Fault detection method for high-voltage frequency converter

Abstract

The invention discloses a fault detection method for a high-voltage frequency converter, and relates to the technical field of data processing. Real-time acquisition of operating parameters of the high-voltage frequency converter is completed through an acquisition module installed in the high-voltage frequency converter. According to the fault detection method of the high-voltage frequency converter, the data parameters of the same type of high-voltage frequency converter are extracted from the database to establish the fault diagnosis model, and the historical fault data of the high-voltage frequency converter are introduced into the fault diagnosis model for training optimization; at the moment, the operation parameters of the high-voltage frequency converter needing fault detection are introduced into the fault diagnosis model to be processed, and a fault diagnosis result is output and obtained, so that the real-time monitoring of the possibility of equipment faults in the real-time data acquisition process can be completed. And the fault problem is not judged according to the experience of personnel in the personnel detection process, so that the detection accuracy is improved while the fault detection efficiency is improved.

Description

Technical Field

[0001] The invention relates to the technical field of data processing, and in particular to a fault detection method for a high-voltage frequency converter. Background Art

[0002] The reference patent name is: A fault detection method and system for a high-voltage inverter (patent publication number: CN117076955A, patent publication date: 2023-11-17), including: establishing a sample data set, which is a fault signal set, performing wavelet packet decomposition, extracting voltage signal features, constructing an abnormal feature set, establishing an abnormal recognition network, and a universal recognition network, using the fault signal frequency as a universal abnormal recognition constraint, establishing a universal abnormal recognition constraint subnetwork, extracting equipment maintenance features, constructing special abnormal recognition constraints, and establishing a special abnormal recognition constraint subnetwork, reconstructing the abnormal recognition network, collecting real-time signal data, and inputting the reconstructed abnormal recognition network after preprocessing to obtain fault detection results, which solves the technical problems that traditional methods have poor accuracy in fault recognition due to complex circuit structures and signal characteristics, and can usually only deal with specific types of faults or specific equipment, and lack of versatility and adaptability.

[0003] Based on the description in the above-mentioned document, in the existing fault detection process of high-voltage inverters, it is necessary to detect data anomalies through data detection, and then rely on personnel to detect data anomalies to determine the specific fault problems. It is not possible to simultaneously perform real-time monitoring and detection operations of faults while collecting real-time data, and find the time node and location of the problem based on the detected fault. For this reason, the present invention provides a fault detection method for a high-voltage inverter. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a fault detection method for a high-voltage inverter, which solves the problem that in the existing fault detection process of a high-voltage inverter, it is necessary to detect data anomalies through data detection, and then rely on personnel to detect data anomalies to determine the specific fault problem, but it is not possible to simultaneously perform real-time monitoring and detection operations of faults while collecting real-time data, and find the time node and location of the occurrence of the problem based on the detected fault.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: A fault detection method for a high-voltage inverter, specifically comprising the following steps:

[0006] A1. The real-time collection of the operating parameters of the high-voltage inverter is completed through the collection module installed inside the high-voltage inverter;

[0007] A2. Extract data parameters of high-voltage inverters of the same type from the database to establish a fault diagnosis model, and introduce historical fault data of the high-voltage inverter into the fault diagnosis model for training and optimization. At this time, introduce the operating parameters of the high-voltage inverter that needs to be fault-detected into the fault diagnosis model for processing, and output the fault diagnosis results;

[0008] A3. Feedback the fault diagnosis results to the display terminal in the form of text and charts for operators or maintenance personnel to review. At the same time, the personnel perform troubleshooting and maintenance operations based on the fault diagnosis results.

[0009] Preferably, the establishment operation of the fault diagnosis model in A2 is:

[0010] a21. First, establish an initial product model by inputting the structure, functional characteristics and historical parameters of the high-voltage inverter;

[0011] a22. Extract historical fault data from the database to form a training set, and introduce the training set into the initial product model for training. The initial product model extracts and trains feature values ​​in the training set based on a convolutional neural network to obtain a fault diagnosis model.

[0012] a23. The fault diagnosis model can identify different types of faults by introducing the detected parameter data.

[0013] Preferably, the operation of introducing the high-voltage inverter operating parameters that need to be detected for faults in A2 into the fault diagnosis model for processing is:

[0014] B1. The operating parameters of the high-voltage inverter to be fault-detected are collected in real time and stored in a database, and then the collected operating parameter data are preprocessed to form an initial data set marked as M;

[0015] B2. Establish a data interaction connection between the acquisition module and the fault diagnosis model, input the data set M into the fault diagnosis model to form a characteristic difference curve through matching and extraction of numerical features, and judge the fault problem based on the change of the characteristic difference curve;

[0016] B3. Transmit the fault diagnosis result to the display terminal, and trace the collection time and collection location data when the problem data is generated based on the fault diagnosis result.

[0017] Preferably, the operation of preprocessing the operating parameter data collected in B1 is:

[0018] b11. Clean the collected operating parameter data to remove errors, incomplete data, and duplicate collected data;

[0019] b12. Perform classification operations based on the cleaned data, classify according to data type, and convert data of the same type to keep the units consistent;

[0020] b13. Gather the processed data to form a data set M classified according to data type.

[0021] Preferably, the data connection between the acquisition module and the fault diagnosis model in B2 is established in the following manner:

[0022] b21, the acquisition module sends a data reception signal to the signal receiving end of the fault diagnosis model, and the fault diagnosis model sends a feedback signal to the signal receiving end of the acquisition module after determining that the signal has been received;

[0023] b22. At this time, after the acquisition module determines that the feedback signal has been received, the data transmission connection between the acquisition module and the fault diagnosis model is established, and the data is transmitted in real time;

[0024] b23. Based on the data transmission process, the acquisition module continuously sends a maintenance signal to the fault diagnosis model. If the feedback signal of the fault diagnosis model is not received within the time threshold T of the signal sent by the acquisition module, it means that the data transmission connection between the acquisition module and the fault diagnosis model is faulty.

[0025] Preferably, the operation of the fault diagnosis model in B2 processing the data set M to form a characteristic difference curve is:

[0026] C1. Introduce parameter data at a certain time node from the data type of data set M into the fault diagnosis model, and extract all normal operating parameter data at the time node through the fault diagnosis model, and form a normal parameter data set with the normal operating parameter data in the fault diagnosis model and mark it as N;

[0027] C2, then forming a characteristic parameter data set with other parameter data at the time node in the data set M and marking it as L, and performing difference calculation between the corresponding parameter values ​​in the normal parameter data set N and the characteristic parameter data set L;

[0028] The expression for difference calculation is: P = p(L n )-p(N n );

[0029] Where P is the difference between the value of the data subset in the characteristic parameter data set L and the value of the data subset in the normal parameter data set N, and p(Ln) represents the value corresponding to the data subset in the characteristic parameter data set, and Ln represents the nth characteristic parameter data set in the order, p(Nn) represents the value corresponding to the data subset in the normal parameter data set, and Nn represents the nth normal parameter data set in the order;

[0030] C3. According to the time node sequence during work, coordinate axis points are established for the parameter value difference between the normal parameter data set N and the characteristic parameter data set L to form a characteristic difference curve.

[0031] Preferably, the operation of establishing the coordinate axis according to the parameter value difference in C3 is:

[0032] c31. Establish the X-axis in chronological order based on the time nodes of data collection during the product work process and mark the coordinate axis nodes.

[0033] c32, using the corresponding data types of the normal parameter dataset N and the characteristic parameter dataset L under the coordinate axis node as the Y axis, and using the parameter value difference range of the normal parameter dataset N and the characteristic parameter dataset L as the coordinate axis node;

[0034] c33, and mark it with the difference between the data subset value in the characteristic parameter data set L under the corresponding data type and the data subset value in the normal parameter data set N, and connect the marked points under the same data type in a curve according to the time node sequence to form a characteristic difference curve.

[0035] Preferably, the operation of determining the fault problem based on the characteristic difference curve change in B2 is:

[0036] D1, inserting the fluctuation threshold of the corresponding data type in the database and the normal parameter data set N into the characteristic difference curve for comparison, and the fluctuation threshold is marked as ±Qm, where m represents different types of parameters;

[0037] D2. According to the change of the characteristic difference curve above and below the X-axis, a straight line y=±Qm is inserted to intersect the characteristic difference curve. The characteristic difference curve area above y=Qm ​​is the positive excess curve segment, and the characteristic difference curve area below y=-Qm is the negative excess curve segment;

[0038] D3. Count the types of parameters located in the positive excess curve segment and the negative excess curve segment at the same time node. At this time, match the data in the fault diagnosis model with the parameter types at the same time node to determine the type of the fault problem.

[0039] Preferably, the starting division point between the positive excess curve segment and the negative excess curve segment in D2 is the abnormal parameter value, and the time node between the abnormal parameter values ​​is used as the starting time node interval.

[0040] Preferably, the operation of the fault diagnosis model in D3 to determine the type of the fault problem is:

[0041] d31. First, select the data type at the same time node and within the normal parameter value range, and match the fault problem data set with the same value as the data type within the normal parameter value range at the same time node of the fault diagnosis model and mark it as R;

[0042] d32, then compare the parameter values ​​in the positive excess curve segment and the negative excess curve segment with the fault problem data set R, and determine the corresponding fault problem according to the comparison result;

[0043] And the expression for fault problem judgment is:

[0044]

[0045] Where U is the determined fault type, (P→R) means comparing the difference between the same type of parameters in the positive and negative excess curve segments with the difference in the fault problem dataset R. It means that two values ​​different from the results are generated according to the comparison results. When the R difference of the compared fault problem data set belongs to the range of the positive exceeding curve segment and the negative exceeding curve segment, the parameter of this type of comparison is also marked as an abnormal parameter value of 1. On the contrary, when the R difference of the compared fault problem data set is within the range of normal parameters, the parameter of this type of comparison is marked as 0. It means that the parameter types marked as 1 are extracted and summarized. S is the fault diagnosis result data set formed by the feedback of historical fault data in the fault diagnosis model. The fault diagnosis result is confirmed by matching the extracted and summarized parameter types with the parameter types in the fault diagnosis result data set.

[0046] The present invention provides a fault detection method for a high-voltage frequency converter. Compared with the prior art, the present invention has the following beneficial effects:

[0047] (1) The fault detection method of the high-voltage inverter establishes a fault diagnosis model by extracting data parameters of the same type of high-voltage inverter from a database, and introduces historical fault data of the high-voltage inverter into the fault diagnosis model for training and optimization. At this time, the operating parameters of the high-voltage inverter to be fault-detected are introduced into the fault diagnosis model for processing, and the fault diagnosis result is output. In this way, real-time monitoring of the possibility of equipment failure can be completed during the real-time data collection process, rather than making a judgment on the fault problem based on the experience of personnel during the detection process. This improves the efficiency of fault detection and also improves the accuracy of detection.

[0048] (2) The fault detection method of the high-voltage inverter inserts the fluctuation threshold of the corresponding data type in the database and the normal parameter data set into the characteristic difference curve for comparison, and obtains the parameter types of the positive excess curve segment and the negative excess curve segment, and then compares the parameter types with the parameter types at the same time node in the fault problem data set to obtain the fault diagnosis result, thereby reducing the impact of normal data fluctuations on the detection results and obtaining a more accurate matching diagnosis result.

[0049] (3) The fault detection method of the high-voltage inverter transmits the fault diagnosis result to the display terminal, and traces the collection time and collection location data when the problem data is generated based on the fault diagnosis result. After the fault problem is found, the information and location of the fault problem are traced through the initial recorded time and collection location data of the abnormal parameter data generated by the fault problem, so as to facilitate subsequent personnel to confirm the problem and carry out corresponding maintenance processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 is a flow chart of the fault detection method of the present invention;

[0051] Figure 2 The figure is a flow chart for establishing the fault diagnosis model of the present invention. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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.

[0053] See also Figure 1-Figure 2 The present invention provides a technical solution: a fault detection method for a high-voltage inverter, which specifically includes the following steps:

[0054] A1. The real-time collection of the operating parameters of the high-voltage inverter is completed through the collection module installed inside the high-voltage inverter;

[0055] A2. Extract data parameters of high-voltage inverters of the same type from the database to establish a fault diagnosis model, and introduce historical fault data of the high-voltage inverter into the fault diagnosis model for training and optimization. At this time, introduce the operating parameters of the high-voltage inverter that needs to be fault-detected into the fault diagnosis model for processing, and output the fault diagnosis results;

[0056] A3. Feedback the fault diagnosis results to the display terminal in the form of text and charts for operators or maintenance personnel to review. At the same time, the personnel perform troubleshooting and maintenance operations based on the fault diagnosis results.

[0057] Among them, a fault diagnosis model is established by extracting data parameters of high-voltage inverters of the same type from the database, and the historical fault data of the high-voltage inverter is introduced into the fault diagnosis model for training and optimization. At this time, the operating parameters of the high-voltage inverter that needs to be fault detected are introduced into the fault diagnosis model for processing, and the fault diagnosis results are output. In this way, real-time monitoring of the possibility of equipment failure can be completed during the real-time data collection process, rather than making judgments on fault problems based on personnel experience during the detection process. This improves the efficiency of fault detection while also improving the accuracy of detection.

[0058] In this embodiment, the establishment operation of the fault diagnosis model in A2 is:

[0059] a21. First, establish an initial product model by inputting the structure, functional characteristics and historical parameters of the high-voltage inverter;

[0060] a22. Extract historical fault data from the database to form a training set, and introduce the training set into the initial product model for training. The initial product model extracts and trains feature values ​​in the training set based on a convolutional neural network to obtain a fault diagnosis model.

[0061] a23. The fault diagnosis model can identify different types of faults by introducing the detected parameter data.

[0062] Among them, the convolutional neural network is an existing learning algorithm, which includes an input layer, a hidden layer and an output layer. It is used to complete the learning of historical fault data for the product model, so as to form a method that can be applied in actual data adoption based on the historical fault data. After the data is input into the fault diagnosis model, the corresponding fault diagnosis result is obtained based on the extraction and matching of data features.

[0063] In this embodiment, the operation of introducing the high-voltage inverter operating parameters that need to be detected for faults in A2 into the fault diagnosis model for processing is:

[0064] B1. The operating parameters of the high-voltage inverter to be fault-detected are collected in real time and stored in a database, and then the collected operating parameter data are preprocessed to form an initial data set marked as M;

[0065] B2. Establish a data interaction connection between the acquisition module and the fault diagnosis model, input the data set M into the fault diagnosis model to form a characteristic difference curve through matching and extraction of numerical features, and judge the fault problem based on the change of the characteristic difference curve;

[0066] B3. Transmit the fault diagnosis result to the display terminal, and trace the collection time and collection location data when the problem data is generated based on the fault diagnosis result.

[0067] When a fault diagnosis result is generated, backtracking is performed based on the abnormal parameter data used in the fault diagnosis result to find the time when the abnormal parameter data was generated, where the sensor was located at the time to collect the data, and where the data was collected. At the same time, based on the backtracking information, personnel determine the approximate location of the fault and confirm and troubleshoot it through personnel.

[0068] Among them, the fault diagnosis result is transmitted to the display terminal, and the collection time and collection location data when the problem data is generated are traced based on the fault diagnosis result. After the fault problem is discovered, the information and location of the fault problem are traced through the initial recorded time and collected location data of the abnormal parameter data generated by the fault problem, so as to facilitate subsequent personnel to confirm the problem and carry out corresponding maintenance processing.

[0069] In this embodiment, the operation of preprocessing the operating parameter data collected in B1 is:

[0070] b11. Clean the collected operating parameter data to remove errors, incomplete data, and duplicate collected data;

[0071] b12. Perform classification operations based on the cleaned data, classify according to data type, and convert data of the same type to keep the units consistent;

[0072] b13. Gather the processed data to form a data set M classified according to data type.

[0073] In this embodiment, the data connection between the acquisition module in B2 and the fault diagnosis model is established in the following manner:

[0074] b21, the acquisition module sends a data reception signal to the signal receiving end of the fault diagnosis model, and the fault diagnosis model sends a feedback signal to the signal receiving end of the acquisition module after determining that the signal has been received;

[0075] b22. At this time, after the acquisition module determines that the feedback signal has been received, the data transmission connection between the acquisition module and the fault diagnosis model is established, and the data is transmitted in real time;

[0076] b23. Based on the data transmission process, the acquisition module continuously sends a maintenance signal to the fault diagnosis model. If the feedback signal of the fault diagnosis model is not received within the time threshold T of the signal sent by the acquisition module, it means that the data transmission connection between the acquisition module and the fault diagnosis model is faulty.

[0077] In this embodiment, the operation of the fault diagnosis model in B2 processing the data set M to form a characteristic difference curve is:

[0078] C1. Introduce parameter data at a certain time node from the data type of data set M into the fault diagnosis model, and extract all normal operating parameter data at the time node through the fault diagnosis model, and form a normal parameter data set with the normal operating parameter data in the fault diagnosis model and mark it as N;

[0079] C2, then forming a characteristic parameter data set with other parameter data at the time node in the data set M and marking it as L, and performing difference calculation between the corresponding parameter values ​​in the normal parameter data set N and the characteristic parameter data set L;

[0080] The expression for difference calculation is: P = p(L n )-p(N n );

[0081] Where P is the difference between the value of the data subset in the characteristic parameter data set L and the value of the data subset in the normal parameter data set N, and p(Ln) represents the value corresponding to the data subset in the characteristic parameter data set, and Ln represents the nth characteristic parameter data set in the order, p(Nn) represents the value corresponding to the data subset in the normal parameter data set, and Nn represents the nth normal parameter data set in the order;

[0082] C3. According to the time node sequence during work, coordinate axis points are established for the parameter value difference between the normal parameter data set N and the characteristic parameter data set L to form a characteristic difference curve.

[0083] In this embodiment, the operation of establishing the coordinate axis according to the parameter value difference in C3 is:

[0084] c31. Establish the X-axis in chronological order based on the time nodes of data collection during the product work process and mark the coordinate axis nodes.

[0085] c32, using the corresponding data types of the normal parameter dataset N and the characteristic parameter dataset L under the coordinate axis node as the Y axis, and using the parameter value difference range of the normal parameter dataset N and the characteristic parameter dataset L as the coordinate axis node;

[0086] c33, and mark it with the difference between the data subset value in the characteristic parameter data set L under the corresponding data type and the data subset value in the normal parameter data set N, and connect the marked points under the same data type in a curve according to the time node sequence to form a characteristic difference curve.

[0087] In this embodiment, the operation of determining the fault problem based on the characteristic difference curve change in B2 is:

[0088] D1, inserting the fluctuation threshold of the corresponding data type in the database and the normal parameter data set N into the characteristic difference curve for comparison, and the fluctuation threshold is marked as ±Qm, where m represents different types of parameters;

[0089] D2. According to the change of the characteristic difference curve above and below the X-axis, a straight line y=±Qm is inserted to intersect the characteristic difference curve. The characteristic difference curve area above y=Qm ​​is the positive excess curve segment, and the characteristic difference curve area below y=-Qm is the negative excess curve segment;

[0090] D3. Count the types of parameters located in the positive excess curve segment and the negative excess curve segment at the same time node. At this time, match the data in the fault diagnosis model with the parameter types at the same time node to determine the type of the fault problem.

[0091] In the present embodiment, the starting segmentation point between the positive excess curve segment and the negative excess curve segment in D2 is the abnormal parameter value, and the time node between the abnormal parameter values ​​is used as the starting time node interval.

[0092] In this embodiment, the operation of the fault diagnosis model in D3 to determine the type of fault problem is:

[0093] d31. First, select the data type at the same time node and within the normal parameter value range, and match the fault problem data set with the same value as the data type within the normal parameter value range at the same time node of the fault diagnosis model and mark it as R;

[0094] d32, then compare the parameter values ​​in the positive excess curve segment and the negative excess curve segment with the fault problem data set R, and determine the corresponding fault problem according to the comparison result;

[0095] And the expression for fault problem judgment is:

[0096]

[0097] Where U is the determined fault type, (P→R) means comparing the difference between the same type of parameters in the positive and negative excess curve segments with the difference in the fault problem dataset R. It means that two values ​​different from the results are generated according to the comparison results. When the R difference of the compared fault problem data set belongs to the range of the positive exceeding curve segment and the negative exceeding curve segment, the parameter of this type of comparison is also marked as an abnormal parameter value of 1. On the contrary, when the R difference of the compared fault problem data set is within the range of normal parameters, the parameter of this type of comparison is marked as 0. It means that the parameter types marked as 1 are extracted and summarized. S is the fault diagnosis result data set formed by the feedback of historical fault data in the fault diagnosis model. The fault diagnosis result is confirmed by matching the extracted and summarized parameter types with the parameter types in the fault diagnosis result data set.

[0098] Among them, the fluctuation thresholds of the corresponding data types in the database and the normal parameter data set are inserted into the characteristic difference curve for comparison, and the parameter types of the positive and negative excess curve segments are obtained. Then, the parameter types are compared with the parameter types at the same time node in the fault problem data set to obtain the fault diagnosis results, which reduces the impact of normal data fluctuations on the detection results and obtains the diagnosis results with more accurate matching.

[0099] Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0100] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0101] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A fault detection method for a high voltage frequency converter, characterized in that: The specific steps include: A1. The real-time collection of the operating parameters of the high-voltage inverter is completed through the collection module installed inside the high-voltage inverter; A2. Extract data parameters of high-voltage inverters of the same type from the database to establish a fault diagnosis model, and introduce historical fault data of the high-voltage inverter into the fault diagnosis model for training and optimization. At this time, introduce the operating parameters of the high-voltage inverter that needs to be fault-detected into the fault diagnosis model for processing, and output the fault diagnosis results; A3. Feedback the fault diagnosis results to the display terminal in the form of text and charts for operators or maintenance personnel to review. At the same time, the personnel perform troubleshooting and maintenance operations based on the fault diagnosis results.

2. The fault detection method for a high-voltage frequency converter according to claim 1, characterized in that: The establishment operation of the fault diagnosis model in A2 is: a21. First, establish an initial product model by inputting the structure, functional characteristics and historical parameters of the high-voltage inverter; a22. Extract historical fault data from the database to form a training set, and introduce the training set into the initial product model for training. The initial product model extracts and trains feature values ​​in the training set based on a convolutional neural network to obtain a fault diagnosis model. a23. The fault diagnosis model can identify different types of faults by introducing the detected parameter data.

3. The fault detection method for a high-voltage frequency converter according to claim 1, characterized in that: The operation of introducing the high-voltage inverter operating parameters that need to be detected for faults in A2 into the fault diagnosis model for processing is: B1. The operating parameters of the high-voltage inverter to be fault-detected are collected in real time and stored in a database, and then the collected operating parameter data are preprocessed to form an initial data set marked as M; B2. Establish a data interaction connection between the acquisition module and the fault diagnosis model, input the data set M into the fault diagnosis model to form a characteristic difference curve through matching and extraction of numerical features, and judge the fault problem based on the change of the characteristic difference curve; B3. Transmit the fault diagnosis result to the display terminal, and trace the collection time and collection location data when the problem data is generated based on the fault diagnosis result.

4. The fault detection method for a high-voltage frequency converter according to claim 3 is characterized in that: The operation of preprocessing the operating parameter data collected in B1 is as follows: b11. Clean the collected operating parameter data to remove errors, incomplete data, and duplicate collected data; b12. Perform classification operations based on the cleaned data, classify according to data type, and convert data of the same type to keep the units consistent; b13. Gather the processed data to form a data set M classified according to data type.

5. The fault detection method for a high-voltage frequency converter according to claim 3, characterized in that: The data connection between the acquisition module and the fault diagnosis model in B2 is established in the following way: b21, the acquisition module sends a data reception signal to the signal receiving end of the fault diagnosis model, and the fault diagnosis model sends a feedback signal to the signal receiving end of the acquisition module after determining that the signal has been received; b22. At this time, after the acquisition module determines that the feedback signal has been received, the data transmission connection between the acquisition module and the fault diagnosis model is established, and the data is transmitted in real time; b23. Based on the data transmission process, the acquisition module continuously sends a maintenance signal to the fault diagnosis model. If the feedback signal of the fault diagnosis model is not received within the time threshold T of the signal sent by the acquisition module, it means that the data transmission connection between the acquisition module and the fault diagnosis model is faulty.

6. A fault detection method for a high-voltage frequency converter according to claim 3, characterized in that: The operation of the fault diagnosis model in B2 processing the data set M to form a characteristic difference curve is: C1. Introduce parameter data at a certain time node from the data type of data set M into the fault diagnosis model, and extract all normal operating parameter data at the time node through the fault diagnosis model, and form a normal parameter data set with the normal operating parameter data in the fault diagnosis model and mark it as N; C2, then forming a characteristic parameter data set with other parameter data at the time node in the data set M and marking it as L, and performing difference calculation between the corresponding parameter values ​​in the normal parameter data set N and the characteristic parameter data set L; The expression for difference calculation is: P = p(L n )-p(N n ); Where P is the difference between the value of the data subset in the characteristic parameter data set L and the value of the data subset in the normal parameter data set N, and p(Ln) represents the value corresponding to the data subset in the characteristic parameter data set, and Ln represents the nth characteristic parameter data set in the order, p(Nn) represents the value corresponding to the data subset in the normal parameter data set, and Nn represents the nth normal parameter data set in the order; C3. According to the time node sequence during work, coordinate axis points are established for the parameter value difference between the normal parameter data set N and the characteristic parameter data set L to form a characteristic difference curve.

7. A fault detection method for a high-voltage frequency converter according to claim 6, characterized in that: The operation of establishing the coordinate axis according to the parameter value difference in C3 is: c31. Establish the X-axis in chronological order based on the time nodes of data collection during the product work process and mark the coordinate axis nodes. c32, using the corresponding data types of the normal parameter dataset N and the characteristic parameter dataset L under the coordinate axis node as the Y axis, and using the parameter value difference range of the normal parameter dataset N and the characteristic parameter dataset L as the coordinate axis node; c33, and mark it with the difference between the data subset value in the characteristic parameter data set L under the corresponding data type and the data subset value in the normal parameter data set N, and connect the marked points under the same data type in a curve according to the time node sequence to form a characteristic difference curve.

8. A fault detection method for a high-voltage frequency converter according to claim 3, characterized in that: The operation of determining the fault problem based on the characteristic difference curve change in B2 is: D1, inserting the fluctuation threshold of the corresponding data type in the database and the normal parameter data set N into the characteristic difference curve for comparison, and the fluctuation threshold is marked as ±Qm, where m represents different types of parameters; D2. According to the change of the characteristic difference curve above and below the X-axis, a straight line y=±Qm is inserted to intersect the characteristic difference curve. The characteristic difference curve area above y=Qm ​​is the positive excess curve segment, and the characteristic difference curve area below y=-Qm is the negative excess curve segment; D3. Count the types of parameters located in the positive excess curve segment and the negative excess curve segment at the same time node. At this time, match the data in the fault diagnosis model with the parameter types at the same time node to determine the type of the fault problem.

9. A fault detection method for a high-voltage frequency converter according to claim 8, characterized in that: The starting dividing points between the positive excess curve segment and the negative excess curve segment in D2 are abnormal parameter values, and the time nodes between the abnormal parameter values ​​are used as the starting time node intervals.

10. A fault detection method for a high-voltage frequency converter according to claim 8, characterized in that: The operation of the fault diagnosis model in D3 to determine the type of fault problem is: d31. First, select the data type at the same time node and within the normal parameter value range, and match the fault problem data set with the same value as the data type within the normal parameter value range at the same time node of the fault diagnosis model and mark it as R; d32, then compare the parameter values ​​in the positive excess curve segment and the negative excess curve segment with the fault problem data set R, and determine the corresponding fault problem according to the comparison result; And the expression for fault problem judgment is: Where U is the determined fault type, (P→R) means comparing the difference between the same type of parameters in the positive and negative excess curve segments with the difference in the fault problem dataset R. It means that two values ​​different from the results are generated according to the comparison results. When the R difference of the compared fault problem data set belongs to the range of the positive exceeding curve segment and the negative exceeding curve segment, the parameter of this type of comparison is also marked as an abnormal parameter value of 1. On the contrary, when the R difference of the compared fault problem data set is within the range of normal parameters, the parameter of this type of comparison is marked as 0. It means that the parameter types marked as 1 are extracted and summarized. S is the fault diagnosis result data set formed by the feedback of historical fault data in the fault diagnosis model. The fault diagnosis result is confirmed by matching the extracted and summarized parameter types with the parameter types in the fault diagnosis result data set.

Images