diagnosis result solution is parsed to extract keywords in the feedback information of the user's original fault diagnosis result solution;

[0117] If the keywords in the feedback information of the user's original fault diagnosis result solution include invalid, unusable or cannot be used, the feedback information of the user's original fault diagnosis result solution is marked as invalid fault solution.

[0118] Specifically, the flexible load control terminal fault diagnosis method of the present invention, step S104 includes:

[0119] Perform feature analysis on the data under the fault condition, and perform cluster analysis on the data under the fault condition, to obtain the data features under the fault condition and the data cluster analysis results under the fault condition;

[0120] Determine the adjustment range of data values ​​of data features under fault conditions and the adjustment range of data cluster analysis results under fault conditions;

[0121] The data is adjusted according to the data value adjustment range of the data feature under the fault condition and the data cluster analysis result adjustment range under the fault condition to obtain the adjusted data under the fault condition and the data set under the fault condition;

[0122] The adjusted data under the fault condition and the data set under the fault condition are used as simulation training data.

[0123] Specifically, the flexible load control terminal fault diagnosis method of the present invention, step S104 includes:

[0124] Integrate the user's feedback on the original fault diagnosis result solution, the abnormal current waveform characteristics and voltage waveform characteristics, the re-acquired flexible load control terminal operation data and the simulation training data into the data set;

[0125] After data cleaning and feature extraction of the data in the dataset, a deep learning algorithm is selected, and the integrated dataset is used to retrain the flexible load control terminal fault diagnosis model to obtain an updated flexible load control terminal fault diagnosis model.

[0126] Specifically, the flexible load control terminal fault diagnosis method of the present invention, step S105 includes:

[0127] Set the initial genetic algorithm parameters, use the parameters of the flexible load control terminal fault diagnosis model as individuals of the genetic algorithm, and generate the initial population for the individuals of the genetic algorithm based on the parameters of the flexible load control terminal fault diagnosis model:

[0128] The fitness function is used to evaluate the fitness of the parameters of each flexible load control terminal fault diagnosis model, and excellent individuals are selected to enter the next generation according to the fitness of the parameters of each flexible load control terminal fault diagnosis model;

[0129] The parameters of the selected flexible load control terminal fault diagnosis model are randomly paired and cross-operated to generate new parameters of the flexible load control terminal fault diagnosis model.

[0130] Specifically, the flexible load control terminal fault diagnosis method of the present invention, step S105 includes:

[0131] After each iterative optimization, the flexible load control terminal fault diagnosis model is reconfigured using model parameters corresponding to the parameters of the current optimal flexible load control terminal fault diagnosis model;

[0132] Diagnose the operation data of the real-time flexible load control terminal and send the diagnosis results and corresponding solutions to the user;

[0133] Collect user feedback on the solution. If the feedback is invalid, continue to optimize. If the feedback is effective, stop optimizing.

[0134] When the user feedback solution is effective or the termination condition is reached, the optimization process of the genetic algorithm is stopped;

[0135] The model parameters finally optimized are configured into the flexible load control terminal fault diagnosis model.

[0136] The present invention effectively solves the maintenance problem of flexible load control terminal which may be caused by the wrong fault diagnosis result. Specifically:

[0137] First, the present invention introduces a real-time feedback and model optimization mechanism. When the user feedback solution is invalid, the model optimization process will be triggered to ensure the accuracy and timeliness of the diagnostic model.

[0138] Secondly, by establishing a user feedback mechanism, the present invention can timely understand the accuracy of the diagnosis results and adjust the diagnosis model according to user feedback, thereby improving the accuracy of subsequent diagnosis.

[0139] In addition, the present invention also uses simulated training data generation technology to generate training data close to the actual fault situation based on the data simulation under the fault situation, so as to improve the diagnostic ability of the model in complex scenarios.

[0140] At the same time, the present invention adopts deep learning and model updating strategies to retrain the model by integrating user feedback, abnormal characteristics, re-acquired operating data and simulated training data, thereby further improving the diagnostic accuracy.

[0141] In terms of model optimization, the present invention also innovatively applies a genetic algorithm. When the diagnostic result after the updated model still does not meet the requirements, the genetic algorithm will further optimize the model until an effective solution is found.

[0142] In addition to the above technical solutions, the present invention also focuses on continuous optimization and deployment, database and data verification, feature extraction and anomaly detection, and the construction of a fault solution knowledge base. These measures together constitute a closed-loop system from data collection, model diagnosis, user feedback to model optimization, enabling the present invention to self-learn and improve, and continuously improve diagnostic accuracy.

[0143] In summary, through the comprehensive application of these technical solutions, the present invention effectively solves the maintenance problem caused by erroneous fault diagnosis results and significantly improves the operation and maintenance efficiency and reliability of the flexible load control terminal.

[0144] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations, and the above-described embodiments of the present invention do not constitute a limitation on the scope of protection of the present invention.

Claims

1. A flexible load control terminal fault diagnosis method, characterized in that: include: Step S101, obtaining operation data of the flexible load control terminal, the operation data of the flexible load control terminal including current data, voltage data, power factor data, temperature data and humidity data; Step S102, preprocessing the operation data of the flexible load control terminal and performing feature extraction to obtain abnormal current waveform features and abnormal voltage waveform features; Step S103, substituting the operation data of the real-time flexible load control terminal into the original flexible load control terminal fault diagnosis model to obtain the original flexible load control terminal fault diagnosis result, matching the original flexible load control terminal fault diagnosis result in the fault solution knowledge base to obtain the original fault diagnosis result solution, receiving the user's feedback information on the original fault diagnosis result solution, and if the user's feedback information on the original fault diagnosis result solution is that the solution is invalid, optimizing the original flexible load control terminal fault diagnosis model; Step S104, collecting user feedback information on the original fault diagnosis result solution, re-acquiring the flexible load control terminal operation data, generating simulation training data according to the data simulation under the fault condition, using the collected user feedback information on the original fault diagnosis result solution, the abnormal current waveform characteristics and the abnormal voltage waveform characteristics, re-acquiring the flexible load control terminal operation data and the simulation training data to re-train the model, and obtain an updated flexible load control terminal fault diagnosis model; Step S105, use the updated flexible load control terminal fault diagnosis model to analyze the operating data of the real-time flexible load control terminal, obtain the fault diagnosis result after the model is updated, and send the updated fault diagnosis result to the user. If the user feedbacks that the solution corresponding to the fault diagnosis result after the model is updated is invalid, the updated flexible load control terminal fault diagnosis model is optimized using a genetic algorithm until the user feedbacks that the solution is effective, and the optimization of the model is stopped, and the optimized flexible load control terminal fault diagnosis model is deployed and used.

2. The flexible load control terminal fault diagnosis method according to claim 1, characterized in that: The step S101 includes: Setting a time period for obtaining the operation data of the flexible load control terminal, obtaining data according to the time period, and classifying and storing the obtained operation data of the flexible load control terminal; Create data labels for the classified and stored data and build a flexible load control terminal database; The data in the flexible load control terminal database is randomly screened to obtain a randomly sampled data group, and the data in the randomly sampled data group is compared with the corresponding historical data of the flexible load control terminal database. If the comparison results are consistent, the data in the flexible load control terminal database has not been tampered with, and the data in the flexible load control terminal database can be used.

3. The flexible load control terminal fault diagnosis method according to claim 1, characterized in that: The step S102 includes: Standardize the operation data of the flexible load control terminal to remove invalid data and duplicate data in the operation data of the flexible load control terminal; The operating data features of the flexible load control terminal are extracted after preprocessing to obtain current features and voltage features.

4. The flexible load control terminal fault diagnosis method according to claim 3, characterized in that: The step S102 includes: The peak value, valley value and average value of the current characteristic and voltage characteristic waveforms are calculated through waveform analysis, and the waveform anomalies of the current characteristic and voltage characteristic are analyzed to obtain the abnormal current waveform characteristics and the abnormal voltage waveform characteristics.

5. The flexible load control terminal fault diagnosis method according to claim 1, characterized in that: The step S103 includes: Obtain fault solutions, perform cluster analysis on the fault solutions, obtain cluster analysis results of the fault solutions, and establish corresponding fault types according to the cluster analysis results of the fault solutions; Establish a knowledge base for fault types and fault solutions, and establish a retrieval mechanism for the knowledge base; After receiving the fault diagnosis result of the original flexible load control terminal, a search algorithm is used to find a matching solution in the fault solution knowledge base.

6. The flexible load control terminal fault diagnosis method according to claim 5, characterized in that: The step S103 includes: After receiving the feedback information of the user's original fault diagnosis result solution, the received feedback information of the user's original fault diagnosis result solution is parsed to extract keywords in the feedback information of the user's original fault diagnosis result solution; If the keywords in the feedback information of the user's original fault diagnosis result solution include invalid, unusable or cannot be used, the feedback information of the user's original fault diagnosis result solution is marked as invalid fault solution.

7. The flexible load control terminal fault diagnosis method according to claim 1, characterized in that: The step S104 includes: Perform feature analysis on the data under the fault condition, and perform cluster analysis on the data under the fault condition, to obtain the data features under the fault condition and the data cluster analysis results under the fault condition; Determine the adjustment range of data values ​​of data features under fault conditions and the adjustment range of data cluster analysis results under fault conditions; The data is adjusted according to the data value adjustment range of the data feature under the fault condition and the data cluster analysis result adjustment range under the fault condition to obtain the adjusted data under the fault condition and the data set under the fault condition; The adjusted data under the fault condition and the data set under the fault condition are used as simulation training data.

8. The flexible load control terminal fault diagnosis method according to claim 7, characterized in that: The step S104 includes: Integrate the user's feedback on the original fault diagnosis result solution, the abnormal current waveform characteristics and voltage waveform characteristics, the re-acquired flexible load control terminal operation data and the simulation training data into the data set; After data cleaning and feature extraction of the data in the dataset, a deep learning algorithm is selected, and the integrated dataset is used to retrain the flexible load control terminal fault diagnosis model to obtain an updated flexible load control terminal fault diagnosis model.

9. The flexible load control terminal fault diagnosis method according to claim 1, characterized in that: The step S105 includes: Set the initial genetic algorithm parameters, use the parameters of the flexible load control terminal fault diagnosis model as individuals of the genetic algorithm, and generate the initial population for the individuals of the genetic algorithm based on the parameters of the flexible load control terminal fault diagnosis model: The fitness function is used to evaluate the fitness of the parameters of each flexible load control terminal fault diagnosis model, and excellent individuals are selected to enter the next generation according to the fitness of the parameters of each flexible load control terminal fault diagnosis model; The parameters of the selected flexible load control terminal fault diagnosis model are randomly paired and cross-operated to generate new parameters of the flexible load control terminal fault diagnosis model.

10. The flexible load control terminal fault diagnosis method according to claim 9, characterized in that: The step S105 includes: After each iterative optimization, the flexible load control terminal fault diagnosis model is reconfigured using model parameters corresponding to the parameters of the current optimal flexible load control terminal fault diagnosis model; Diagnose the operation data of the real-time flexible load control terminal and send the diagnosis results and corresponding solutions to the user; Collect user feedback on the solution. If the feedback is invalid, continue to optimize. If the feedback is effective, stop optimizing. When the user feedback solution is effective or the termination condition is reached, the optimization process of the genetic algorithm is stopped; The model parameters finally optimized are configured into the flexible load control terminal fault diagnosis model.

Images