CN117250942B - Fault prediction method, device, equipment and storage medium for determining model - Google Patents

Fault prediction method, device, equipment and storage medium for determining model Download PDF

Info

Publication number
CN117250942B
CN117250942B CN202311516569.9A CN202311516569A CN117250942B CN 117250942 B CN117250942 B CN 117250942B CN 202311516569 A CN202311516569 A CN 202311516569A CN 117250942 B CN117250942 B CN 117250942B
Authority
CN
China
Prior art keywords
data
historical
feature
fault
determining
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202311516569.9A
Other languages
Chinese (zh)
Other versions
CN117250942A (en
Inventor
徐永刚
孙成思
何瀚
王灿
谭尚庚
刘昆奇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chengdu Statan Testing Technology Co ltd
Original Assignee
Chengdu Statan Testing Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Chengdu Statan Testing Technology Co ltd filed Critical Chengdu Statan Testing Technology Co ltd
Priority to CN202311516569.9A priority Critical patent/CN117250942B/en
Publication of CN117250942A publication Critical patent/CN117250942A/en
Application granted granted Critical
Publication of CN117250942B publication Critical patent/CN117250942B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B23/00—Testing or monitoring of control systems or parts thereof
    • G05B23/02—Electric testing or monitoring
    • G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
    • G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
    • G05B23/0256—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults injecting test signals and analyzing monitored process response, e.g. injecting the test signal while interrupting the normal operation of the monitored system; superimposing the test signal onto a control signal during normal operation of the monitored system
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques

Landscapes

  • Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Automation & Control Theory (AREA)
  • Telephonic Communication Services (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The application provides a fault prediction method, a device, equipment and a storage medium for determining a model, wherein the method for determining the fault prediction model is used for extracting characteristics of acquired historical characteristic data based on a partial least square analysis method and a historical fault result; and determining a relationship function between the input feature extraction data and the fault prediction result according to the historical feature extraction data, so as to determine a fault prediction model according to the relationship function. The method for determining the fault prediction model can further extract the mapping relation between the feature extraction data and the fault prediction result through a partial least square analysis method, further determine a relation function according to the extracted mapping relation, and determine the fault prediction model with higher fault prediction accuracy based on the relation function.

Description

Fault prediction method, device, equipment and storage medium for determining model
Technical Field
The present invention relates to the field of fault prediction technologies, and in particular, to a fault prediction method, a device, equipment and a storage medium for determining a model.
Background
In industrial production, various components of an automated test equipment such as an automatic integrated circuit tester (Automatic Test Equipment, ATE) inevitably have many faults, for example, electrical faults, control faults, mechanical faults and the like of a motor, and the faults may cause interruption of a test flow, so that the whole production test flow is affected.
The existing method for predicting possible faults of ATE equipment based on the fault prediction model has the technical problem of poor fault prediction accuracy.
Disclosure of Invention
In view of the foregoing, an object of an embodiment of the present application is to provide a fault prediction method, a device, equipment and a storage medium for determining a model, which are used for solving the technical problem of "poor fault prediction accuracy of the existing fault prediction model".
In a first aspect, an embodiment of the present application provides a method for determining a failure prediction model, where the method includes:
acquiring a plurality of groups of historical characteristic data of at least one testing machine in the communication process and a historical fault result corresponding to the historical characteristic data; each group of history characteristic data comprises at least one of communication signals of the testing machine, temperature of the testing machine, pressure of the testing machine, data exchange state of a communication board card, running state of software of the testing machine, classification progress of a sorting machine and transmission rate of an online line;
Performing feature extraction on the historical feature data based on a partial least square analysis method and the historical fault result to obtain historical feature extraction data;
determining a relationship function between the input feature extraction data and the fault prediction result based on the historical feature extraction data;
and determining a fault prediction model according to the relation function.
In the implementation process, the method for determining the fault prediction model performs feature extraction on the obtained historical feature data based on a partial least squares analysis method and a historical fault result; and determining a relationship function between the input feature extraction data and the fault prediction result according to the historical feature extraction data, so as to determine a fault prediction model according to the relationship function. The method for determining the fault prediction model can further extract the mapping relation between the feature extraction data and the fault prediction result through a partial least square analysis method, further determine a relation function according to the extracted mapping relation, determine a fault prediction model with higher fault prediction accuracy based on the relation function, and solve the technical problem of poor fault prediction accuracy of the existing fault prediction model.
In addition, the characteristic extraction is carried out on the historical characteristic data based on the partial least square analysis method and the historical fault result, and the relation function between the input characteristic extraction data and the fault prediction result is determined according to the historical characteristic extraction data obtained by the characteristic extraction, so that the calculated amount in the relation function determination process can be greatly reduced, and the complexity of the training process of the fault prediction model is further reduced.
Optionally, in an embodiment of the present application, the input feature extraction data includes input time domain feature data and input frequency domain feature data; the feature extraction is performed on the historical feature data based on a partial least square analysis method and the historical fault result to obtain historical feature extraction data, and the method comprises the following steps: extracting time domain features of the historical feature data to obtain historical time domain feature data; extracting frequency domain features of the historical feature data to obtain historical frequency domain feature data; and carrying out feature extraction on the historical time domain feature data and the historical frequency domain feature data based on a partial least square analysis method and the historical fault result to obtain historical feature extraction data.
In the implementation process, the obtained historical feature data can be more comprehensively analyzed through time domain feature extraction and frequency domain feature extraction so as to obtain feature parameters (historical time domain feature data and historical frequency domain feature data) with higher relevance to the fault category. And then carrying out feature extraction on the historical time domain feature data and the historical frequency domain feature data based on the partial least square analysis method and the historical fault result, so that historical feature extraction data with better relativity with the fault prediction result can be obtained, and determining a relation function between the input feature extraction data and the fault prediction result according to the historical feature extraction data, thereby further improving the fault prediction accuracy of the fault prediction model.
Optionally, in an embodiment of the present application, the feature extracting the historical time domain feature data and the historical frequency domain feature data based on the partial least squares analysis method and the historical fault result to obtain historical feature extraction data includes: determining an input variable matrix according to the historical time domain feature data and the historical frequency domain feature data; determining an output variable matrix according to the historical fault result; determining a feature vector set according to the input variable matrix, the output variable matrix and a preset covariance relation between the input variable matrix and the output variable matrix; determining the feature vector which meets the preset feature value condition in the feature vector set as the historical feature extraction data; the preset eigenvalue condition is related to the magnitude of the eigenvalue corresponding to each eigenvector in the eigenvector set.
In the implementation process, an input variable matrix carrying input variable information can be determined according to the historical time domain feature data and the historical frequency domain feature data; according to the historical fault result, an output variable matrix carrying output variable information can be determined; according to a preset covariance relation, the correlation relation between an input variable matrix and an output variable matrix is adjusted, and a feature vector set corresponding to the output variable matrix is determined; and selecting a feature vector with strong interpretation capability on the fault prediction result from the feature vector set as historical feature extraction data according to a preset feature value condition. Because the historical feature extraction data has stronger interpretation capability on the fault prediction result, the mapping relationship between the input feature extraction data and the fault prediction result can be well reflected according to the relation function between the input feature extraction data and the fault prediction result determined by the historical feature extraction data. Therefore, the fault prediction model obtained based on the method for determining the fault prediction model can improve the accuracy of fault prediction of the model while reducing the calculated amount in the model training process.
Optionally, in an embodiment of the present application, the determining, based on the historical feature extraction data, a relationship function between the input feature extraction data and the fault prediction result includes: grouping the historical feature extraction data to obtain a training extraction data set and a test extraction data set; wherein the training extraction data set comprises a plurality of training feature data, and the test extraction data set comprises a plurality of test feature data; calculating a distance value between the training feature data and each test feature data; determining a plurality of neighbor feature data in the test feature data according to the distance value; the number of the neighbor feature data is determined according to the number of the feature vectors meeting the preset feature value condition in the feature vector set; determining a fault classification result of the training feature data according to the historical fault result corresponding to the neighbor feature data; wherein the fault classification result comprises whether a fault exists or not and the fault type; and determining a relation function between input feature extraction data and the fault prediction result according to the training feature data and the fault classification result corresponding to each training feature data.
In the implementation process, the historical feature extraction data are divided into a training extraction data set and a test extraction data set; determining a fault classification result of each training feature data (namely determining a fault classification result of each training feature data through K nearest neighbor classification) according to the test extraction data in the test extraction data set and the corresponding historical fault results thereof; the fault classification result corresponding to the training feature data can be determined more accurately, so that the accuracy of a relation function determined based on the training feature data and the fault classification result corresponding to each training feature data is improved, and the fault prediction accuracy of the determined fault prediction model is improved.
Optionally, in an embodiment of the present application, the historical fault result includes: at least one of failure or communication abnormality of an upper computer, resource board abnormality, communication abnormality of a tester and information interaction error does not occur; the obtaining a plurality of groups of historical feature data of at least one testing machine in the communication process and a historical fault result corresponding to the historical feature data comprises the following steps: acquiring a plurality of groups of historical characteristic data of at least one testing machine in the communication process; determining a first data exchange amount between the upper computer and the testing machine, a second data exchange amount of the resource board, a third data exchange amount between the testing machine and the sorting machine and a classification matching degree between the testing machine and the sorting machine according to the historical characteristic data; determining whether the communication abnormality of the upper computer occurs according to the first data exchange amount and a first preset exchange amount threshold; determining whether the resource board is abnormal or not according to the second data exchange amount and a second preset exchange amount threshold; determining whether communication abnormality of the testing machine occurs according to the third data exchange amount and a third preset exchange amount threshold; and determining whether information interaction errors occur according to the change state of the classification matching degree.
In the implementation process, by determining the abnormality judgment data (the first data exchange amount between the upper computer and the testing machine, the second data exchange amount of the resource board, the third data exchange amount between the testing machine and the sorting machine and the classification matching degree between the testing machine and the sorting machine) corresponding to each set of history feature data, and determining whether the upper computer communication abnormality, the resource board abnormality, the testing machine communication abnormality or the information interaction error occurs according to the abnormality judgment data and the preset abnormality judgment conditions, the history fault result corresponding to the history feature data can be obtained. In addition, the existing fault prediction technology can only position whether a fault occurs or not, and cannot accurately position which fault occurs, the scheme of the application can not only judge whether the target testing machine has a fault in the communication process, but also predict the fault type of the target testing machine in the communication process, thereby providing a reference basis for subsequent fault processing and improving the efficiency of fault processing.
In a second aspect, embodiments of the present application further provide a fault prediction method, where the fault prediction method includes:
acquiring communication characteristic data of a target testing machine in a communication process; the communication characteristic data comprise at least one of communication signals of the target testing machine, target testing machine temperature, target testing machine pressure, communication board card data exchange state, testing machine software running state, sorting machine sorting progress and online line transmission rate;
Extracting the characteristics of the communication characteristic data to obtain characteristic extraction data;
inputting the feature extraction data into a fault prediction model, and determining a fault prediction result corresponding to the communication feature data according to the output of the fault prediction model; wherein the failure prediction model is determined according to the failure prediction model determination method as described in any one of the first aspects above.
In a third aspect, embodiments of the present application further provide a fault prediction apparatus, where the fault prediction apparatus includes:
the communication characteristic data acquisition module is used for acquiring communication characteristic data of the target testing machine in a communication process; the communication characteristic data comprise at least one of communication signals of the target testing machine, target testing machine temperature, target testing machine pressure, communication board card data exchange state, testing machine software running state, sorting machine sorting progress and online line transmission rate;
the feature extraction module is used for carrying out feature extraction on the communication feature data to obtain feature extraction data;
the fault prediction module is used for inputting the feature extraction data into a fault prediction model and determining a fault prediction result corresponding to the communication feature data according to the output of the fault prediction model; wherein the failure prediction model is determined according to the failure prediction model determination method as described in any one of the first aspects above.
In a fourth aspect, an embodiment of the present application further provides a device for determining a failure prediction model, where the device for determining a failure prediction model includes:
the historical data acquisition module is used for acquiring historical characteristic data of at least one testing machine in the communication process and a historical fault result corresponding to the historical characteristic data; the historical characteristic data comprise at least one of communication signals of the testing machine, testing machine temperature, testing machine pressure, communication board card data exchange state, testing machine software running state, sorting machine sorting progress and online line transmission rate;
the historical feature extraction module is used for carrying out feature extraction on the historical feature data based on a partial least square analysis method to obtain historical feature extraction data;
the relation function determining module is used for determining a relation function between the input feature extraction data and the fault result based on the historical feature extraction data and the historical fault result;
and the model determining module is used for determining a fault prediction model according to the relation function.
In a fifth aspect, embodiments of the present application further provide an electronic device, including: a memory and a processor, the memory storing a computer program executable by the processor, the computer program, when executed by the processor, performing the method of determining a fault prediction device as described in the first aspect above or the method of predicting a fault as described in the second aspect above.
In a sixth aspect, embodiments of the present application further provide a computer readable storage medium having stored thereon computer program instructions which, when executed by a processor, perform a method of determining a fault prediction device as described in the first aspect above or a method of predicting a fault as described in the second aspect above.
The beneficial effects of this application are: the mapping relation between the feature extraction data and the fault prediction result can be extracted more deeply through the partial least square analysis method, a relation function is determined according to the extracted mapping relation, a fault prediction model with higher fault prediction accuracy is determined based on the relation function, and the technical problem of poor fault prediction accuracy of the existing fault prediction model is solved. The method has the advantages that the characteristic extraction is carried out on the historical characteristic data based on the partial least square analysis method and the historical fault result, the relation function between the input characteristic extraction data and the fault prediction result is determined according to the historical characteristic extraction data obtained by the characteristic extraction, the calculated amount in the relation function determination process can be greatly reduced, and the complexity of the training process of the fault prediction model is further reduced.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a method for determining a failure prediction model according to an embodiment of the present application;
FIG. 2 is a characteristic value deviation graph provided in an embodiment of the present application;
fig. 3 is a schematic flow chart of a fault prediction method provided in an embodiment of the present application;
fig. 4 is a schematic structural diagram of a determining device of a fault prediction model according to an embodiment of the present application;
fig. 5 is a schematic structural diagram of a fault prediction device according to an embodiment of the present application;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
Embodiments of the technical solutions of the present application will be described in detail below with reference to the accompanying drawings. The following examples are only for more clearly illustrating the technical solutions of the present application, and thus are only examples, and are not intended to limit the scope of protection of the present application.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.
In the description of the embodiments of the present application, the technical terms "first," "second," etc. are used merely to distinguish between different objects and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated, a particular order or a primary or secondary relationship. In the description of the embodiments of the present application, the meaning of "plurality" is two or more unless explicitly defined otherwise.
Please refer to fig. 1, which is a flowchart illustrating a method for determining a fault prediction model according to an embodiment of the present application. The method for determining the fault prediction model may include the steps of:
step 101, acquiring a plurality of groups of historical characteristic data of at least one testing machine in the communication process and a historical fault result corresponding to the historical characteristic data; each group of history characteristic data comprises at least one of communication signals of the testing machine, temperature of the testing machine, pressure of the testing machine, data exchange state of a communication board card, running state of software of the testing machine, classification progress of a sorting machine and transmission rate of an online line;
102, carrying out feature extraction on the historical feature data based on a partial least square analysis method and the historical fault result to obtain historical feature extraction data;
step 103, determining a relation function between the input feature extraction data and the fault prediction result based on the historical feature extraction data;
and 104, determining a fault prediction model according to the relation function.
In step 101, the tester may be an ATE device or an automated test device such as an online tester. The method can randomly extract multiple groups of historical characteristic data of one testing machine at different time points in the operation process and after repair, and can also extract multiple groups of historical characteristic data of multiple testing machines of the same model or the same performance series at different time points in the operation process and after repair. Specifically, 1000 sets of historical feature data at different time points can be extracted, or 1500 sets of historical feature data at different time points can be extracted, and the number of sets of the extracted historical feature data can be adjusted according to the actual model precision requirement.
Wherein, in step 102, the partial least squares analysis may project the independent and dependent variables, respectively, into the new space such that the correlation between the two is maximized. The mapping relation between the feature extraction data and the fault prediction result can be extracted more deeply through the partial least square analysis method, further, a relation function is determined according to the extracted mapping relation, and a fault prediction model with higher fault prediction accuracy is determined based on the relation function.
In step 103, the failure prediction result may include whether a failure occurs, or whether a failure occurs and the type of the failure occurred. A "relationship function between input feature extraction data and failure prediction result" may be determined based on the historical feature extraction data and the historical failure result; the historical feature data can be divided into two groups (a training group and a testing group), the fault classification result of the historical feature data of the training group is determined according to the historical feature data of the testing group and the corresponding historical fault result, and the relation function between the input feature extraction data and the fault prediction result is determined according to the historical feature data of the training group and the determined fault classification result.
Wherein, in step 104, the fault prediction model may determine a fault prediction result corresponding to the inputted feature extraction data according to the relation function.
Therefore, the method for determining the fault prediction model provided by the embodiment of the application can further extract the mapping relation between the feature extraction data and the fault prediction result through the partial least square analysis method, further determine the relation function according to the extracted mapping relation, determine the fault prediction model with higher fault prediction accuracy based on the relation function, and solve the technical problem of poor fault prediction accuracy of the existing fault prediction model.
In some alternative embodiments, the input feature extraction data includes input time domain feature data and input frequency domain feature data; step 102, performing feature extraction on the historical feature data based on a partial least square analysis method and the historical fault result to obtain historical feature extraction data, including: extracting time domain features of the historical feature data to obtain historical time domain feature data; extracting frequency domain features of the historical feature data to obtain historical frequency domain feature data; and carrying out feature extraction on the historical time domain feature data and the historical frequency domain feature data based on a partial least square analysis method and the historical fault result to obtain historical feature extraction data.
Wherein, the historical characteristic data can be respectively subjected to characteristic extraction based on a time domain and a frequency domain so as to obtain characteristic parameters (the historical time domain characteristic data and the historical frequency domain characteristic data) which can be directly related to the fault category. Taking the communication signal of the testing machine as an example, in the time domain, the signal characteristics of the communication signal of the testing machine such as the average value, variance, peak value, peak-to-peak value and the like can be measured; in the frequency domain, communication power in different frequency ranges, for example, communication power in a low frequency range (which may be 10-20 Hz), communication power in a medium frequency range (which may be 40-60 Hz), communication power in a high frequency range (greater than 100 Hz), or frequency of a peak, etc., may be calculated. The number of sets of the extracted historical time domain feature data and the historical frequency domain feature data can be 48, 36 or other numerical values, and the specific number of sets can be determined according to the actual model precision requirement.
In some optional embodiments, the feature extracting the historical time domain feature data and the historical frequency domain feature data based on the partial least squares analysis and the historical fault result to obtain historical feature extracted data includes: determining an input variable matrix according to the historical time domain feature data and the historical frequency domain feature data; determining an output variable matrix according to the historical fault result; determining a feature vector set according to the input variable matrix, the output variable matrix and a preset covariance relation between the input variable matrix and the output variable matrix; determining the feature vector which meets the preset feature value condition in the feature vector set as the historical feature extraction data; the preset eigenvalue condition is related to the magnitude of the eigenvalue corresponding to each eigenvector in the eigenvector set.
The "preset covariance relation between the input variable matrix and the output variable matrix" may be that the covariance between the input variable matrix and the output variable matrix is as large as possible. The input variable matrix can be determined according to the historical time domain characteristic data and the historical frequency domain characteristic data Determining an output variable matrix according to the historical fault result>For matrix->Sum matrix->Performing centering processing by calculating matrix ∈>Sum matrix->To obtain a regression coefficient matrix +.>The method comprises the steps of carrying out a first treatment on the surface of the Wherein (1)>The method comprises the steps of carrying out a first treatment on the surface of the By->Calculation matrix->In regression coefficient matrix->Lower projection matrix->By->Calculation matrix->In regression coefficient matrix->Lower projection matrix->The method comprises the steps of carrying out a first treatment on the surface of the Recalculating the resulting projection matrix +.>And->Covariance matrix between the two, according to the projection matrix in the case of maximum rank of covariance matrix +.>The included feature vector "determines a feature vector set. Specifically, the projection matrix may be constructed based on the input variable matrix +.>Constructing a projection matrix based on the output variable matrixThe method comprises the steps of carrying out a first treatment on the surface of the Based on constraint (+)>，/>，/>) Presetting a protocolVariance relation [ ]) A set of feature vectors is determined. The preset feature value condition may be that k feature vectors with the largest feature value are selected as the historical feature extraction data. The specific value of k can be determined according to the inflection point method by arranging each feature vector in the feature vector set according to the magnitude of the feature value. Referring to fig. 2, fig. 2 is a characteristic value deviation chart provided in an embodiment of the present application. The abscissa of fig. 2 is the sequence number of the feature vectors after sorting (the feature values are sorted from large to small), and the ordinate is the feature values corresponding to the feature vectors with different sequence numbers. Taking the eigenvalue deviation graph shown in fig. 2 as an example, it can be seen that after the eigenvalues are arranged according to the magnitude of the eigenvalues, the front 3 eigenvalues are larger but the attenuation speed is very fast, while the eigenvalues behind the 3 rd eigenvalue are basically very small, the attenuation speed is very slow, and k=3 is determined according to the inflection point method.
In some alternative embodiments, step 103, determining a relationship function between the input feature extraction data and the fault prediction result based on the historical feature extraction data, comprises: grouping the historical feature extraction data to obtain a training extraction data set and a test extraction data set; wherein the training extraction data set comprises a plurality of training feature data, and the test extraction data set comprises a plurality of test feature data; calculating a distance value between the training feature data and each test feature data; determining a plurality of neighbor feature data in the test feature data according to the distance value; the number of the neighbor feature data is determined according to the number of the feature vectors meeting the preset feature value condition in the feature vector set; determining a fault classification result of the training feature data according to the historical fault result corresponding to the neighbor feature data; wherein the fault classification result comprises whether a fault exists or not and the fault type; and determining a relation function between input feature extraction data and the fault prediction result according to the training feature data and the fault classification result corresponding to each training feature data.
The test extraction data set may account for 15% or 20% of the number of sets of the acquired historical feature extraction data, and the data proportion relationship between the test extraction data set and the training extraction data set may be adjusted according to the actual model accuracy requirement. Taking the number of sets of the historical feature extraction data as 1500, and taking the example that the test extraction data set accounts for 20% of the number of sets of the obtained historical feature extraction data, the test extraction data set is 300 sets, and the training extraction data set is 1200 sets. The closest test feature data to each training feature data in the test extraction data set may be found out and used as neighboring feature data by calculating the euclidean distance between each training feature data in the training extraction data set and all test feature data in the test extraction data set. The number of neighboring feature data may be equal to "the number of feature vectors satisfying the preset feature value condition in the feature vector set".
The historical fault result can comprise at least one of failure or upper computer communication abnormality, resource board abnormality, tester communication abnormality and information interaction error. Taking the number of the adjacent feature data as 3 as an example, if the historical fault results of two adjacent feature data are abnormal communication of the upper computer, the fault classification result of the corresponding training feature data is abnormal communication of the upper computer; if the historical fault results of the two adjacent feature data are non-faulty, the fault classification result of the corresponding training feature data is non-faulty; if the historical fault results of the two adjacent feature data are that the communication of the upper computer is abnormal and the resource board is abnormal, the fault classification result of the corresponding training feature data is that the communication of the upper computer is abnormal and the resource board is abnormal.
Wherein, with the historical fault result includes: the historical fault results specifically comprise fault classification results of 16 possible faults such as no fault or at least one of communication abnormality of an upper computer, resource board abnormality, communication abnormality of a testing machine and information interaction error of the upper computer (1 possible condition), only one fault (4 possible conditions such as communication abnormality of the upper computer, communication abnormality of the resource board, communication abnormality of the testing machine or information interaction error of the testing machine) and two faults (6 possible conditions are not listed in detail), three faults (4 possible conditions are not listed in detail) and four faults (at the same time, communication abnormality of the upper computer, communication abnormality of the resource board, communication abnormality of the testing machine and information interaction error fault of the testing machine and 1 possible condition) are generated; the 16 fault classification results may be identified (e.g., with 0, 1, 2 … … 15, respectively). The training feature data and the fault classification result (which may be the fault classification result represented by the fault identifier) corresponding to each training feature data may be input into a regression algorithm to perform regression relation fitting, so as to obtain a fault prediction model including a regression relation function. Specifically, the training feature data may be determined as variables, and the regression coefficient of each variable may be calculated using a least square method or a gradient descent method to obtain a "regression relation function between the input feature extraction data and the failure prediction result".
In some alternative embodiments, the historical fault results include: at least one of failure or communication abnormality of an upper computer, resource board abnormality, communication abnormality of a tester and information interaction error does not occur; step 101, obtaining a plurality of sets of historical feature data of at least one testing machine in a communication process and a historical fault result corresponding to the historical feature data, wherein the step comprises the following steps: acquiring a plurality of groups of historical characteristic data of at least one testing machine in the communication process; determining a first data exchange amount between the upper computer and the testing machine, a second data exchange amount of the resource board, a third data exchange amount between the testing machine and the sorting machine and a classification matching degree between the testing machine and the sorting machine according to the historical characteristic data; determining whether the communication abnormality of the upper computer occurs according to the first data exchange amount and a first preset exchange amount threshold; determining whether the resource board is abnormal or not according to the second data exchange amount and a second preset exchange amount threshold; determining whether communication abnormality of the testing machine occurs according to the third data exchange amount and a third preset exchange amount threshold; and determining whether information interaction errors occur according to the change state of the classification matching degree.
The first preset exchange amount threshold value can be 2bit/s, and the fault of abnormal communication of the upper computer is determined under the condition that the first data exchange amount is smaller than 2 bit/s; the first preset swap size threshold may also be 3 bit/s or 5 bit/s. The second preset exchange amount threshold value can be 4bit/s, and the abnormal fault of the resource board is determined under the condition that the second data exchange amount is smaller than 4 bit/s; the second preset swap size threshold may also be 2bit/s or 5 bit/s. The third preset exchange amount threshold value can be 5bit/s, and under the condition that the third data exchange amount is smaller than 5bit/s, the fault of abnormal communication of the tester is determined; the third preset swap size threshold may also be 3 bit/s or 4 bit/s. The fault in which the information interaction error occurs can be determined in the case that the classification matching degree is lowered or is lower than the matching degree threshold value. Specific values such as the first preset exchange amount threshold, the second preset exchange amount threshold, the third preset exchange amount threshold, the matching degree threshold and the like can be adjusted according to actual application scenes (for example, the model of ATE equipment or the actual working environment of the ATE equipment).
The method comprises the steps of extracting a mapping relation between feature extraction data and a fault prediction result by using a partial least square analysis method, determining a relation function according to the extracted mapping relation, and determining a fault prediction model with high fault prediction accuracy based on the relation function. Compared with the method for extracting the mapping relation between the feature extraction data and the fault prediction result by adopting the principal component analysis method, the method for determining the relation function according to the extracted mapping relation and determining the principal component fault prediction model based on the relation function has higher accuracy. In addition, in a specific fault classification process, a K nearest neighbor classifier is adopted to determine a fault classification result of each training feature data, and compared with a 'vector machine fault classification result of each training feature data determined by a support vector machine', the accuracy of the fault classification result obtained based on the implementation mode provided by the application is higher.
Specifically, classification models for comparison are respectively established based on two feature extraction methods (principal component analysis method and partial least squares analysis method), and two classification methods (support vector machine and K nearest neighbor). Aiming at the binary problem of fault prediction, a binary classification model is established. Initially, the support vector machine and K nearest neighbor classifier use the original dataset (which may be 48 feature data) for classifier construction. All 48 features were then filtered using principal component analysis and partial least squares analysis, respectively, and classifier testing was performed using the filtered optimal feature set (feature sets could be tested using the same platform and the same standard), with the results shown in tables 1 and 2 below. The accuracy of the fault prediction model constructed by the partial least square analysis method is obviously improved compared with that of the principal component fault prediction model, the algorithm accuracy of the fault prediction model constructed by the partial least square analysis method is improved by 3-4%, and the running speed is improved by more than 50%. In addition, the results of table 1 also show that the failure prediction model established based on the K nearest neighbor classifier can obtain the best correct classification rate of up to 98.5% in the binary classification problem.
TABLE 1
TABLE 2
For sixteen classification problems, the correct classification rates for the models determined according to the modeling method described above are shown in table 3 below. The result shows that based on the partial least square analysis method and the fault prediction model obtained by training the K nearest neighbor classifier, single faults or multiple faults can be classified more effectively, and the average accuracy of the model can be about 95%. However, as can be seen from the results, the support vector machine is not suitable for the multi-classification problem, and the accuracy rate of sixteen classifications is only about 63%; the reason for this may be noise or overfitting due to pursuit of a full fit (i.e. a full fit to training data and not good predictive performance).
TABLE 3 Table 3
Referring to fig. 3, fig. 3 is a flow chart of a fault prediction method according to an embodiment of the present application. The fault prediction method may include the steps of:
step 201, obtaining communication characteristic data of a target testing machine in a communication process; the communication characteristic data comprise at least one of communication signals of the target testing machine, target testing machine temperature, target testing machine pressure, communication board card data exchange state, testing machine software running state, sorting machine sorting progress and online line transmission rate;
Step 202, extracting the characteristics of the communication characteristic data to obtain characteristic extraction data;
step 203, inputting the feature extraction data into a fault prediction model, and determining a fault prediction result corresponding to the communication feature data according to the output of the fault prediction model; wherein the failure prediction model is determined according to the failure prediction model determination method according to any one of the first aspect.
Wherein the feature extraction data may include time domain feature data and frequency domain feature data; step 202, performing feature extraction on the communication feature data to obtain feature extraction data, which may include: extracting time domain features of the communication feature data to obtain the time domain feature data; and carrying out frequency domain feature extraction on the communication feature data to obtain the frequency domain feature data.
In some alternative embodiments, the fault prediction result includes: at least one of failure or communication abnormality of an upper computer, resource board abnormality, communication abnormality of a tester and information interaction error does not occur; in step 203, after inputting the feature extraction data into a fault prediction model, and determining a fault prediction result corresponding to the communication feature data according to the output of the fault prediction model, the method further includes: and carrying out maintenance and inspection on the target testing machine and related equipment according to the fault prediction result.
The fault prediction method can accurately diagnose and predict the communication fault of the target testing machine, effectively improve the safety and reliability of the target testing machine and reduce the occurrence of disastrous accidents. The fault prediction method is used for continuously and online state monitoring and data analysis of the target testing machine, and can also diagnose and predict the fault development trend of the target testing machine so as to make a predictive maintenance plan in advance and implement maintenance and inspection behaviors. And the shutdown maintenance time of the equipment can be effectively reduced, the hidden trouble of faults can be found as soon as possible, and the fault deterioration is avoided.
Referring to fig. 4, fig. 4 is a schematic structural diagram of a fault prediction device according to an embodiment of the present application. The failure prediction apparatus includes:
the communication characteristic data acquisition module 301 is configured to acquire communication characteristic data of the target test machine in a communication process; the communication characteristic data comprise at least one of communication signals of the target testing machine, target testing machine temperature, target testing machine pressure, communication board card data exchange state, testing machine software running state, sorting machine sorting progress and online line transmission rate;
the feature extraction module 302 is configured to perform feature extraction on the communication feature data to obtain feature extraction data;
The fault prediction module 303 is configured to input the feature extraction data into a fault prediction model, and determine a fault prediction result corresponding to the communication feature data according to an output of the fault prediction model; wherein the failure prediction model is determined according to the failure prediction model determination method as described in any one of the first aspects above.
In some alternative embodiments, the fault prediction result includes: at least one of failure or communication abnormality of an upper computer, resource board abnormality, communication abnormality of a tester and information interaction error does not occur; the failure prediction apparatus further includes: and the maintenance checking module is used for carrying out maintenance checking on the target testing machine and related equipment according to the fault prediction result.
Referring to fig. 5, fig. 5 is a schematic structural diagram of a determining device for a fault prediction model according to an embodiment of the present application. The device for determining the fault prediction model comprises:
a historical data obtaining module 401, configured to obtain historical feature data of at least one testing machine in a communication process and a historical fault result corresponding to the historical feature data; the historical characteristic data comprise at least one of communication signals of the testing machine, testing machine temperature, testing machine pressure, communication board card data exchange state, testing machine software running state, sorting machine sorting progress and online line transmission rate;
A historical feature extraction module 402, configured to perform feature extraction on the historical feature data based on a partial least squares analysis method, so as to obtain historical feature extraction data;
a relation function determining module 403, configured to determine a relation function between the input feature extraction data and the fault result based on the historical feature extraction data and the historical fault result;
the model determining module 404 is configured to determine a fault prediction model according to the relation function.
In some alternative embodiments, the input feature extraction data includes input time domain feature data and input frequency domain feature data; the history feature extraction module 402 is specifically configured to: extracting time domain features of the historical feature data to obtain historical time domain feature data; extracting frequency domain features of the historical feature data to obtain historical frequency domain feature data; and carrying out feature extraction on the historical time domain feature data and the historical frequency domain feature data based on a partial least square analysis method and the historical fault result to obtain historical feature extraction data.
In some alternative embodiments, the input feature extraction data includes input time domain feature data and input frequency domain feature data; the history feature extraction module 402 is specifically configured to: extracting time domain features of the historical feature data to obtain historical time domain feature data; extracting frequency domain features of the historical feature data to obtain historical frequency domain feature data; and carrying out feature extraction on the historical time domain feature data and the historical frequency domain feature data based on a partial least square analysis method and the historical fault result to obtain historical feature extraction data.
In some alternative embodiments, the history feature extraction module 402 is specifically further configured to: determining an input variable matrix according to the historical time domain feature data and the historical frequency domain feature data; determining an output variable matrix according to the historical fault result; determining a feature vector set according to the input variable matrix, the output variable matrix and a preset covariance relation between the input variable matrix and the output variable matrix; determining the feature vector which meets the preset feature value condition in the feature vector set as the historical feature extraction data; the preset eigenvalue condition is related to the magnitude of the eigenvalue corresponding to each eigenvector in the eigenvector set.
In some alternative embodiments, the relationship function determination module 403 is specifically configured to: grouping the historical feature extraction data to obtain a training extraction data set and a test extraction data set; wherein the training extraction data set comprises a plurality of training feature data, and the test extraction data set comprises a plurality of test feature data; calculating a distance value between the training feature data and each test feature data; determining a plurality of neighbor feature data in the test feature data according to the distance value; the number of the neighbor feature data is determined according to the number of the feature vectors meeting the preset feature value condition in the feature vector set; determining a fault classification result of the training feature data according to the historical fault result corresponding to the neighbor feature data; wherein the fault classification result comprises whether a fault exists or not and the fault type; and determining a relation function between input feature extraction data and the fault prediction result according to the training feature data and the fault classification result corresponding to each training feature data.
In some alternative embodiments, the historical fault results include: at least one of failure or communication abnormality of an upper computer, resource board abnormality, communication abnormality of a tester and information interaction error does not occur; the historical data acquisition module 401 is specifically configured to: acquiring a plurality of groups of historical characteristic data of at least one testing machine in the communication process; determining a first data exchange amount between the upper computer and the testing machine, a second data exchange amount of the resource board, a third data exchange amount between the testing machine and the sorting machine and a classification matching degree between the testing machine and the sorting machine according to the historical characteristic data; determining whether the communication abnormality of the upper computer occurs according to the first data exchange amount and a first preset exchange amount threshold; determining whether the resource board is abnormal or not according to the second data exchange amount and a second preset exchange amount threshold; determining whether communication abnormality of the testing machine occurs according to the third data exchange amount and a third preset exchange amount threshold; and determining whether information interaction errors occur according to the change state of the classification matching degree.
It should be understood that the determining means (failure prediction means) of the failure prediction model, whose specific function can be seen from the above description, corresponds to the above-described embodiment of the determining method (failure prediction method) of the failure prediction model, and the detailed description is omitted here appropriately for avoiding repetition, and the steps involved in the above-described embodiment of the method can be performed. The determining means of the fault prediction model (fault prediction means) comprises at least one software functional module which can be stored in the form of software or firmware (firmware) in a memory or solidified in the Operating System (OS) of the device.
Referring to fig. 6, fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present application. An electronic device 500 provided in an embodiment of the present application includes: processor 501 and memory 502, which are interconnected and communicate with each other by a communication bus 503 and/or other form of connection mechanism (not shown). The memory 502 stores a computer program executable by the processor 501, which, when executed by the processor 501, performs the above-described method of determining a failure prediction model or failure prediction method.
Embodiments of the present application also provide a computer readable storage medium having stored thereon computer program instructions that, when executed by the processor 501, perform the above method of determining a failure prediction model or the method of failure prediction.
The storage medium may be implemented by any type of volatile or nonvolatile Memory device or combination thereof, such as static random access Memory (Static Random Access Memory, SRAM), electrically erasable Programmable Read-Only Memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), erasable Programmable Read-Only Memory (Erasable Programmable Read Only Memory, EPROM), programmable Read-Only Memory (PROM), read-Only Memory (ROM), magnetic Memory, flash Memory, magnetic disk, or optical disk.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. The apparatus embodiments described above are merely illustrative, for example, of the flowcharts and block diagrams in the figures that illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, the functional modules in the embodiments of the present application may be integrated together to form a single part, or each module may exist alone, or two or more modules may be integrated to form a single part.
The foregoing description is merely an optional implementation of the embodiments of the present application, but the scope of the embodiments of the present application is not limited thereto, and any person skilled in the art may easily think about changes or substitutions within the technical scope of the embodiments of the present application, and the changes or substitutions should be covered in the scope of the embodiments of the present application.

Claims (7)

1. A method of determining a fault prediction model, the method comprising:
acquiring a plurality of groups of historical characteristic data of at least one testing machine in the communication process and a historical fault result corresponding to the historical characteristic data; each group of history characteristic data comprises at least one of communication signals of the testing machine, temperature of the testing machine, pressure of the testing machine, data exchange state of a communication board card, running state of software of the testing machine, classification progress of a sorting machine and transmission rate of an online line;
performing feature extraction on the historical feature data based on a partial least square analysis method and the historical fault result to obtain historical feature extraction data;
Determining a relationship function between the input feature extraction data and the fault prediction result based on the historical feature extraction data;
determining a fault prediction model according to the relation function;
the input feature extraction data comprises input time domain feature data and input frequency domain feature data; the feature extraction is performed on the historical feature data based on a partial least square analysis method and the historical fault result to obtain historical feature extraction data, and the method comprises the following steps: extracting time domain features of the historical feature data to obtain historical time domain feature data; extracting frequency domain features of the historical feature data to obtain historical frequency domain feature data; performing feature extraction on the historical time domain feature data and the historical frequency domain feature data based on a partial least square analysis method and the historical fault result to obtain historical feature extraction data;
the feature extraction is performed on the historical time domain feature data and the historical frequency domain feature data based on the partial least square analysis method and the historical fault result to obtain historical feature extraction data, and the method comprises the following steps: determining an input variable matrix according to the historical time domain feature data and the historical frequency domain feature data; determining an output variable matrix according to the historical fault result; determining a feature vector set according to the input variable matrix, the output variable matrix and a preset covariance relation between the input variable matrix and the output variable matrix; determining the feature vector which meets the preset feature value condition in the feature vector set as the historical feature extraction data; wherein, the preset characteristic value condition is related to the magnitude of the characteristic value corresponding to each characteristic vector in the characteristic vector set;
Wherein the determining a relationship function between the input feature extraction data and the failure prediction result based on the history feature extraction data includes: grouping the historical feature extraction data to obtain a training extraction data set and a test extraction data set; wherein the training extraction data set comprises a plurality of training feature data, and the test extraction data set comprises a plurality of test feature data; calculating a distance value between the training feature data and each test feature data; determining a plurality of neighbor feature data in the test feature data according to the distance value; the number of the neighbor feature data is determined according to the number of the feature vectors meeting the preset feature value condition in the feature vector set; determining a fault classification result of the training feature data according to the historical fault result corresponding to the neighbor feature data; wherein the fault classification result comprises whether a fault exists or not and the fault type; and determining a relation function between input feature extraction data and the fault prediction result according to the training feature data and the fault classification result corresponding to each training feature data.
2. The method of claim 1, wherein the historical fault results comprise: at least one of failure or communication abnormality of an upper computer, resource board abnormality, communication abnormality of a tester and information interaction error does not occur; the obtaining a plurality of groups of historical feature data of at least one testing machine in the communication process and a historical fault result corresponding to the historical feature data comprises the following steps:
acquiring a plurality of groups of historical characteristic data of at least one testing machine in the communication process;
determining a first data exchange amount between the upper computer and the testing machine, a second data exchange amount of the resource board, a third data exchange amount between the testing machine and the sorting machine and a classification matching degree between the testing machine and the sorting machine according to the historical characteristic data;
determining whether the communication abnormality of the upper computer occurs according to the first data exchange amount and a first preset exchange amount threshold;
determining whether the resource board is abnormal or not according to the second data exchange amount and a second preset exchange amount threshold;
determining whether communication abnormality of the testing machine occurs according to the third data exchange amount and a third preset exchange amount threshold;
and determining whether information interaction errors occur according to the change state of the classification matching degree.
3. A method of fault prediction, the method comprising:
acquiring communication characteristic data of a target testing machine in a communication process; the communication characteristic data comprise at least one of communication signals of the target testing machine, target testing machine temperature, target testing machine pressure, communication board card data exchange state, testing machine software running state, sorting machine sorting progress and online line transmission rate;
extracting the characteristics of the communication characteristic data to obtain characteristic extraction data;
inputting the feature extraction data into a fault prediction model, and determining a fault prediction result corresponding to the communication feature data according to the output of the fault prediction model; wherein the fault prediction model is determined according to the method of claim 1 or 2.
4. A fault prediction device, the device comprising:
the communication characteristic data acquisition module is used for acquiring communication characteristic data of the target testing machine in a communication process; the communication characteristic data comprise at least one of communication signals of the target testing machine, target testing machine temperature, target testing machine pressure, communication board card data exchange state, testing machine software running state, sorting machine sorting progress and online line transmission rate;
The feature extraction module is used for carrying out feature extraction on the communication feature data to obtain feature extraction data;
the fault prediction module is used for inputting the feature extraction data into a fault prediction model and determining a fault prediction result corresponding to the communication feature data according to the output of the fault prediction model; wherein the fault prediction model is determined according to the method of claim 1 or 2.
5. A device for determining a failure prediction model, the device comprising:
the historical data acquisition module is used for acquiring historical characteristic data of at least one testing machine in the communication process and a historical fault result corresponding to the historical characteristic data; the historical characteristic data comprise at least one of communication signals of the testing machine, testing machine temperature, testing machine pressure, communication board card data exchange state, testing machine software running state, sorting machine sorting progress and online line transmission rate;
the historical feature extraction module is used for carrying out feature extraction on the historical feature data based on a partial least square analysis method to obtain historical feature extraction data;
the relation function determining module is used for determining a relation function between the input feature extraction data and the fault prediction result based on the historical feature extraction data and the historical fault result;
The model determining module is used for determining a fault prediction model according to the relation function;
the input feature extraction data comprises input time domain feature data and input frequency domain feature data; the history feature extraction module is specifically configured to: extracting time domain features of the historical feature data to obtain historical time domain feature data; extracting frequency domain features of the historical feature data to obtain historical frequency domain feature data; performing feature extraction on the historical time domain feature data and the historical frequency domain feature data based on a partial least square analysis method and the historical fault result to obtain historical feature extraction data;
wherein, the history feature extraction module is specifically further configured to: determining an input variable matrix according to the historical time domain feature data and the historical frequency domain feature data; determining an output variable matrix according to the historical fault result; determining a feature vector set according to the input variable matrix, the output variable matrix and a preset covariance relation between the input variable matrix and the output variable matrix; determining the feature vector which meets the preset feature value condition in the feature vector set as the historical feature extraction data; wherein, the preset characteristic value condition is related to the magnitude of the characteristic value corresponding to each characteristic vector in the characteristic vector set;
The relation function determining module is specifically configured to: grouping the historical feature extraction data to obtain a training extraction data set and a test extraction data set; wherein the training extraction data set comprises a plurality of training feature data, and the test extraction data set comprises a plurality of test feature data; calculating a distance value between the training feature data and each test feature data; determining a plurality of neighbor feature data in the test feature data according to the distance value; the number of the neighbor feature data is determined according to the number of the feature vectors meeting the preset feature value condition in the feature vector set; determining a fault classification result of the training feature data according to the historical fault result corresponding to the neighbor feature data; wherein the fault classification result comprises whether a fault exists or not and the fault type; and determining a relation function between input feature extraction data and the fault prediction result according to the training feature data and the fault classification result corresponding to each training feature data.
6. An electronic device, the electronic device comprising:
A memory;
a processor;
the memory having stored thereon a computer program executable by the processor for performing the method of any of claims 1-3 when executed by the processor.
7. A computer readable storage medium, characterized in that the computer readable storage medium has stored therein computer program instructions, which when executed by a processor, perform the method of any of claims 1-3.
CN202311516569.9A 2023-11-15 2023-11-15 Fault prediction method, device, equipment and storage medium for determining model Active CN117250942B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311516569.9A CN117250942B (en) 2023-11-15 2023-11-15 Fault prediction method, device, equipment and storage medium for determining model

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311516569.9A CN117250942B (en) 2023-11-15 2023-11-15 Fault prediction method, device, equipment and storage medium for determining model

Publications (2)

Publication Number Publication Date
CN117250942A CN117250942A (en) 2023-12-19
CN117250942B true CN117250942B (en) 2024-02-27

Family

ID=89131696

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311516569.9A Active CN117250942B (en) 2023-11-15 2023-11-15 Fault prediction method, device, equipment and storage medium for determining model

Country Status (1)

Country Link
CN (1) CN117250942B (en)

Citations (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2005045548A2 (en) * 2003-11-06 2005-05-19 Elan Group Ltd System and process for predictively classifying objects
CN109657982A (en) * 2018-12-20 2019-04-19 三重能有限公司 A kind of fault early warning method and device
CN111680726A (en) * 2020-05-28 2020-09-18 国网上海市电力公司 Transformer fault diagnosis method and system based on neighbor component analysis and k neighbor learning fusion
CN112101142A (en) * 2020-08-27 2020-12-18 深圳市行健自动化股份有限公司 Slurry pump running state evaluation method, monitoring terminal and computer readable storage medium
CN113156243A (en) * 2021-04-09 2021-07-23 南方电网电动汽车服务有限公司 Fault prediction method and prediction system
CN113837596A (en) * 2021-09-22 2021-12-24 广东电网有限责任公司 Fault determination method and device, electronic equipment and storage medium
CN115329876A (en) * 2022-08-17 2022-11-11 平安科技（深圳）有限公司 Equipment fault processing method and device, computer equipment and storage medium
CN115545970A (en) * 2022-10-13 2022-12-30 国网山西省电力公司 Power grid fault analysis method, system, equipment and medium based on digital twinning
CN115587333A (en) * 2022-10-10 2023-01-10 北京智谱华章科技有限公司 Failure analysis fault point prediction method and system based on multi-classification model
CN115705279A (en) * 2021-08-10 2023-02-17 中国移动通信集团浙江有限公司 Intelligent fault early warning method and device based on index data
CN116166993A (en) * 2022-12-26 2023-05-26 国网河北省电力有限公司电力科学研究院 Power line fault type identification method and device, power system and storage medium
CN116226646A (en) * 2023-05-05 2023-06-06 国家石油天然气管网集团有限公司 Method, system, equipment and medium for predicting health state and residual life of bearing
CN116225958A (en) * 2023-03-23 2023-06-06 中国工商银行股份有限公司 Fault prediction method and device, storage medium and electronic equipment
CN116304928A (en) * 2023-03-21 2023-06-23 北京思维实创科技有限公司 Power supply equipment fault prediction method, device, equipment and storage medium
CN116361104A (en) * 2023-02-17 2023-06-30 重庆长安新能源汽车科技有限公司 Big data-based application fault prediction method, device, equipment and storage medium
CN116383606A (en) * 2023-03-17 2023-07-04 无锡市第五人民医院 Constant-current temperature acquisition method and system for distributed medical equipment
CN116399588A (en) * 2023-03-06 2023-07-07 太原理工大学 Rolling bearing fault diagnosis method based on WPD and AFRB-LWUNet under small sample
CN116739154A (en) * 2023-05-26 2023-09-12 华为技术有限公司 Fault prediction method and related equipment thereof
CN116822383A (en) * 2023-08-31 2023-09-29 成都态坦测试科技有限公司 Equipment life prediction model construction method and device, readable storage medium and equipment
CN117033912A (en) * 2023-10-07 2023-11-10 成都态坦测试科技有限公司 Equipment fault prediction method and device, readable storage medium and electronic equipment
CN117033990A (en) * 2023-06-26 2023-11-10 华能大理风力发电有限公司祥云分公司 Photovoltaic array fault feature extraction and modeling method

Patent Citations (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2005045548A2 (en) * 2003-11-06 2005-05-19 Elan Group Ltd System and process for predictively classifying objects
CN109657982A (en) * 2018-12-20 2019-04-19 三重能有限公司 A kind of fault early warning method and device
CN111680726A (en) * 2020-05-28 2020-09-18 国网上海市电力公司 Transformer fault diagnosis method and system based on neighbor component analysis and k neighbor learning fusion
CN112101142A (en) * 2020-08-27 2020-12-18 深圳市行健自动化股份有限公司 Slurry pump running state evaluation method, monitoring terminal and computer readable storage medium
CN113156243A (en) * 2021-04-09 2021-07-23 南方电网电动汽车服务有限公司 Fault prediction method and prediction system
CN115705279A (en) * 2021-08-10 2023-02-17 中国移动通信集团浙江有限公司 Intelligent fault early warning method and device based on index data
CN113837596A (en) * 2021-09-22 2021-12-24 广东电网有限责任公司 Fault determination method and device, electronic equipment and storage medium
CN115329876A (en) * 2022-08-17 2022-11-11 平安科技（深圳）有限公司 Equipment fault processing method and device, computer equipment and storage medium
CN115587333A (en) * 2022-10-10 2023-01-10 北京智谱华章科技有限公司 Failure analysis fault point prediction method and system based on multi-classification model
CN115545970A (en) * 2022-10-13 2022-12-30 国网山西省电力公司 Power grid fault analysis method, system, equipment and medium based on digital twinning
CN116166993A (en) * 2022-12-26 2023-05-26 国网河北省电力有限公司电力科学研究院 Power line fault type identification method and device, power system and storage medium
CN116361104A (en) * 2023-02-17 2023-06-30 重庆长安新能源汽车科技有限公司 Big data-based application fault prediction method, device, equipment and storage medium
CN116399588A (en) * 2023-03-06 2023-07-07 太原理工大学 Rolling bearing fault diagnosis method based on WPD and AFRB-LWUNet under small sample
CN116383606A (en) * 2023-03-17 2023-07-04 无锡市第五人民医院 Constant-current temperature acquisition method and system for distributed medical equipment
CN116304928A (en) * 2023-03-21 2023-06-23 北京思维实创科技有限公司 Power supply equipment fault prediction method, device, equipment and storage medium
CN116225958A (en) * 2023-03-23 2023-06-06 中国工商银行股份有限公司 Fault prediction method and device, storage medium and electronic equipment
CN116226646A (en) * 2023-05-05 2023-06-06 国家石油天然气管网集团有限公司 Method, system, equipment and medium for predicting health state and residual life of bearing
CN116739154A (en) * 2023-05-26 2023-09-12 华为技术有限公司 Fault prediction method and related equipment thereof
CN117033990A (en) * 2023-06-26 2023-11-10 华能大理风力发电有限公司祥云分公司 Photovoltaic array fault feature extraction and modeling method
CN116822383A (en) * 2023-08-31 2023-09-29 成都态坦测试科技有限公司 Equipment life prediction model construction method and device, readable storage medium and equipment
CN117033912A (en) * 2023-10-07 2023-11-10 成都态坦测试科技有限公司 Equipment fault prediction method and device, readable storage medium and electronic equipment

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
数据驱动的航空发动机气路故障诊断与剩余寿命预测研究;李汶骏;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》(第2022年02期);C031-118 *
舵机多应力加速试验及寿命评估技术研究;赵少康;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》(第2022年04期);C032-60 *

Also Published As

Publication number Publication date
CN117250942A (en) 2023-12-19

Similar Documents

Publication Publication Date Title
CN115034248A (en) Automatic diagnostic method, system and storage medium for equipment
CN108196143A (en) Power Transformer Faults depth diagnostic method and terminal device
CN113488414B (en) Wafer production monitoring method, system and electronic equipment
CN111354496B (en) Nuclear power plant accident online diagnosis and state tracking prediction method
CN114077919A (en) System for predicting machining anomalies
US11276300B2 (en) Method for learning latest data considering external influences in early warning system and system for same
CN109844779B (en) Method and system for analyzing measurement-yield correlation
CN111309584B (en) Data processing method, device, electronic equipment and storage medium
US20210166362A1 (en) Wafer map identification method and computer-readable recording medium
CN117763432A (en) Communication fault prediction method, device, equipment and storage medium
CN117250942B (en) Fault prediction method, device, equipment and storage medium for determining model
CN116453437B (en) Display screen module testing method, device, equipment and storage medium
Decker et al. Does your model think like an engineer? explainable ai for bearing fault detection with deep learning
CN117786565A (en) Method and device for determining semiconductor abnormal chip and electronic equipment
CN114674511B (en) Bridge modal anomaly early warning method for eliminating time-varying environmental factor influence
CN111061254B (en) PHM system performance evaluation method and system
Ye et al. Self-learning and adaptive board-level functional fault diagnosis
CN111310989B (en) Method and device for predicting part machining success rate and readable storage medium
CN106547690A (en) Based on the Software Reliability Modeling method and device under new fitting criterion
CN114580982B (en) Method, device and equipment for evaluating data quality of industrial equipment
CN112148536A (en) Method and device for detecting deep learning chip, electronic equipment and computer storage medium
CN118312796B (en) High-precision FFT test optimization method, system and storage medium of tester
CN112506784B (en) System and method for evaluating product performance with autonomous learning capability
CN115576850B (en) Data index testing method and device, electronic equipment and storage medium
CN112506803B (en) Big data testing method and system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant