An intelligent acquisition, analysis and processing method for detection data of transformer cross-cut sheet materials

Abstract

The present invention relates to the technical field of data query and analysis, and specifically to an intelligent acquisition, analysis and processing method for transformer cross-cut sheet detection data. First, a production database for cross-cut sheets is constructed. Subsequently, if the detection result of the cross-cut sheet is unqualified, feature extraction is performed on the unqualified cross-cut sheet to obtain unqualified sample features; a first set of cross-cut sheets is obtained according to the main correlation retrieval similarity, a second set of cross-cut sheets is obtained according to the secondary correlation retrieval conditions, and an unqualified cross-cut sheet set is constructed based on the first set of cross-cut sheets and the second set of cross-cut sheets; an unqualified factor feature vector set is constructed according to the causal strength index; the elements in the unqualified factor feature vector set are processed according to the effective frequency to obtain the main factors of the unqualified cross-cut sheet; the real-time data of the steps or processes where the main factors are located are collected and compared and analyzed with the standard data, and a warning is issued.

Description

Technical Field

[0001] The present invention relates to the technical field of data query and analysis, and in particular to a method for intelligently collecting, analyzing and processing transformer cross-cut sheet material detection data. Background Art

[0002] In the field of transformer cross-cut sheet material inspection, traditional data processing methods rely on offline manual input and single-machine calculations, resulting in insufficient real-time and integrity of inspection data; on the other hand, data storage is scattered in isolated systems and lacks a unified data integration framework, making it impossible to achieve collaborative analysis of multi-source heterogeneous data, and the value of historical data has not been effectively mined.

[0003] Although the existing system has introduced computer-aided analysis, it still has significant defects: first, the cross-protocol integration capability of multimodal data is weak, and a standardized data bus architecture has not been built, resulting in interruptions in real-time data flows; second, the feature extraction and analysis of data rely on static algorithms and lack dynamic weight allocation capabilities such as those based on machine learning models, and cannot adaptively combine multi-source data; third, the causal reasoning logic is rigid, and dynamic reasoning of the root cause of the defect has not been achieved, resulting in low accuracy in solution generation.

[0004] Therefore, a method for intelligent collection, analysis and processing of transformer cross-cut sheet detection data is proposed. Summary of the invention

[0005] The purpose of the present invention is to provide a method for intelligently collecting, analyzing and processing detection data of transformer cross-cut sheets, which improves production efficiency and enhances early warning capabilities by analyzing the detection data and production data of transformer cross-cut sheets. First, a cross-cut sheet production database is constructed. Then, if the detection result of the cross-cut sheet is unqualified, feature extraction is performed on the unqualified cross-cut sheet to obtain unqualified sample features; a first cross-cut sheet set is obtained based on the primary association retrieval similarity, a second cross-cut sheet set is obtained based on the secondary association retrieval condition, and an unqualified cross-cut sheet set is constructed based on the first cross-cut sheet set and the second cross-cut sheet set; an unqualified factor feature vector set is constructed based on the causal strength index; elements in the unqualified factor feature vector set are processed according to the effective frequency to obtain the main factors of the unqualified cross-cut sheet; real-time data of the steps or processes where the main factors are located are collected and compared with the standard data for analysis, and an early warning is issued.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for intelligently collecting, analyzing and processing transformer cross-cut sheet material detection data, comprising:

[0008] Construct a production database of transverse shear sheets based on all production data of transformer transverse shear sheets;

[0009] Further, the cross-cut sheet production database includes a core detection database, a detection equipment parameter database, a processing equipment parameter database, an order database, a material database, and a defect case database. The core detection database includes an image database, a dimension database, a burr database, and a deformation database. The image database stores the images captured by the detection equipment. The dimension database stores the structural data of the cross-cut sheet. The burr database stores the burr distribution data of the cross-cut sheet. The deformation database stores the thickness deformation data of the cross-cut sheet. The detection equipment parameter database is used to store the timing operation parameters of the detection equipment when detecting the cross-cut sheet, including the speed measurement device parameters, the frequency converter parameters, and the line scan camera parameters. The processing equipment parameter database is used to store the timing processing parameters when processing the cross-cut sheet, including the force-displacement curve, the thermal parameters, and the die state parameters during the shearing process. The order database is used to store the order data. The material database is used to store the material data of the cross-cut sheet. The defect case database is used to store the data of the cross-cut sheet with unqualified quality.

[0010] If the detection result of the cross-cut sheet is unqualified in quality, then feature extraction is performed on the unqualified cross-cut sheet to obtain unqualified sample features. According to the unqualified sample features and the main association retrieval similarity, the defect case database is retrieved to obtain the first cross-cut sheet set. According to the unqualified sample features and the secondary association retrieval conditions, the core detection database is retrieved to obtain the second cross-cut sheet set. An unqualified cross-cut sheet set is constructed based on the first cross-cut sheet set and the second cross-cut sheet set.

[0011] Further, the unqualified sample features include defect image features, process parameter hash values, and material features. The unqualified cross-cut sheets in the defect case database are retrieved according to the main association retrieval similarity to obtain the first cross-cut sheet set , and the formula for the main association retrieval similarity is:

[0012] ;

[0013] Among them, represents the main association retrieval similarity between the unqualified sample and the unqualified sample in the defect case database, represents the unqualified sample features of the th unqualified sample in the defect case database, represents the die length calculation formula, represents the natural base, represents the time decay coefficient, represents the time difference between the current sample and the historical sample;

[0014] Query the core detection database according to the secondary association retrieval conditions to obtain the second set of transverse cut sheet materials , and the formula for the secondary association retrieval conditions is: , where represents the sample characteristics of the th transverse cut sheet material , represents the standard deviation of historical data, represents the false detection probability, represents the logarithmic function.

[0015] Furthermore, the construction formula for the set of unqualified transverse cut sheet materials is:

[0016] ;

[0017] where represents the set of unqualified transverse cut sheet materials, represents the first set of transverse cut sheet materials, represents the second set of transverse cut sheet materials, represents the elements in the set of unqualified transverse cut sheet materials, represents the intersection operation, represents the union operation, represents retaining the first combinations with the highest similarity product, represents the Cartesian product operation.

[0018] Obtain the unqualified transverse cut sheet materials in the set of unqualified transverse cut sheet materials and obtain the unqualified factor feature vector according to the causal strength index, and further obtain the set of unqualified factor feature vectors; process the elements in the set of unqualified factor feature vectors according to the effective frequency to obtain the main factors of the unqualified transverse cut sheet materials;

[0019] Furthermore, the calculation formula for the causal strength index is:

[0020] ;

[0021] where represents the causal strength index of event causing event , represents the time decay factor, represents the number of times event occurs independently, represents the number of times event occurs independently, represents the conditional entropy, represents the expectation of event , represents the expectation of event . represents the logarithmic function with base 2, represents the total number of event samples, represents the time step index, which is combined with the time decay factor to calculate the weight decay of historical events on the current causal relationship, represents the total number of time steps, represents the transfer entropy, represents the event at the th expectation, represents the event at the th expectation.

[0022] Furthermore, the specific steps to obtain the main factors of unqualified cross-cut sheet materials include:

[0023] Statistically analyze the effective frequencies of each factor in the set of unqualified factor feature vectors, sort them according to the number of occurrences, and determine the factors exceeding the threshold as the main factors causing the unqualified cross-cut sheet materials. The formula for calculating the effective frequency is:

[0024] ;

[0025] wherein, represents the effective frequency of the th factor, represents the number of occurrences of the th factor in the set, represents the total number of times, represents the summation of the causal strength indicators related to the th factor, represents the weight coefficient of the frequency, represents the weight coefficient of the timeliness correction term, represents the natural base, represents the time decay coefficient, represents the time step index.

[0026] Collect the real-time data of the steps where the main factors are located and compare and analyze it with the standard data.

[0027] Furthermore, collect the real-time data of the steps corresponding to the main factors, construct a real-time data vector, calculate the real-time data vector and the standard data vector, and calculate the Euclidean distance. If it exceeds the threshold, it is determined that there are still unqualified factors in the production process and a warning is issued.

[0028] Compared with the prior art, the beneficial effects of the present invention are:

[0029] 1. A transverse cutting sheet production database including a core detection database, a detection equipment parameter database, a processing equipment parameter database, an order database, a material database, and a defect case database is constructed, realizing the comprehensive integration of data, providing rich data support for subsequent analysis, enabling accurate quality traceability and intelligent defect analysis, providing a basis for analyzing key influencing factors in the processing process, and supporting the optimization and improvement of production processes.

[0030] 2. By mainly associating and retrieving similarities, similar unqualified samples in the defect case database can be accurately matched to form a first set of transverse cutting sheets, which helps to quickly locate similar defect types for easy analysis and processing; the secondary association retrieval conditions for the comprehensive retrieval of transverse cutting sheets form a second set of transverse cutting sheets, which helps to capture potential quality problems; by combining the intersection and union of the first and second sets of transverse cutting sheets to construct a set of unqualified transverse cutting sheets, potential association patterns under different parameter combinations can be explored, which is beneficial for multi-dimensional analysis and in-depth searching for the root causes of quality problems.

[0031] 3. By calculating the causal strength index for the set of unqualified transverse cutting sheets, the influence degree of each factor on product quality unqualified can be quantified, and then the main unqualified factors can be identified; by counting the effective frequencies of each factor in the set of unqualified factor feature vectors and sorting according to the number of occurrences, the main factors causing the unqualified transverse cutting sheets can be determined, which helps to optimize the production process. Brief Description of the Drawings

[0032] Figure 1 It is a flowchart of an intelligent acquisition, analysis, and processing method for transformer transverse cutting sheet detection data provided by an embodiment of the present invention;

[0033] Figure 2 It is a schematic structural diagram of a transverse cutting sheet production database provided by an embodiment of the present invention;

[0034] Figure 3 It is a flowchart of constructing a set of unqualified transverse cutting sheets of the present invention. Detailed Embodiments

[0035] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0036] Embodiment 1

[0037] A certain company has introduced an intelligent acquisition, analysis, and processing method for the detection data of transformer cross-cut sheet materials provided by the present invention. It aims to analyze all the data of the cross-cut sheet materials to determine the factors causing unqualified quality and issue timely warnings. The method process is as follows Figure 1 shown, and the specific implementation is as follows:

[0038] Firstly, a production database for cross-cut sheet materials of transformers is constructed based on all the production data of the cross-cut sheet materials, and the structure is as Figure 2 shown.

[0039] Table 1. Example of order database

[0040]

[0041] Furthermore, the production database for cross-cut sheet materials includes a core detection database, a detection equipment parameter database, a processing equipment parameter database, an order database, a material database, and a defect case database; among them, the core detection database includes an image database, a dimension database, a burr database, and a deformation database. The image database stores the images taken by the detection equipment; the dimension database stores the structural data of the cross-cut sheet materials, the burr database stores the burr distribution data of the cross-cut sheet materials, and the deformation database stores the thickness deformation data of the cross-cut sheet materials; the detection equipment parameter database is used to store the timing operation parameters of the detection equipment when detecting the cross-cut sheet materials each time, including the speed measurement device parameters, frequency converter parameters, and line scan camera parameters; the processing equipment parameter database is used to store the timing processing parameters when processing the cross-cut sheet materials, including the force-displacement curve, thermal parameters, and die state parameters during the shearing process; the order database is used to store order data, and Table 1 shows an example of part of the order database; the material database is used to store the material data of the cross-cut sheet materials; the defect case database is used to store the data of the cross-cut sheet materials with unqualified quality, and Table 2 shows an example of part of the defect case database.

[0042] Table 2. Example of defect case database

[0043]

[0044] Construct a production database for cross-cut sheet materials that includes a core detection database, a detection equipment parameter database, a processing equipment parameter database, an order database, a material database, and a defect case database, comprehensively integrate the data, and form a complete production data chain to provide rich data support for subsequent intelligent analysis.

[0045] If the inspection result of the horizontally cut sheet material is unqualified, feature extraction is performed on the unqualified horizontally cut sheet material to obtain unqualified sample features; according to the unqualified sample features and the main association retrieval similarity, the defect case library is retrieved to obtain the first set of horizontally cut sheet materials. According to the unqualified sample features and the secondary association retrieval conditions, the core inspection database is retrieved to obtain the second set of horizontally cut sheet materials. An unqualified horizontally cut sheet material set is constructed based on the first set of horizontally cut sheet materials and the second set of horizontally cut sheet materials.

[0046] Furthermore, the process of constructing the unqualified horizontally cut sheet material set is as Figure 3 shown, and the specific process includes:

[0047] If the inspection result of the horizontally cut sheet material is unqualified, the data of the unqualified horizontally cut sheet material stored in the core inspection database is obtained and feature extraction is performed to obtain unqualified sample features , including defect image features, process parameter hash values, and material features; the unqualified horizontally cut sheet materials in the defect case library are queried according to the main association retrieval similarity to obtain the first set of horizontally cut sheet materials , and the formula for the main association retrieval similarity is:

[0048] ;

[0049] Among them, represents the main association retrieval similarity between the unqualified sample and the unqualified sample in the defect case library, represents the unqualified sample feature of the th unqualified sample in the defect case library, represents the mold length calculation formula, represents the natural base, represents the time decay coefficient, represents the time difference between the current sample and the historical sample;

[0050] The data in the core inspection database is queried according to the unqualified sample features and the secondary association retrieval conditions to obtain the second set of horizontally cut sheet materials , and the formula for the secondary association retrieval conditions is:

[0051] ;

[0052] Among them, represents the sample feature of the th horizontally cut sheet material , represents the unqualified sample feature of the unqualified sample , represents the historical data standard deviation, represents the allowable false positive probability, represents the logarithmic function;

[0053] By extracting the features of the unqualified cross-cut sheet materials and using the main correlation retrieval similarity to match similar samples from the defect case library, the historical case most similar to the current unqualified sample can be quickly and accurately located, providing a clear reference basis for subsequent cause analysis; the secondary correlation retrieval conditions are used to query all the cross-cut sheet material data, construct the second cross-cut sheet material set, and form a comprehensive coverage of the data.

[0054] Furthermore, the construction formula for constructing the unqualified cross-cut sheet material set is:

[0055] ;

[0056] where, represents the unqualified cross-cut sheet material set, represents the first cross-cut sheet material set, represents the second cross-cut sheet material set, represents the element in the unqualified cross-cut sheet material set, represents the intersection operation, represents the union operation, represents retaining the top cross-parameter combinations with the highest similarity product, represents the Cartesian product operation, which describes all possible combinations of the first cross-cut sheet material set and the second cross-cut sheet material set, and is used to explore the potential association patterns of cross-parameter combinations.

[0057] By performing the combination of intersection, union, and Cartesian product on the first cross-cut sheet material set and the second cross-cut sheet material set through operations, and retaining the part with the highest similarity product among all possible combinations, the comprehensive capture of data is achieved, the key samples are accurately screened, providing data support for subsequent analysis of the factors of unqualified cross-cut sheet materials, and providing support for data-driven quality analysis and management.

[0058] Obtain the unqualified cross-cut sheet materials in the unqualified cross-cut sheet material set and obtain the unqualified factor feature vector according to the causal strength index, and further obtain the unqualified factor feature vector set; process the unqualified factor feature vector set according to the effective frequency to obtain the main factors of the unqualified cross-cut sheet materials;

[0059] Furthermore, the calculation formula for the causal strength index is:

[0060] ;

[0061] where, represents event causing event The causal strength index, represents the time decay factor, emphasizing the importance of recent events, represents the event The number of independent occurrences, represents the event The number of independent occurrences, represents the conditional entropy, measuring the uncertainty of the causal relationship, represents the event The expectation, represents the event The expectation, represents the logarithm function with base 2, represents the total event sample size, and the formula is valid only when it is greater than 1000, represents the time step index, used to measure the relative time interval of event occurrence, combined with the time decay factor to calculate the weight decay of historical events on the current causal relationship, represents the total time step length, limiting the effective time range of the causal relationship to avoid infinite historical tracing, represents the event At the th expectation, represents the event At the th expectation, represents the transfer entropy, calculating the information flow from event to event The formula is:

[0062] ;

[0063] Among them, represents the Shannon entropy of event , represents the event The conditional entropy.

[0064] The causal strength index comprehensively considers the internal uncertainty of the system and the information flow between events by integrating the time decay factor and transfer entropy, accurately characterizes and quantifies the causal strength between events, reduces the ambiguity in causal analysis, realizes the precise matching and data quantification of unqualified cross-cut sheet materials and the factors causing their unqualifiedness, and quickly determines the specific range of unqualified factors.

[0065] Furthermore, the specific steps to obtain the unqualified factors of the unqualified cross-cut sheet materials include:

[0066] Statistically analyze the effective frequencies of each factor in the set of characteristic vectors of unqualified factors, sort them according to the number of occurrences, and determine the factors exceeding the threshold as the main factors causing the unqualified transverse cut sheets. The calculation formula for the effective frequency is as follows:

[0067] ;

[0068] Among them, represents the effective frequency of the th factor, represents the number of occurrences of the th factor in the set, represents the total number of occurrences, represents the summation of the causal strength indicators related to the th factor, represents the weight coefficient of the frequency, represents the weight coefficient of the timeliness correction term, represents the natural base, represents the time decay coefficient, represents the time step index. Table 3 shows the unqualified factors and their effective frequencies in this query. If the effective frequency is greater than 0.7, the unqualified factor is identified as the main factor for unqualified transverse cut sheets.

[0069] Table 3. Some unqualified factors and their effective frequencies

[0070]

[0071] Construct the effective frequency based on the number of occurrences, causal strength indicators, and timeliness correction terms, and sort each unqualified factor according to the effective frequency, which can accurately quantify the influence degree of each unqualified factor on the transverse cut sheets. Identify the main unqualified factors through the threshold, which helps to prioritize the attention and solution of the factors that have the greatest impact on the quality of the transverse cut materials, and improve the efficiency of quality improvement.

[0072] Collect the real-time data of the steps or processes where the unqualified factors are located and compare and analyze them with the standard data. If there are still problems, issue a warning.

[0073] Furthermore, collect the real-time data of the steps or items corresponding to the unqualified factors and construct a real-time data vector, calculate the real-time data vector and the standard data vector, and calculate the Euclidean distance. If it exceeds the threshold, it is determined that there are still unqualified factors in the production process and a warning is issued.

[0074] By collecting and analyzing the real-time data of the steps or positions where the unqualified factors are located, a closed-loop is constructed to trace the generating factors from the unqualified transverse cut sheets, which can realize the real-time detection and early warning of key parameters in the production process, and achieve the analysis and optimization of the production links.

[0075] Through an intelligent acquisition, analysis and processing method for transformer cross-cut sheet detection data provided by the present invention, it is possible to comprehensively integrate and intelligently analyze the whole-process production data of cross-cut sheets, thereby accurately extracting the characteristics of unqualified samples, constructing a set of unqualified cross-cut sheets by using the main correlation retrieval similarity formula and secondary correlation retrieval conditions, and clarifying the main factors affecting product quality by using the causal strength index and effective frequency analysis. Finally, by comparing the data of the key steps corresponding to the main factors collected in real time with the standard data, anomalies in the production process can be detected in a timely manner and early warnings can be issued, thus providing data-driven quality monitoring and continuous improvement support for enterprises.

[0076] Embodiment 2

[0077] A certain company applied an intelligent acquisition, analysis and processing method for transformer cross-cut sheet detection data provided by the present invention to improve the query and analysis capabilities of cross-cut sheet detection data and realize the optimization of the production link. The specific implementation method is as follows:

[0078] Construct a cross-cut sheet production database based on all the production data of the transformer cross-cut sheets;

[0079] Furthermore, the cross-cut sheet production database includes a core detection database, a detection equipment parameter database, a processing equipment parameter database, an order database, a material database and a defect case database; the core detection database includes an image database, a dimension database, a burr database and a deformation database. The image database stores the images taken by the detection equipment, the dimension database stores the structural data of the cross-cut sheets, the burr database stores the burr distribution data of the cross-cut sheets, and the deformation database stores the thickness deformation data of the cross-cut sheets; the detection equipment parameter database is used to store the timing operation parameters of the detection equipment when detecting the cross-cut sheets, including the speed measurement device parameters, the frequency converter parameters and the line scan camera parameters; the processing equipment parameter database is used to store the timing processing parameters when processing the cross-cut sheets, including the force-displacement curve, the thermal parameters and the die state parameters during the shearing process; the order database is used to store order data; the material database is used to store the material data of the cross-cut sheets; the defect case database is used to store the data of the cross-cut sheets with unqualified quality.

[0080] If the detection result of the cross-cut sheet is unqualified in quality, then extract the characteristics of the unqualified cross-cut sheet to obtain the unqualified sample characteristics; retrieve the defect case database according to the unqualified sample characteristics and the main correlation retrieval similarity to obtain the first set of cross-cut sheets, retrieve the core detection database according to the unqualified sample characteristics and the secondary correlation retrieval conditions to obtain the second set of cross-cut sheets, and construct a set of unqualified cross-cut sheets according to the first set of cross-cut sheets and the second set of cross-cut sheets;

[0081] Further, the characteristics of unqualified samples include defective image features, process parameter hash values, and material features; retrieve unqualified cross-cut sheet materials in the defective case library according to the main association retrieval similarity to obtain the first set of cross-cut sheet materials , and the formula for the main association retrieval similarity is:

[0082] ;

[0083] Among them, represents the main association retrieval similarity between the unqualified sample and the unqualified sample in the defective case library , represents the th unqualified sample in the defective case library, represents the die length calculation formula, represents the natural logarithm base, represents the time decay coefficient, represents the time difference between the current sample and the historical sample;

[0084] Retrieve the core detection database according to the secondary association retrieval conditions to obtain the second set of cross-cut sheet materials , and the formula for the secondary association retrieval conditions is: , where represents the th sample feature of the cross-cut sheet material , represents the standard deviation of historical data, represents the false detection probability, represents the logarithmic function.

[0085] Further, the construction formula for the set of unqualified cross-cut sheet materials is:

[0086] ;

[0087] Among them, represents the set of unqualified cross-cut sheet materials, represents the first set of cross-cut sheet materials, represents the second set of cross-cut sheet materials, represents the element in the set of unqualified cross-cut sheet materials, represents the intersection operation, represents the union operation, represents retaining the top combinations with the highest similarity product, represents the Cartesian product operation.

[0088] Obtain the unqualified cross-cut sheet materials in the set of unqualified cross-cut sheet materials and obtain the unqualified factor feature vector according to the causal strength index, and further obtain the set of unqualified factor feature vectors; process the elements in the set of unqualified factor feature vectors according to the effective frequency to obtain the main factors causing the unqualified cross-cut sheet materials.

[0089] Further, the calculation formula of the causal strength index is:

[0090] ;

[0091] Wherein, represents the causal strength index of event causing event , represents the time decay factor, represents the number of times event occurs independently, represents the number of times event occurs independently, represents the conditional entropy, represents the expectation of event , represents the expectation of event , represents the logarithmic function with base 2, represents the total event sample size, represents the time step index, which is combined with the time decay factor to calculate the weight decay of historical events on the current causal relationship, represents the total time step length, represents the transfer entropy, represents the expectation of event at the th time, represents the expectation of event at the th time.

[0092] Further, the specific steps to obtain the main factors of the unqualified cross-cut sheet materials include:

[0093] Statistical the effective frequency of each factor in the set of unqualified factor feature vectors, and sort them according to the number of occurrences. The factors exceeding the threshold are judged as the main factors causing the unqualified cross-cut sheet materials. The calculation formula of the effective frequency is:

[0094] ;

[0095] Wherein, represents the effective frequency of the th factor, represents the The number of occurrences of a factor in the set, Indicates the total number of times, Indicates the summation of the causal strength indicators related to the th factor, Indicates the weight coefficient of the frequency, Indicates the weight coefficient of the timeliness correction term, Indicates the natural base, Indicates the time decay coefficient, Indicates the time step index.

[0096] Collect the real-time data of the step where the main factor is located and conduct a comparative analysis with the standard data.

[0097] Furthermore, collect the real-time data of the step corresponding to the main factor, construct a real-time data vector, calculate the real-time data vector and the standard data vector, and calculate the Euclidean distance. If it exceeds the threshold, it is determined that there are still unqualified factors in the production process and a warning is issued. Table 4 shows the real-time data monitoring results of the key steps corresponding to the main factors. It can be seen that there may be problems in the shearing link, and a warning for the corresponding step is issued in a timely manner.

[0098] Table 4. Real-time data monitoring results

[0099]

[0100] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent acquisition, analysis and processing method for detecting data of transformer cross-cut sheet materials, characterized in that Including: Construct a production database for cross-cut transformer laminations based on all production data of cross-cut transformer laminations; If the inspection result of the cross-cut transformer lamination is unqualified, extract the features of the unqualified cross-cut transformer lamination to obtain the unqualified sample features; Characteristics of unqualified samples including defective image features, process parameter hash values, and material features; retrieving unqualified transverse shearing sheets in the defective case library according to the similarity query of the main correlation retrieval to obtain the first set of transverse shearing sheets , and the formula for the similarity of the main correlation retrieval is as follows: ； Among them, represents unqualified samples and the unqualified samples in the defect case library of the main associated retrieval similarity represents the th unqualified sample in the defect case library characteristics of the unqualified sample represents the modulus length calculation formula represents the natural base represents the time decay coefficient represents the time difference between the current sample and the historical sample; Query the core detection database according to the secondary association retrieval conditions to obtain the second set of transverse cut sheet materials , and the formula for the secondary association retrieval conditions is: , where represents the th sample feature of the transverse cut sheet material , represents the standard deviation of historical data, represents the false detection probability, represents the logarithmic function; Construct a set of unqualified cross-cut transformer laminations based on the first set of cross-cut transformer laminations and the second set of cross-cut transformer laminations; The construction formula for constructing the set of unqualified cross-cut transformer laminations is: ； Among them, represents the set of unqualified transverse shearing blanks, represents the first set of transverse shearing blanks, represents the second set of transverse shearing blanks, represents an element in the set of unqualified transverse shearing blanks, represents the intersection operation, represents the union operation, represents keeping the top combinations with the highest similarity product, represents the Cartesian product operation; Obtain the unqualified cross-cut transformer laminations in the set of unqualified cross-cut transformer laminations and obtain the unqualified factor feature vector according to the causal strength index, and further obtain the set of unqualified factor feature vectors; process the elements in the set of unqualified factor feature vectors according to the effective frequency to obtain the main factors of the unqualified cross-cut transformer laminations; Collect the real-time data of the step where the main factor is located and compare and analyze it with the standard data.

2. The intelligent acquisition, analysis and processing method for the detection data of the transverse shearing sheet of a transformer according to claim 1, wherein, The production database for cross-cut transformer laminations includes a core inspection database, a detection equipment parameter database, a processing equipment parameter database, an order database, a material database, and a defect case database; among them, the core inspection database includes an image database, a dimension database, a burr database, and a deformation database. The image database stores the images taken by the detection equipment, the dimension database stores the structural data of the cross-cut transformer laminations, the burr database stores the burr distribution data of the cross-cut transformer laminations, and the deformation database stores the thickness deformation data of the cross-cut transformer laminations; the detection equipment parameter database is used to store the timing operation parameters of the detection equipment when detecting cross-cut transformer laminations, including the speed measurement device parameters, frequency converter parameters, and line scan camera parameters; the processing equipment parameter database is used to store the timing processing parameters when processing cross-cut transformer laminations, including the force-displacement curve, thermal parameters, and die state parameters during the shearing process; the order database is used to store order data; the material database is used to store the material data of cross-cut transformer laminations; the defect case database is used to store the data of unqualified cross-cut transformer laminations.

3. An intelligent acquisition, analysis and processing method for detecting data of transformer cross-cut sheet materials according to claim 1, characterized in that The calculation formula for the causal strength index is: ； Among them, represents an event causing the event of the causal strength index, represents the time decay factor, represents an event the number of independent occurrences, represents an event the number of independent occurrences, represents the conditional entropy, represents an event the expectation of, represents an event the expectation of, represents the logarithm function with base 2, represents the total event sample size, represents the time step index, combined with the time decay factor to calculate the weight decay of historical events on the current causal relationship, represents the total time step length, represents the transfer entropy, represents an event at the th expectation, represents an event at the th expectation.

4. An intelligent acquisition, analysis and processing method for transformer cross-cut sheet detection data according to claim 1, characterized in that The specific steps for obtaining the main factors of the unqualified cross-cut transformer laminations include: Count the effective frequencies of each factor in the set of unqualified factor feature vectors, sort them according to the number of occurrences, and judge the factors exceeding the threshold as the main factors causing the unqualified cross-cut transformer laminations. The calculation formula for the effective frequency is: ； Among them, represents the effective frequency of the th factor, represents the number of times the th factor appears in the set, represents the total number of times, represents the summation of the causal strength index related to the th factor, represents the weight coefficient of the frequency, represents the weight coefficient of the timeliness correction term, represents the natural base, represents the time decay coefficient, represents the time step index.

5. The intelligent acquisition, analysis and processing method for the detection data of the cross-cut sheet of a transformer according to claim 1, wherein, Collect the real-time data of the step corresponding to the main factor and construct a real-time data vector, calculate the real-time data vector and the standard data vector, and calculate the Euclidean distance. If it exceeds the threshold, it is judged that there are still unqualified factors in the production process and a warning is issued.

Images