A test monitoring method and system for a combined transformer test power supply

Abstract

This invention discloses a test monitoring method and system for a combined transformer test power supply, comprising: testing and monitoring the target test power supply to obtain power supply test monitoring information; preprocessing the power supply test monitoring information to obtain preprocessing result information; constructing an anomaly detection model, inputting the preprocessing result information to detect abnormal operating data occurring during power supply testing, and obtaining anomaly detection information; performing trend analysis based on the anomaly detection information to predict future changes in power supply test monitoring data and detect abnormal trends, and obtaining anomaly trend detection information; extracting anomaly features based on the anomaly trend detection information, pre-setting risk assessment rules, and performing risk assessment based on the extracted anomaly features, and obtaining risk assessment information; determining the existence of test safety risks based on the risk assessment information, and if so, using a fuzzy controller to generate a control scheme for control to ensure the safety and reliability of power supply testing.

Description

Technical Field

[0001] This invention relates to the field of power supply testing and monitoring technology, and in particular to a testing and monitoring method and system for a combined transformer test power supply. Background Technology

[0002] With the development of power systems, the demand for transformer power supplies has increased due to the needs of different scenarios. A single power supply is insufficient to support these needs, necessitating the use of multiple power supplies in combination—that is, combined transformer power supplies. By employing different combination methods, the operating efficiency of the power grid can be improved. However, combining power supplies requires further enhancements in the monitoring and control capabilities of power supply operation to ensure safety. Before being put into use, the corresponding combination methods need to be tested to ensure stability and reliability.

[0003] Because the testing requirements for transformer power supplies vary, comprehensive testing is required under different frequencies and load conditions to verify their operating performance under different operating conditions. The variable testing conditions and environment can bring unknown testing risks. Therefore, how to accurately respond to and control testing risks during power supply testing is an urgent problem to be solved. Summary of the Invention

[0004] This invention overcomes the shortcomings of the prior art and provides a test monitoring method and system for a combined transformer test power supply, the important purpose of which is to improve the safety and reliability of power supply testing.

[0005] To achieve the above objectives, the first aspect of the present invention provides a test monitoring method for a combined transformer test power supply, comprising:

[0006] The target test power supply is tested and monitored to obtain power supply test and monitoring information. The power supply test and monitoring information is preprocessed to obtain preprocessed result information.

[0007] An anomaly detection model is constructed, and the preprocessed result information is input to detect abnormal operating data that occurs during power supply testing, thereby obtaining anomaly detection information.

[0008] Based on the anomaly detection information, trend analysis is performed to predict changes in power supply test monitoring data at future times, and abnormal trends are detected to obtain anomaly trend detection information.

[0009] Based on the abnormal trend detection information, abnormal features are extracted, risk assessment rules are preset, and risk assessment is performed according to the extracted abnormal features to obtain risk assessment information.

[0010] If a test safety risk is identified based on the risk assessment information, a control scheme is generated using a fuzzy controller to control the test power supply of the target combined transformer.

[0011] In this solution, the step of testing and monitoring the target test power supply to obtain power supply test monitoring information, and preprocessing the power supply test monitoring information to obtain preprocessed result information, specifically includes:

[0012] A sensor array is set up in the test power supply of the target combined transformer, and power supply test monitoring is performed based on the set sensor array to obtain power supply test monitoring information;

[0013] The power supply test and monitoring information is processed by using a low-pass filter to remove high-frequency signal noise, and the power supply test and monitoring information data after signal filtering is calibrated.

[0014] A standard reference signal is set, and the power supply test monitoring information after signal filtering is compared with the standard reference signal. The sensor response is calibrated according to the comparison result, and the calibrated power supply test monitoring information is obtained.

[0015] The calibrated power supply test monitoring information is subjected to time-series processing. The standard deviation of the time-series processed power supply test monitoring information is calculated and compared with the preset standard deviation range.

[0016] Data that are outside the preset standard deviation range are defined as outliers and removed. Linear interpolation is then used for interpolation compensation to obtain the preprocessed results.

[0017] In this solution, the construction of the anomaly detection model, which involves inputting the preprocessed result information to detect operational anomalies during power supply testing, and obtaining anomaly detection information, specifically includes:

[0018] An anomaly detection model is constructed based on isolated forest, preprocessing result information is obtained and a time series is generated, which is then input into the anomaly detection model to detect abnormal operating data that occurs during power supply testing.

[0019] The RANSAC algorithm is introduced, which randomly selects several sample data from the input time series, performs linear fitting on the selected sample data, and calculates the difference between the actual observed value and the fitted predicted value based on the linear fitting to generate residual features.

[0020] Feature extraction is performed on the input time series, and a new feature matrix is ​​constructed by combining the residual features. The total number of trees in the isolated forest is determined based on the new feature matrix, and a feature space is constructed.

[0021] A random binary search tree is constructed to detect low-density and isolated data points in the feature space, which are defined as isolated points. The path length of each isolated point in the corresponding tree is calculated as an anomaly score, and compared with a preset threshold to obtain anomaly detection information.

[0022] In this solution, the step of performing trend analysis based on the anomaly detection information to predict future changes in power supply test monitoring data and detect abnormal trends to obtain anomaly trend detection information specifically includes:

[0023] A trend prediction model is constructed using a Markov algorithm. The anomaly detection information is input into the trend prediction model for analysis, and the original sequence is generated based on the input anomaly detection information.

[0024] The original sequence is subjected to feature extraction to form a feature matrix, and the similarity is calculated with the feature matrix corresponding to each state transition matrix in the preset state space to obtain a similarity value.

[0025] The calculated similarity value is compared with a preset threshold, and the corresponding state transition matrix is ​​obtained based on the comparison result. The state transition probability at the next moment is calculated by combining the feature matrix of the original sequence.

[0026] The calculated state transition probabilities for the next time step are sorted, and the state with the highest transition probability is selected as the prediction result to generate a prediction sequence, thus obtaining trend prediction result information.

[0027] The trend prediction results are input into the anomaly detection model for detection. Based on the detection results, abnormal data at future times are marked to obtain anomaly trend detection information and generate anomaly trend detection map.

[0028] In this solution, the step of extracting abnormal features based on the abnormal trend detection information, setting up risk assessment rules, and performing risk assessment based on the extracted abnormal features to obtain risk assessment information specifically includes:

[0029] Obtain abnormal trend detection information, and extract abnormal data features and corresponding time features based on the abnormal trend detection information to obtain abnormal feature information;

[0030] Risk assessment is performed based on the abnormal feature information, the abnormal feature information is calculated with preset benchmark parameters to obtain the degree of abnormal deviation, and the corresponding time features are correlated with the degree of abnormal deviation based on the abnormal feature information.

[0031] A preset risk assessment rule is established, and the calculated degree of abnormal deviation is compared with the risk assessment rule to assess the test risks existing in the current power supply test and obtain risk assessment information.

[0032] In this scheme, the step of determining the existence of test safety risks based on the risk assessment information, and if such risks exist, using a fuzzy controller to generate a control scheme to regulate the test power supply of the target combined transformer, specifically includes:

[0033] Obtain risk assessment information, preset test security thresholds, compare the risk assessment information with the test security thresholds, and analyze whether there are any test security risks.

[0034] If the judgment result indicates that there is a test safety risk in the current power supply test, then the fuzzy controller is used to generate a control scheme to control the test power supply of the target combined transformer.

[0035] Obtain abnormal feature information and generate control targets based on the abnormal feature information; set input variables based on the degree of abnormal deviation; perform fuzzification processing on the input variables and map them into fuzzy sets; and construct membership functions based on the fuzzy sets.

[0036] Based on big data retrieval, control examples of different types of power supply test anomalies are obtained. Fuzzy rules for control measures of different types of power supply test anomalies are established by combining expert analysis method, and a fuzzy rule base is formed.

[0037] Using the set input variables, fuzzy rule matching is performed in the fuzzy rule base to obtain matching rules. The activation degree of each matching rule is calculated based on the Min-Max inference method. The outputs of multiple rules are combined through Max operation to obtain a fuzzy output set.

[0038] The fuzzy output set is solved by the centroid method, and the corresponding membership function is obtained based on the fuzzy output set. The membership function is then updated based on the activation degree of each output fuzzy set.

[0039] The centroids of each fuzzy output set are calculated based on the updated membership function and then weighted and averaged. A control output value is generated based on the weighted average result, and a control scheme is constructed based on the control output value and input to the actuator for risk control.

[0040] A second aspect of the present invention provides a test monitoring system for a combined transformer test power supply. The system includes a memory and a processor. The memory contains a test monitoring method program for the combined transformer test power supply. When the processor executes the test monitoring method program for the combined transformer test power supply, it performs the following steps:

[0041] The target test power supply is tested and monitored to obtain power supply test and monitoring information. The power supply test and monitoring information is preprocessed to obtain preprocessed result information.

[0042] An anomaly detection model is constructed, and the preprocessed result information is input to detect abnormal operating data that occurs during power supply testing, thereby obtaining anomaly detection information.

[0043] Based on the anomaly detection information, trend analysis is performed to predict changes in power supply test monitoring data at future times, and abnormal trends are detected to obtain anomaly trend detection information.

[0044] Based on the abnormal trend detection information, abnormal features are extracted, risk assessment rules are preset, and risk assessment is performed according to the extracted abnormal features to obtain risk assessment information.

[0045] If a test safety risk is identified based on the risk assessment information, a control scheme is generated using a fuzzy controller to control the test power supply of the target combined transformer.

[0046] This invention discloses a test monitoring method and system for a combined transformer test power supply, comprising: testing and monitoring the target test power supply to obtain power supply test monitoring information; preprocessing the power supply test monitoring information to obtain preprocessing result information; constructing an anomaly detection model, inputting the preprocessing result information to detect abnormal operating data occurring during power supply testing, and obtaining anomaly detection information; performing trend analysis based on the anomaly detection information to predict future changes in power supply test monitoring data and detect abnormal trends, and obtaining anomaly trend detection information; extracting anomaly features based on the anomaly trend detection information, pre-setting risk assessment rules, and performing risk assessment based on the extracted anomaly features, and obtaining risk assessment information; determining the existence of test safety risks based on the risk assessment information, and if so, using a fuzzy controller to generate a control scheme for control to ensure the safety and reliability of power supply testing. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments or examples of the present invention, the drawings used in the embodiments or examples will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained according to these drawings without creative effort.

[0048] Figure 1 A flowchart of a test monitoring method for a combined transformer test power supply provided in an embodiment of the present invention;

[0049] Figure 2 Provided for one embodiment of the present invention;

[0050] Figure 3 A block diagram of a test monitoring system for a combined transformer test power supply provided in an embodiment of the present invention;

[0051] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0052] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0053] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0054] Figure 1 A flowchart of a test monitoring method for a combined transformer test power supply provided in an embodiment of the present invention;

[0055] like Figure 1 As shown, the present invention provides a flowchart of a test monitoring method for a combined transformer test power supply, including:

[0056] S102, Test and monitor the target test power supply to obtain power supply test and monitoring information, perform data preprocessing on the power supply test and monitoring information to obtain preprocessing result information;

[0057] S104, Construct an anomaly detection model, input the preprocessing result information to detect abnormal operating data that occurs during power supply testing, and obtain anomaly detection information;

[0058] S106, Based on the anomaly detection information, perform trend analysis to predict future changes in power supply test monitoring data and detect abnormal trends to obtain anomaly trend detection information;

[0059] S108, Extract abnormal features based on the abnormal trend detection information, preset risk assessment rules, perform risk assessment based on the extracted abnormal features, and obtain risk assessment information;

[0060] S110, based on the risk assessment information, if a test safety risk is determined to exist, a control scheme is generated using a fuzzy controller to control the test power supply of the target combined transformer.

[0061] It should be noted that this invention provides a test monitoring method and system for a combined transformer test power supply. A sensor array is set up for the power supply during the test process. Based on the set sensor array, power supply test monitoring information is acquired, including test data such as voltage, current, and temperature. This data is then preprocessed to obtain preprocessed result information. Next, an anomaly detection model is constructed using a random forest algorithm. The preprocessed result information is input into the model to detect abnormal data occurring during the test, obtaining anomaly detection information. Subsequently, trend prediction is performed based on the anomaly detection information to analyze future changes in test data and detect whether anomalies will occur in future test data, obtaining anomaly trend detection information. Finally, a risk assessment is performed based on the anomaly trend detection information to determine whether there is a test risk. If a test risk exists, a fuzzy controller is used to generate a control scheme to regulate the target combined transformer test power supply, thereby improving the safety and reliability of the test process.

[0062] Furthermore, in a preferred embodiment of the present invention, the step of testing and monitoring the target test power supply to obtain power supply test monitoring information, and performing data preprocessing on the power supply test monitoring information to obtain preprocessed result information, specifically includes:

[0063] A sensor array is set up in the test power supply of the target combined transformer, and power supply test monitoring is performed based on the set sensor array to obtain power supply test monitoring information;

[0064] The power supply test and monitoring information is processed by using a low-pass filter to remove high-frequency signal noise, and the power supply test and monitoring information data after signal filtering is calibrated.

[0065] A standard reference signal is set, and the power supply test monitoring information after signal filtering is compared with the standard reference signal. The sensor response is calibrated according to the comparison result, and the calibrated power supply test monitoring information is obtained.

[0066] The calibrated power supply test monitoring information is subjected to time-series processing. The standard deviation of the time-series processed power supply test monitoring information is calculated and compared with the preset standard deviation range.

[0067] Data that are outside the preset standard deviation range are defined as outliers and removed. Linear interpolation is then used for interpolation compensation to obtain the preprocessed results.

[0068] It should be noted that during the testing of the target combined transformer power supply, voltage, current, and temperature data are collected in real time using a set sensor array. First, data preprocessing is required. This involves initial filtering of the collected data to remove abnormal parameters that may be caused by sensor interference. Next, sensor calibration is performed to correct any potential sensor deviations. Specific correction schemes include linear and nonlinear corrections, determined based on the sensor's response type. Since different sensors have different acquisition times, further sequencing is necessary to unify the acquisition times and create complete data. Finally, the acquired data is cleaned, outlier removed, and interpolated to obtain accurate test monitoring data, providing a foundation for subsequent analysis.

[0069] Furthermore, in a preferred embodiment of the present invention, the step of constructing an anomaly detection model, inputting the preprocessed result information to detect operational anomaly data occurring during power supply testing, and obtaining anomaly detection information specifically includes:

[0070] An anomaly detection model is constructed based on isolated forest, preprocessing result information is obtained and a time series is generated, which is then input into the anomaly detection model to detect abnormal operating data that occurs during power supply testing.

[0071] The RANSAC algorithm is introduced, which randomly selects several sample data from the input time series, performs linear fitting on the selected sample data, and calculates the difference between the actual observed value and the fitted predicted value based on the linear fitting to generate residual features.

[0072] Feature extraction is performed on the input time series, and a new feature matrix is ​​constructed by combining the residual features. The total number of trees in the isolated forest is determined based on the new feature matrix, and a feature space is constructed.

[0073] A random binary search tree is constructed to detect low-density and isolated data points in the feature space, which are defined as isolated points. The path length of each isolated point in the corresponding tree is calculated as an anomaly score, and compared with a preset threshold to obtain anomaly detection information.

[0074] It's important to note that to ensure test safety, accurate identification of anomalous data during testing is crucial. An anomaly detection model is built based on the Isolation Forest model. The core of the Isolation Forest model lies in identifying outliers in the data by randomly selecting features and constructing binary trees. To further improve the accuracy of anomaly detection, the RANSAC algorithm is introduced. This algorithm performs linear fitting on several samples from a time series, calculates the difference between the actual observed values ​​and the fitted predicted values, and generates residual features. These residual features reflect the degree of deviation between the data and the linear model, enabling more sensitive detection of potential anomalies. By extracting features from the time series and combining them with the residual features to construct a new feature matrix, the model's ability to identify anomalies can be effectively improved. After constructing the new feature matrix, the total number of trees in the Isolation Forest model is determined, and a feature space is constructed accordingly. Next, a random binary search tree is constructed in the feature space to detect data points with low distribution density that appear isolated in the space. After detecting isolated points, their path lengths in the corresponding trees are calculated; shorter paths indicate a more anomalous data point. Finally, the path lengths of these isolated points are used as anomaly scores and compared with a preset threshold to determine whether the data is abnormal, thus obtaining the final anomaly detection information. This accurately identifies abnormal operational data occurring during power supply testing, ensuring the stability and reliability of the testing process.

[0075] Furthermore, in a preferred embodiment of the present invention, the step of performing trend analysis based on the anomaly detection information to predict future changes in power supply test monitoring data and detect abnormal trends to obtain anomaly trend detection information specifically includes:

[0076] A trend prediction model is constructed using a Markov algorithm. The anomaly detection information is input into the trend prediction model for analysis, and the original sequence is generated based on the input anomaly detection information.

[0077] The original sequence is subjected to feature extraction to form a feature matrix, and the similarity is calculated with the feature matrix corresponding to each state transition matrix in the preset state space to obtain a similarity value.

[0078] The calculated similarity value is compared with a preset threshold, and the corresponding state transition matrix is ​​obtained based on the comparison result. The state transition probability at the next moment is calculated by combining the feature matrix of the original sequence.

[0079] The calculated state transition probabilities for the next time step are sorted, and the state with the highest transition probability is selected as the prediction result to generate a prediction sequence, thus obtaining trend prediction result information.

[0080] The trend prediction results are input into the anomaly detection model for detection. Based on the detection results, abnormal data at future times are marked to obtain anomaly trend detection information and generate anomaly trend detection map.

[0081] It is important to note that after detecting abnormal parameters in the real-time acquired data, predictions need to be made based on these anomalies to determine whether future data changes will pose a testing risk, as anomalies during testing may be normal power supply fluctuations. A trend prediction model is constructed using a Markov algorithm, with anomaly detection information used as input to form the original sequence for trend prediction. Features are extracted from the original sequence to form its feature matrix. Then, similarity calculations are performed between this feature matrix and the feature matrices corresponding to each state transition matrix in the preset state space. This determines the degree of matching between the original sequence and the existing state model, providing a basis for the next prediction step. The state transition matrix most similar to the original sequence is selected. The state transition probability for the next time step is calculated using this matrix. After obtaining the state transition probabilities for the next time step, these probabilities are sorted, and the state with the highest transition probability is selected as the prediction result, generating a trend prediction result sequence. Finally, the trend prediction result information is re-inputted into the anomaly detection model for detection. This identifies abnormal data appearing in the prediction results, representing anomalies that may exist in the future, thus obtaining anomaly trend detection information.

[0082] Furthermore, in a preferred embodiment of the present invention, the step of extracting abnormal features based on the abnormal trend detection information, pre-setting risk assessment rules, and performing risk assessment based on the extracted abnormal features to obtain risk assessment information specifically includes:

[0083] Obtain abnormal trend detection information, and extract abnormal data features and corresponding time features based on the abnormal trend detection information to obtain abnormal feature information;

[0084] Risk assessment is performed based on the abnormal feature information, the abnormal feature information is calculated with preset benchmark parameters to obtain the degree of abnormal deviation, and the corresponding time features are correlated with the degree of abnormal deviation based on the abnormal feature information.

[0085] A preset risk assessment rule is established, and the calculated degree of abnormal deviation is compared with the risk assessment rule to assess the test risks existing in the current power supply test and obtain risk assessment information.

[0086] It should be noted that after completing trend prediction and anomaly detection, a risk assessment is further conducted based on the anomaly detection results to determine whether any testing risks exist. Risk assessment rules can be further constructed by analyzing historical anomaly test data that has shown testing risks, setting different risk levels and characteristic profiles for each risk level, thereby assessing the risks of the current test.

[0087] Furthermore, in a preferred embodiment of the present invention, the step of determining the existence of a test safety risk based on the risk assessment information, and if such a risk exists, using a fuzzy controller to generate a control scheme to control the test power supply of the target combined transformer, specifically includes:

[0088] Obtain risk assessment information, preset test security thresholds, compare the risk assessment information with the test security thresholds, and analyze whether there are any test security risks.

[0089] If the judgment result indicates that there is a test safety risk in the current power supply test, then the fuzzy controller is used to generate a control scheme to control the test power supply of the target combined transformer.

[0090] Obtain abnormal feature information and generate control targets based on the abnormal feature information; set input variables based on the degree of abnormal deviation; perform fuzzification processing on the input variables and map them into fuzzy sets; and construct membership functions based on the fuzzy sets.

[0091] Based on big data retrieval, control examples of different types of power supply test anomalies are obtained. Fuzzy rules for control measures of different types of power supply test anomalies are established by combining expert analysis method, and a fuzzy rule base is formed.

[0092] Using the set input variables, fuzzy rule matching is performed in the fuzzy rule base to obtain matching rules. The activation degree of each matching rule is calculated based on the Min-Max inference method. The outputs of multiple rules are combined through Max operation to obtain a fuzzy output set.

[0093] The fuzzy output set is solved by the centroid method, and the corresponding membership function is obtained based on the fuzzy output set. The membership function is then updated based on the activation degree of each output fuzzy set.

[0094] The centroids of each fuzzy output set are calculated based on the updated membership function and then weighted and averaged. A control output value is generated based on the weighted average result, and a control scheme is constructed based on the control output value and input to the actuator for risk control.

[0095] It should be noted that risk assessment information is used to determine whether risk control is necessary. If risk control is required, a fuzzy controller is used to generate a corresponding control scheme to regulate the target transformer test power supply and eliminate safety risks. Control targets, such as voltage or temperature, are generated based on abnormal feature information. Then, input variables, i.e., the abnormal parameter characteristics at the current moment, are set and fuzzified based on the degree of deviation of the abnormality, converting them into fuzzy sets. To formulate effective control schemes, big data technology is used to retrieve control examples of different types of power supply test anomalies. Simultaneously, expert analysis is used to establish fuzzy rules for control measures for different types of power supply test anomalies, forming a complete fuzzy rule base. Next, fuzzy rules are matched in the fuzzy rule base using the set input variables. The activation degree of each matched rule is calculated using the Min-Max inference method, and the outputs of multiple rules are combined using Max operations to obtain the final fuzzy output set. After obtaining the fuzzy output set, it is solved using the centroid method. First, the corresponding membership functions are obtained based on the fuzzy output set, and the membership functions are updated based on the activation degree of each output fuzzy set. Then, the centroid of each fuzzy output set is calculated, and the final control output value is generated by weighted averaging. This value is then input to the actuator to regulate the operating status of the transformer test power supply, ensuring that the risks during the test process are effectively controlled.

[0096] Figure 2 A flowchart of a test risk monitoring and control method for a combined transformer test power supply provided in an embodiment of the present invention;

[0097] like Figure 2 As shown, this invention provides a flowchart of a test risk monitoring and control method for a combined transformer test power supply, including:

[0098] S202, Set a sensor array in the test power supply of the target combined transformer, perform power supply test monitoring based on the set sensor array to obtain power supply test monitoring information, and preprocess the power supply test monitoring information;

[0099] S204. Input the preprocessing results into the anomaly detection model to detect the abnormal data that occurs during power supply testing. Based on the anomaly detection results, perform trend prediction and detect abnormal trends to obtain anomaly trend detection information.

[0100] S206. Based on the abnormal trend detection information, a risk assessment is performed to determine if there is a test risk. If there is a test risk, a control scheme is generated using a fuzzy controller to control the test power supply of the target combined transformer.

[0101] Furthermore, the test monitoring method for a combined transformer test power supply provided by this invention also includes the following steps:

[0102] When a test risk occurs, obtain the test monitoring information and corresponding abnormal detection information of the test power supply of the target combined transformer, conduct test risk abnormality cause analysis, and obtain test risk abnormality cause analysis information;

[0103] Based on the analysis information of the abnormal test risk, determine whether it is a test temperature abnormality risk. If it is a test temperature abnormality risk, analyze the performance release degree of the target combined transformer test power supply when the test risk occurs based on the test monitoring information.

[0104] A performance evaluation model is constructed and performance evaluation indicators are set. The characterization parameters corresponding to each performance evaluation indicator when a test risk occurs are calculated by combining test monitoring information. The parameters are then input into the performance evaluation model to perform performance evaluation and obtain performance evaluation result information.

[0105] Based on the performance evaluation results, it is determined whether the performance of the target combined transformer test power supply is beneficially improved when the risk of abnormal test temperature occurs. If it is beneficially improved, the test parameters corresponding to the current abnormal test temperature are defined as performance release parameters and constitute a performance mode.

[0106] Based on the abnormal detection information corresponding to the occurrence of test temperature risk, abnormal temperature features are extracted, a safe temperature benchmark is obtained and set as the optimization target, and a genetic algorithm is used to solve the abnormal temperature control parameters to obtain temperature control parameter analysis information.

[0107] Based on the temperature control parameter analysis information, a temperature control strategy scheme for the performance mode is constructed and set as the activation condition restriction for the current performance mode.

[0108] It should be noted that during power supply testing, if the identified anomaly is a temperature anomaly, the controllability of temperature allows for analysis of the power supply's performance under these conditions. This involves determining whether the current test parameters, assuming other parameters remain normal, lead to a better power supply usage pattern. If the performance of the target combined transformer test power supply shows a beneficial improvement despite the temperature anomaly risk (i.e., a positive improvement compared to the preset performance), it indicates that the current test parameters can further unleash the corresponding power supply performance. In this case, by understanding the current temperature anomaly characteristics, simply adjusting the temperature anomaly can yield a better power supply usage pattern, thus providing a better understanding of the actual capabilities of the target combined transformer test power supply.

[0109] Figure 3A test monitoring system 3 for a combined transformer test power supply is provided according to an embodiment of the present invention. The system includes: a memory 31 and a processor 32. The memory 31 contains a test monitoring method program for the combined transformer test power supply. When the test monitoring method program for the combined transformer test power supply is executed by the processor 32, it performs the following steps:

[0110] The target test power supply is tested and monitored to obtain power supply test and monitoring information. The power supply test and monitoring information is preprocessed to obtain preprocessed result information.

[0111] An anomaly detection model is constructed, and the preprocessed result information is input to detect abnormal operating data that occurs during power supply testing, thereby obtaining anomaly detection information.

[0112] Based on the anomaly detection information, trend analysis is performed to predict changes in power supply test monitoring data at future times, and abnormal trends are detected to obtain anomaly trend detection information.

[0113] Based on the abnormal trend detection information, abnormal features are extracted, risk assessment rules are preset, and risk assessment is performed according to the extracted abnormal features to obtain risk assessment information.

[0114] If a test safety risk is identified based on the risk assessment information, a control scheme is generated using a fuzzy controller to control the test power supply of the target combined transformer.

[0115] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0116] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0117] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0118] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0119] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0120] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A test monitoring method for a combined transformer test power supply, characterized by, The method comprises the following steps: obtaining power test monitoring information by testing and monitoring the target test power supply, and performing data preprocessing on the power test monitoring information to obtain preprocessing result information; constructing an anomaly detection model, inputting the preprocessing result information to detect running anomaly data occurring during power test, and obtaining anomaly detection information; performing trend analysis based on the anomaly detection information, predicting power test monitoring data changes at future time, and detecting anomaly trends to obtain anomaly trend detection information; extracting anomaly features based on the anomaly trend detection information, presetting risk assessment rules, and performing risk assessment according to the extracted anomaly features to obtain risk assessment information; judging whether there is a test safety risk according to the risk assessment information, and if there is, generating a regulation and control scheme by a fuzzy controller to regulate and control the target combined transformer test power supply; obtaining test monitoring information and corresponding anomaly detection information of the target combined transformer test power supply when a test risk occurs, performing test risk anomaly cause analysis, and obtaining test risk anomaly cause analysis information; judging whether it is a test temperature anomaly risk according to the test risk anomaly cause analysis information, and if it is a test temperature anomaly risk, analyzing the performance release degree of the target combined transformer test power supply when a test risk occurs according to the test monitoring information; constructing a performance evaluation model and setting performance evaluation indexes, combining the test monitoring information to calculate the corresponding characteristic parameters of each performance evaluation index when a test risk occurs, inputting into the performance evaluation model for performance evaluation, and obtaining performance evaluation result information; judging whether the performance of the target combined transformer test power supply is beneficial growth when a test temperature anomaly risk occurs based on the performance evaluation result information, and if it is beneficial growth, defining the test parameters corresponding to the current test temperature anomaly as performance release parameters and forming a performance mode; extracting abnormal temperature features based on the corresponding anomaly detection information when a test temperature risk occurs, obtaining a safe temperature benchmark and setting it as an optimization target, and solving abnormal temperature regulation parameters by a genetic algorithm to obtain temperature regulation parameter analysis information; forming a temperature regulation strategy scheme of the performance mode according to the temperature regulation parameter analysis information, and setting it as the enabling condition limit of the current performance mode; wherein, judging whether there is a test safety risk according to the risk assessment information, and if there is, generating a regulation and control scheme by a fuzzy controller to regulate and control the target combined transformer test power supply, specifically comprising: obtaining risk assessment information, presetting a test safety threshold, judging the risk assessment information and the test safety threshold, and analyzing whether there is a test safety risk; if the judgment result is that the current power test exists a test safety risk, generating a regulation and control scheme by a fuzzy controller to regulate and control the target combined transformer test power supply; obtaining abnormal feature information and generating a control target according to the abnormal feature information, setting input variables based on abnormal deviation degree, fuzzy processing the input variables, mapping them into a fuzzy set, and constructing a membership function according to the fuzzy set; Based on big data retrieval, abnormal regulation instances of different types of power supply tests are obtained, fuzzy rules of abnormal regulation measures of different types of power supply tests are established by combining expert analysis method, and a fuzzy rule base is formed; By using the set input variables, fuzzy rule matching is performed in the fuzzy rule base to obtain matching rules, the activation degree of each matching rule is calculated based on Min-Max reasoning method, and the outputs of multiple rules are combined through Max operation to obtain a fuzzy output set; The fuzzy output set is solved by the barycenter method, the corresponding membership function is obtained according to the fuzzy output set, and the membership function is updated based on the activation degree of each output fuzzy set; The centroids of each fuzzy output set are calculated according to the updated membership function and weighted average is performed, a regulation output value is generated according to the weighted average result, a regulation scheme is formed based on the regulation output value and input to the actuator for risk regulation; Wherein, the trend analysis based on the abnormal detection information, predicting the power supply test monitoring data change at future time, and detecting the abnormal trend, obtaining the abnormal trend detection information, specifically including: A trend prediction model is constructed by using Markov algorithm, the abnormal detection information is input into the trend prediction model for analysis, and an original sequence is generated according to the input abnormal detection information; The feature extraction of the original sequence forms a feature matrix of the original sequence, and the similarity calculation is performed with the feature matrix corresponding to each state transition matrix in the preset state space to obtain a similarity value; The calculated similarity value is compared with the preset threshold value, and the corresponding state transition matrix is obtained according to the comparison result, and the state transition probability at next time is calculated combined with the feature matrix of the original sequence; The calculated state transition probability at next time is sorted, and the state with the highest transition probability is selected as the prediction result to generate a prediction sequence, and a trend prediction result information is obtained; The trend prediction result information is input into the abnormal detection model for detection, and the abnormal data at future time is marked according to the detection result, and the abnormal trend detection information is obtained and an abnormal trend detection graph is generated.

2. The test monitoring method of a combined transformer test power supply according to claim 1, characterized in that, The target test power supply is tested and monitored to obtain power supply test monitoring information, and the power supply test monitoring information is preprocessed to obtain preprocessing result information, specifically including: A sensor array is set on the target combined transformer test power supply, and power supply test monitoring is performed based on the set sensor array to obtain power supply test monitoring information; The power supply test monitoring information is filtered by using a low-pass filter to remove high-frequency signal noise, and the power supply test monitoring information data after signal filtering is calibrated; A standard reference signal is set, the power supply test monitoring information after signal filtering is compared with the standard reference signal, sensor response calibration is performed according to the comparison result, and calibrated power supply test monitoring information is obtained; The calibrated power supply test monitoring information is processed in time sequence, the data standard deviation of the power supply test monitoring information after time sequence processing is calculated, and the data standard deviation is compared with a preset standard deviation range, The data corresponding to the standard deviation of the data not within the preset standard deviation range is defined as an abnormal value for outlier rejection, and linear interpolation is used for interpolation compensation to obtain pre-processing result information.

3. The test monitoring method of a combined transformer test power supply according to claim 1, characterized in that, The abnormal detection model is constructed, the pre-processing result information is input to detect the running abnormal data during the power test, and abnormal detection information is obtained, specifically including: Based on the isolation forest, the pre-processing result information is obtained and a time sequence is generated, which is input into the abnormal detection model to detect the running abnormal data during the power test; The RANSAC algorithm is introduced, a plurality of sample data are randomly selected according to the input time sequence, linear fitting is performed on the selected sample data, and the difference between the actual observation value and the fitting prediction value is calculated according to the linear fitting to generate residual features; The input time sequence is feature extracted, a new feature matrix is constructed in combination with the residual features, the total number of trees in the isolation forest is determined according to the new feature matrix, and a feature space is constructed; The random binary search tree is constructed to detect the data points with low density and isolation in the feature space, which are defined as isolated points, the path length of each isolated point in the corresponding tree is calculated as an abnormal score, which is compared with a preset threshold to obtain abnormal detection information.

4. The test monitoring method of a combined transformer test power supply according to claim 1, characterized in that, The abnormal trend detection information is extracted, a preset risk assessment rule is set, the risk assessment is performed according to the extracted abnormal features, and risk assessment information is obtained, specifically including: The abnormal trend detection information is obtained, the abnormal data features and corresponding time features are extracted according to the abnormal trend detection information, and abnormal feature information is obtained; The risk assessment is performed according to the abnormal feature information, the abnormal feature information is calculated with a preset reference parameter to obtain an abnormal deviation degree, and the corresponding time features are associated with the abnormal deviation degree according to the abnormal feature information; The preset risk assessment rule is set, the calculated abnormal deviation degree is compared with the risk assessment rule to assess the test risk of the current power test, and risk assessment information is obtained.

5. A test monitoring system for a combined transformer test power supply, characterized by The system comprises a memory and a processor, the memory contains a test monitoring method program of the combined transformer test power supply, and the test monitoring method program of the combined transformer test power supply is executed by the processor to realize the test monitoring method steps of the combined transformer test power supply according to any one of claims 1-4.

Images