CN115169832A - Sensitivity analysis method and system based on curve form change - Google Patents
Sensitivity analysis method and system based on curve form change Download PDFInfo
- Publication number
- CN115169832A CN115169832A CN202210693186.8A CN202210693186A CN115169832A CN 115169832 A CN115169832 A CN 115169832A CN 202210693186 A CN202210693186 A CN 202210693186A CN 115169832 A CN115169832 A CN 115169832A
- Authority
- CN
- China
- Prior art keywords
- quality
- process parameters
- sensitivity
- change
- reference sample
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 277
- 230000008859 change Effects 0.000 title claims abstract description 75
- 238000010206 sensitivity analysis Methods 0.000 title claims abstract description 30
- 230000008569 process Effects 0.000 claims abstract description 219
- 239000000523 sample Substances 0.000 claims abstract description 65
- 239000013074 reference sample Substances 0.000 claims abstract description 59
- 230000035945 sensitivity Effects 0.000 claims abstract description 46
- 238000005259 measurement Methods 0.000 claims abstract description 11
- 238000006243 chemical reaction Methods 0.000 claims abstract description 9
- 239000000047 product Substances 0.000 claims description 55
- 238000005070 sampling Methods 0.000 claims description 20
- 238000007689 inspection Methods 0.000 claims description 9
- 238000012360 testing method Methods 0.000 claims description 8
- 230000009466 transformation Effects 0.000 claims description 8
- 238000009776 industrial production Methods 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000004660 morphological change Effects 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 5
- 239000012467 final product Substances 0.000 claims description 5
- 230000002194 synthesizing effect Effects 0.000 claims description 5
- 230000001131 transforming effect Effects 0.000 claims description 2
- 238000004519 manufacturing process Methods 0.000 description 18
- 238000010586 diagram Methods 0.000 description 12
- 239000000463 material Substances 0.000 description 12
- 230000015654 memory Effects 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 238000009826 distribution Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 5
- 238000003860 storage Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003908 quality control method Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000556 factor analysis Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 239000002243 precursor Substances 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06395—Quality analysis or management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Manufacturing & Machinery (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- General Factory Administration (AREA)
Abstract
The invention discloses a sensitivity analysis method and a system based on curve form change, wherein the method comprises the following steps: acquiring process parameters and quality index data to be analyzed, and constructing an original sample data set; constructing a relation model of the process parameters and the quality index data; selecting a reference sample in the original sample data set based on the relation model; converting the process parameters in the reference sample, and respectively predicting the quality categories of the reference sample before and after conversion by using a relation model to obtain the probability variation result of the corresponding quality category in the process of process parameter variation; the sensitivity of the process parameters is measured by utilizing the gradient relation between the process parameter mean value change and the quality category probability change, and the key process parameters influencing the product quality fluctuation are obtained based on the measurement result of the sensitivity. The invention solves the problem of identifying key factors causing product quality fluctuation in the industrial control process.
Description
Technical Field
The invention relates to the technical field of process industrial quality control and optimization, in particular to a sensitivity analysis method and system based on curve morphological change.
Background
In the process of industrial production, an effective analysis strategy needs to be established to identify key factors causing product quality fluctuation in the whole production process, so that relevant problems can be quickly positioned and solved, and high-quality products can be obtained. In the current industrial process, the problem of unstable product quality often occurs.
In the actual production process, the key process parameter variables causing the product quality fluctuation need to be rapidly and accurately identified, so that the quality analysis and the problem solution are performed in a targeted manner. Therefore, it is necessary to utilize a sensitivity analysis method based on the change of the curve form to find out the key factors causing the fluctuation of the product quality in the process industrial production. However, the actual production process may include complex raw material fluctuation, manual intervention, process state change and the like to cause product quality fluctuation, and the conventional analysis method cannot accurately locate the key factors causing the product quality fluctuation and also cannot quantify the influence degree.
Disclosure of Invention
The invention provides a sensitivity analysis method and system based on curve form change, and aims to solve the problems that key factors causing product quality fluctuation cannot be accurately positioned and the influence degree cannot be quantified in the prior art.
In order to solve the technical problems, the invention provides the following technical scheme:
in one aspect, the invention provides a sensitivity analysis method based on curve morphological change, which comprises the following steps:
acquiring process parameters and quality index data to be analyzed, and constructing an original sample data set; wherein, in the original sample data set, the process parameters correspond to the quality index data; the quality category of the sample data is divided into two categories of qualified quality and unqualified quality according to a preset quality index standard;
constructing a relation model of the process parameters and the quality index data;
selecting a reference sample in the original sample data set based on the relationship model;
converting the process parameters in the reference sample, and respectively predicting the quality categories of the reference sample before and after conversion by using the relational model to obtain the probability variation result of the corresponding quality category in the process of process parameter variation;
and measuring the sensitivity of the process parameters by using the gradient relation between the process parameter mean value change and the quality category probability change to obtain key process parameters influencing the product quality fluctuation.
Further, the obtaining of the process parameters and the quality index data to be analyzed and the construction of the original sample data set include:
acquiring a real-time process parameter curve generated by a preset process variable in the process industrial production process and final product inspection quality index data;
and taking the product inspection and testing sampling time as a reference, intercepting process parameter curves in a period of time before and after the sampling time and corresponding to the product quality index data of the current sampling time, and dividing the data into two types of qualified quality and unqualified quality according to a preset quality index specification so as to construct the original sample data set.
Further, the building of the relational model between the process parameters and the quality index data includes:
performing replacement sampling and SMOTE up-sampling on the original sample data set by adopting a Bagging-SMOTE integrated sampling method to obtain a sample subset with balanced quality categories;
training the sample subset by using a classification method of a support matrix machine to obtain each base model;
and (4) integrating the training results of all the base models by adopting a voting method to obtain the relation model.
Further, the selecting of the reference sample in the original sample data set based on the relationship model comprises:
and performing quality type prediction on each sample in the original sample data set by using the relational model, and selecting the sample which is predicted to have a quality qualified probability difference value and a quality unqualified probability difference value smaller than a set threshold value as a reference sample.
Further, the set threshold is larger than the quality category probability difference value when the difference between the numbers of the base models of which the quality of the sample is judged to be qualified and unqualified in all the base models is 1.
Further, transforming the process parameters in the reference sample, and predicting the quality categories of the reference sample before and after transformation by using the relational model to obtain the probability variation result of the corresponding quality category in the process of process parameter variation, which comprises:
and sequentially selecting the reference sample, sequentially replacing the corresponding process parameter curves in the reference sample by using the average value of each process parameter curve in the original sample data set, and respectively predicting the quality categories before and after replacement by using the relational model to obtain the probability change result of the corresponding quality categories in the process of changing the process parameter curves.
Further, the measuring the sensitivity of the process parameter by using the gradient relationship between the process parameter mean value change and the quality category probability change to obtain the key process parameter influencing the product quality fluctuation comprises:
extracting a quality class probability change result in the process of changing the process parameter curve in each reference sample, and measuring the sensitivity of the process parameter by the gradient of the curve mean change and the quality class probability change in the process of changing the process parameter curve;
synthesizing the technological parameter sensitivity measurement results in the technological parameter curve transformation process of the multiple reference samples to obtain the sensitivity of the corresponding technological parameters;
and normalizing the sensitivity result of the process parameters to obtain key process parameters influencing the quality fluctuation of the product.
In another aspect, the present invention further provides a sensitivity analysis system based on the change of the curve form, including:
the process parameter and quality index data acquisition module is used for acquiring process parameters and quality index data to be analyzed and constructing an original sample data set; wherein, in the original sample data set, the process parameters correspond to the quality index data; the quality category of the sample data is divided into two categories of qualified quality and unqualified quality according to a preset quality index standard;
the relational model construction module is used for constructing a relational model of the process parameters and the quality index data;
a reference sample selection and data change module for selecting a reference sample in the original sample data set based on the relationship model; converting the process parameters in the reference sample, and respectively predicting the quality classes of the reference sample before and after conversion by using the relation model to obtain the probability change result of the corresponding quality class in the process of process parameter change;
and the sensitivity measurement module is used for measuring the sensitivity of the process parameters by utilizing the gradient relation between the process parameter mean value change and the quality category probability change to obtain key process parameters influencing the product quality fluctuation.
In yet another aspect, the present invention also provides an electronic device comprising a processor and a memory; wherein the memory has stored therein at least one instruction that is loaded and executed by the processor to implement the above-described method.
In yet another aspect, the present invention also provides a computer-readable storage medium having at least one instruction stored therein, which is loaded and executed by a processor to implement the above-mentioned method.
The technical scheme provided by the invention has the beneficial effects that at least:
the method solves the problem that key factors causing product quality fluctuation can not be accurately identified in the process industrial process, can well identify key process parameters causing product quality fluctuation in the process industrial process, and can quantify the influence degree of process parameter change on the product quality through sensitivity, thereby effectively improving the efficiency and the accuracy of identifying the key factors causing the product quality fluctuation in the process industrial production process.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the description below are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a schematic diagram illustrating an implementation flow of a sensitivity analysis method based on a change in a curve form according to a first embodiment of the present invention;
FIG. 2 is a graph illustrating normalized sensitivity analysis results provided by a first embodiment of the present invention;
FIG. 3 is a graph comparing a key raw process parameter curve provided by a first embodiment of the present invention between qualified and unqualified samples;
FIG. 4 is a schematic diagram of a flow chart for executing a sensitivity analysis method based on curve form change according to a second embodiment of the present invention;
FIG. 5 is a flow chart of the selection of a reference sample, the variation of a curve form based on a mean value, and the sensitivity measurement provided by the second embodiment of the present invention;
FIG. 6 is a graph illustrating the results of a process data and quality data sensitivity analysis for a material manufacturing process according to a second embodiment of the present invention;
FIG. 7 is a graph comparing the mean change of a univariate curve with the change of the probability of quality failure according to the second embodiment of the present invention; wherein, (a) is a graph of the variation of the unqualified quality probability along with the mean value of a torque curve, (b) is a graph of the variation of the unqualified quality probability along with the mean value of a temperature 2 curve, and (c) is a graph of the variation of the unqualified quality probability along with the mean value of a feeding quantity 1 curve;
FIG. 8 is a plot of torque variation curves between quality-qualified and quality-unqualified samples according to a second embodiment of the present invention;
FIG. 9 is a variable mean histogram provided in accordance with a second embodiment of the present invention; wherein, (a) is a torque and temperature 2 curve mean value combination distribution diagram, and (b) is a torque and temperature 5 curve mean value combination distribution diagram.
Detailed Description
To make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
First embodiment
The embodiment provides a sensitivity analysis method based on curve form change, which can be realized by an electronic device, and the electronic device can be a terminal or a server. The execution flow of the sensitivity analysis method based on the change of the curve form is shown in fig. 1, and the method comprises the following steps:
s1, acquiring process parameters and quality index data to be analyzed, and constructing an original sample data set; wherein, in the original sample data set, the process parameters correspond to the quality index data; the quality category of the sample data is divided into two categories of qualified quality and unqualified quality according to a preset quality index standard;
specifically, in this embodiment, the implementation process of S1 is as follows:
s11, acquiring a real-time process parameter curve generated by a preset process variable in the process of industrial production and final product inspection quality index data;
and S12, taking the product inspection and testing sampling time as a reference, intercepting process parameter curves in a period of time before and after the sampling time and corresponding to the product quality index data of the current sampling time, and dividing the data into two types of qualified data and unqualified data according to a preset quality index specification so as to construct the original sample data set.
S2, constructing a relation model of the process parameters and the quality index data;
specifically, in this embodiment, the implementation process of S2 is as follows:
s21, performing replacement sampling and SMOTE upsampling on the original sample data set by adopting a Bagging-SMOTE integrated sampling method to obtain a sample subset with balanced quality classes;
s22, training the sample subset by using a classification method of a support matrix machine to obtain each base model;
and S23, synthesizing training results of all the base models by adopting a voting method to obtain the relation model.
S3, selecting a reference sample in the original sample data set based on the relation model;
specifically, in this embodiment, the implementation process of S3 is as follows:
and performing quality type prediction on each sample in the original sample data set by using the relational model, and selecting the sample with the difference between the probability of qualified quality and the probability of unqualified quality smaller than a set threshold as a reference sample.
It should be noted that, the set threshold is related to the number of the base models, and generally should be greater than the quality class probability difference when the difference between the numbers of the base models of a certain sample that is determined to be qualified in quality and unqualified in quality is 1, so as to ensure the selection of the reference sample.
S4, converting the process parameters in the reference sample, and respectively predicting the quality categories of the reference sample before and after conversion by using the relation model to obtain the probability variation result of the corresponding quality category in the process of process parameter variation;
specifically, in this embodiment, the implementation process of S4 is as follows:
and sequentially selecting the reference sample, sequentially replacing corresponding process parameter curves in the reference sample by using the average value of all process parameter curves in the original sample data set, and respectively predicting the quality categories before and after replacement by using the relational model to obtain the probability change result of the corresponding quality categories in the process of changing the process parameter curves.
It should be noted that, since there may be a plurality of selected reference samples, each process parameter variable curve of each reference sample needs to be sequentially replaced by each process data curve in the process parameter data set to be analyzed in sequence based on the mean value.
And S5, measuring the sensitivity of the process parameters by using the gradient relation between the process parameter mean value change and the quality category probability change to obtain key process parameters influencing the product quality fluctuation.
Specifically, in this embodiment, the implementation process of S5 is as follows:
s51, extracting a quality class probability change result in the process of process parameter curve transformation in each reference sample, and measuring the sensitivity of the process parameters by the gradient of curve mean change and quality class probability change in the process of process parameter curve transformation;
s52, synthesizing the technological parameter sensitivity measurement results in the technological parameter curve transformation process of the multiple reference samples to obtain the sensitivity of the corresponding technological parameters;
s53, normalizing the process parameter sensitivity result to obtain the critical process parameters influencing the product quality fluctuation.
An exemplary graph of the normalized sensitivity analysis result is shown in fig. 2, and fig. 3 is a comparison graph of a key original process parameter curve between qualified and unqualified samples.
In summary, the sensitivity analysis method based on the curve form change provided by the embodiment solves the problem that the key factors causing the product quality fluctuation cannot be accurately identified in the process of the process industry, can well identify the key process parameters causing the product quality fluctuation in the process industry, and can quantify the influence degree of the process parameter change on the product quality through the sensitivity, thereby effectively improving the efficiency and the accuracy of identifying the key factors causing the product quality fluctuation in the process industry production process.
Second embodiment
In the embodiment, the sensitivity analysis method is applied to actual process data and quality inspection and test data generated in a certain flow industrial production process so as to verify and analyze the effectiveness of the method.
A material is an important precursor for producing a product, and the structure and quality of the material have a great influence on the quality of a final product, so that quality control of the material is an important issue for enterprises. During the production of such materials, abnormal changes in process parameters can lead to dramatic fluctuations in the quality of the material. If the main factors causing the product quality fluctuation cannot be found quickly and accurately, and the process parameters are adjusted in time, huge economic losses can be caused to enterprises. By adopting a sensitivity analysis method, main reasons influencing the product quality are discovered in time, and great help is provided for enterprises to establish process specifications, adjust process parameters in the production process in time and improve the product quality. In contrast, in this embodiment, the sensitivity analysis method of the present invention is used to analyze the sensitivity, and the main factors causing the product quality fluctuation are mined, which mainly includes the following steps:
the method comprises the steps of firstly, acquiring a real-time process parameter data set and final product inspection and test quality index data generated by preset process variables in the material production process, and intercepting process parameter curves and product quality index data in a period of time before and after the product inspection and test sampling time as a reference to correspond to each other; and dividing the data into qualified data and unqualified data according to the quality index specification to obtain process data and quality data to be analyzed. Then, a relation model between the process data and the quality data is constructed by using an integrated classification method based on a support matrix machine, and then, a reference sample is selected and the curve form change based on a mean value is carried out based on the relation model between the process data and the quality data to obtain a corresponding quality category probability change result in the process of process curve change; and finally, measuring the sensitivity of the process parameters by using the gradient relation between the process curve mean value change and the quality category probability change to obtain key process parameters causing product quality fluctuation. In the embodiment, a sensitivity method based on the change of the curve form is applied to a material production process, and key factors causing product quality fluctuation in the material production process are analyzed, wherein the analysis process is as follows:
for actual process data and a quality data set generated in a certain material production process in the process industry, the number of the base classifiers is determined to be 15, a threshold value is set to be 0.1 when a reference sample is selected, and the number of the reference samples is 9.
Based on the above, for the problem of analyzing the key factors causing the product quality fluctuation, the method of the present embodiment utilizes the sensitivity method based on the curve form change to perform the key factor analysis, and the specific calculation flow is shown in fig. 4, and includes the following steps:
1) According to the test sampling time t i Intercept T forward 1 Time period, backward intercept T 2 The process data and the quality data of the time period correspond to each other, the samples are divided into qualified quality and unqualified quality according to the quality index specification, the qualified quality and the unqualified quality are respectively represented by labels 0 and 1, and the tensor X belonging to the R of the process data to be analyzed is obtained N×J×T And the corresponding quality label vector Y ∈ R N ;
2) For the obtained process and quality data setsThe method comprises the steps of utilizing a Bagging-SMOTE algorithm to perform replaced sampling and SMOTE upsampling to balance the number of qualified and unqualified samples, then adopting a support matrix machine method to train a base classifier, and finally combining the base classifier into an integrated classification model through a voting method to represent the relationship between process data and quality data;
3) Traversing data setsSelecting a reference sample according to a set probability threshold value p to obtain a reference sample set Z belonging to R T×J×Q ;
4) Traversing the reference sample in the reference sample set, sequentially replacing the variable curve in the reference sample based on the mean value of the variable curve, and recording the quality probability class change in the sequential transformation process to obtain a probability class change tensor Y belonging to R Q×J×m ;
5) According to the relationship between the variation of variable curve matrix and the variation of gradient between probability classesThe sensitivity of the process parameters is measured, and the obtained gradient sensitivity measurement is normalized to obtain a sensitivity measurement vector S' belonging to R J . The flow of reference sample selection, mean-based curve morphology change, and sensitivity measurement is shown in fig. 5.
By specifically analyzing the analysis result of the key factors causing the product quality fluctuation in the material production process, the following conclusion can be drawn:
(1) As shown in fig. 6, torque, temperature 2 and temperature 5 are three variables with greater sensitivity, and are the main factors causing the quality fluctuation of the material.
(2) The sensitivity of the torque is the largest, and it can be seen from fig. 7 that the variation of the mean value of the torque curve causes the drastic variation of the quality failure probability, while the variation of the mean value of the curve of the feeding quantity 1 with smaller sensitivity only causes the slight variation of the quality failure probability; it is also seen from the torque curve distribution in fig. 8 that there is a significant difference in the distribution of the torque curves between the quality-qualified and the quality-unqualified samples; from fig. 9, (a) torque and temperature 2 curve mean combination distribution (b) torque and temperature 5 curve mean combination distribution there is also a distinct regional distribution between quality-qualified and quality-unqualified samples; it is shown that the method of the present embodiment analyzes the critical factors causing the product quality fluctuation accurately and effectively.
(3) The degree of influence of the process parameters on the quality fluctuation can be well represented by the normalized sensitivity value obtained by the method.
In conclusion, these results are consistent with the expected results, which show that the curve form change sensitivity-based analysis method provided by the invention is feasible and has good effect on practical industrial data.
Third embodiment
The embodiment provides a sensitivity analysis system based on curve form change, which comprises the following modules:
the process parameter and quality index data acquisition module is used for acquiring process parameters and quality index data to be analyzed and constructing an original sample data set; wherein, in the original sample data set, the process parameters correspond to the quality index data; dividing the quality category of the sample data into two categories of qualified quality and unqualified quality according to a preset quality index specification;
the relational model construction module is used for constructing a relational model of the process parameters and the quality index data;
a reference sample selection and data change module for selecting a reference sample in the original sample data set based on the relationship model; converting the process parameters in the reference sample, and respectively predicting the quality categories of the reference sample before and after conversion by using the relational model to obtain the probability variation result of the corresponding quality category in the process of process parameter variation;
and the sensitivity measurement module is used for measuring the sensitivity of the process parameters by utilizing the gradient relation between the process parameter mean value change and the quality category probability change to obtain the key process parameters influencing the product quality fluctuation.
The sensitivity analysis system based on the change of the curve form of the present embodiment corresponds to the sensitivity analysis method based on the change of the curve form of the first embodiment; the functions realized by the functional modules in the sensitivity analysis system based on the curve form change of the embodiment correspond to the flow steps in the sensitivity analysis method based on the curve form change of the first embodiment one by one; therefore, it is not described herein.
Fourth embodiment
The present embodiment provides an electronic device, which includes a processor and a memory; wherein the memory has stored therein at least one instruction that is loaded and executed by the processor to implement the method of the first embodiment.
The electronic device may generate a large difference due to different configurations or performances, and may include one or more processors (CPUs) and one or more memories, where at least one instruction is stored in the memory, and the instruction is loaded by the processor and performs the following steps:
s1, acquiring process parameters and quality index data to be analyzed, and constructing an original sample data set; wherein, in the original sample data set, the process parameters correspond to the quality index data; the quality category of the sample data is divided into two categories of qualified quality and unqualified quality according to a preset quality index standard;
s2, constructing a relation model of the process parameters and the quality index data;
s3, selecting a reference sample in the original sample data set based on the relation model;
s4, converting the process parameters in the reference sample, and respectively predicting the quality categories of the reference sample before and after conversion by using the relation model to obtain the probability variation result of the corresponding quality category in the process of process parameter variation;
and S5, measuring the sensitivity of the process parameters by using the gradient relation between the process parameter mean value change and the quality category probability change to obtain key process parameters influencing the product quality fluctuation.
The electronic equipment of the embodiment solves the problem that the key factors causing the product quality fluctuation can not be accurately identified in the process of the process industry by executing the method, can well identify the key process parameters causing the product quality fluctuation in the process of the process industry, and can quantify the influence degree of the process parameter change on the product quality through the sensitivity, thereby effectively improving the efficiency and the accuracy of identifying the key factors causing the product quality fluctuation in the process industry production process.
Fifth embodiment
The present embodiments provide a computer-readable storage medium having at least one instruction stored therein, the instruction being loaded and executed by a processor to implement the above-mentioned method. The computer readable storage medium may be, among others, ROM, random Access Memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like. The instructions stored therein may be loaded by a processor in the terminal and perform the following steps:
s1, acquiring process parameters and quality index data to be analyzed, and constructing an original sample data set; wherein, in the original sample data set, the process parameters correspond to the quality index data; the quality category of the sample data is divided into two categories of qualified quality and unqualified quality according to a preset quality index standard;
s2, constructing a relation model of the process parameters and the quality index data;
s3, selecting a reference sample in the original sample data set based on the relation model;
s4, converting the process parameters in the reference sample, and respectively predicting the quality categories of the reference sample before and after conversion by using the relation model to obtain the probability variation result of the corresponding quality category in the process of process parameter variation;
and S5, measuring the sensitivity of the process parameters by using the gradient relation between the process parameter mean value change and the quality category probability change to obtain key process parameters influencing the product quality fluctuation.
The method stored by the storage medium of the embodiment solves the problem that the key factors causing the product quality fluctuation can not be accurately identified in the process of the process industry, can well identify the key process parameters causing the product quality fluctuation in the process of the process industry, and can quantify the influence degree of the process parameter change on the product quality through the sensitivity, thereby effectively improving the efficiency and the accuracy of identifying the key factors causing the product quality fluctuation in the process industry production process.
Furthermore, it should be noted that the present invention may be provided as a method, apparatus or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present invention may take the form of a computer program product embodied on one or more computer-usable storage media having computer-usable program code embodied in the media.
Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, embedded processor, or other programmable data processing terminal to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing terminal to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks. These computer program instructions may also be loaded onto a computer or other programmable data processing terminal to cause a series of operational steps to be performed on the computer or other programmable terminal to produce a computer implemented process such that the instructions which execute on the computer or other programmable terminal provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It is further noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrases "comprising one of \ 8230; \8230;" does not exclude the presence of additional like elements in a process, method, article, or terminal device that comprises the element.
Finally, it should be noted that while the above describes a preferred embodiment of the invention, it will be appreciated by those skilled in the art that, once the basic inventive concepts have been learned, numerous changes and modifications may be made without departing from the principles of the invention, which shall be deemed to be within the scope of the invention. Therefore, it is intended that the appended claims be interpreted as including the preferred embodiment and all changes and modifications that fall within the true scope of the embodiments of the present invention.
Claims (8)
1. A sensitivity analysis method based on curve morphological change is characterized by comprising the following steps:
acquiring process parameters and quality index data to be analyzed, and constructing an original sample data set; wherein, in the original sample data set, the process parameters correspond to the quality index data; the quality category of the sample data is divided into two categories of qualified quality and unqualified quality according to a preset quality index standard;
constructing a relation model of the process parameters and the quality index data;
selecting a reference sample in the original sample data set based on the relationship model;
converting the process parameters in the reference sample, and respectively predicting the quality categories of the reference sample before and after conversion by using the relational model to obtain the probability variation result of the corresponding quality category in the process of process parameter variation;
and measuring the sensitivity of the process parameters by using the gradient relation between the process parameter mean value change and the quality category probability change to obtain key process parameters influencing the product quality fluctuation.
2. The method for sensitivity analysis based on curve morphological change of claim 1, wherein the obtaining of the process parameters and quality index data to be analyzed and the construction of the original sample data set comprise:
acquiring a real-time process parameter curve generated by a preset process variable in the process of industrial production and final product inspection and test quality index data;
and taking the product inspection and testing sampling time as a reference, intercepting process parameter curves in a period of time before and after the sampling time and corresponding to the product quality index data of the current sampling time, and dividing the data into two types of qualified quality and unqualified quality according to a preset quality index specification so as to construct the original sample data set.
3. The method for analyzing sensitivity based on curve morphological change of claim 1, wherein the establishing of the relational model of the process parameters and the quality index data comprises:
performing replacement sampling and SMOTE up-sampling on the original sample data set by adopting a Bagging-SMOTE integrated sampling method to obtain a sample subset with balanced quality categories;
training the sample subset by using a classification method of a support matrix machine to obtain each base model;
and synthesizing training results of all the base models by adopting a voting method to obtain the relational model.
4. The method for sensitivity analysis based on curve morphological change according to claim 1, wherein the selecting a reference sample in the original sample data set based on the relational model comprises:
and performing quality type prediction on each sample in the original sample data set by using the relational model, and selecting the sample with the difference between the probability of qualified quality and the probability of unqualified quality smaller than a set threshold as a reference sample.
5. The method for sensitivity analysis based on curve form change according to claim 4, wherein the set threshold is larger than the difference of the probability of the quality class when the number of the base models of which the sample is judged to be qualified and unqualified in all the base models is 1.
6. The method for sensitivity analysis based on curve form change according to claim 2, wherein the step of transforming the process parameters in the reference sample, and the step of predicting the quality classes of the reference sample before and after transformation by using the relationship model to obtain the probability change results of the corresponding quality classes in the process of process parameter change comprises the steps of:
and sequentially selecting the reference sample, sequentially replacing corresponding process parameter curves in the reference sample by using the average value of all process parameter curves in the original sample data set, and respectively predicting the quality categories before and after replacement by using the relational model to obtain the probability change result of the corresponding quality categories in the process of changing the process parameter curves.
7. The method of claim 6, wherein the step of measuring the sensitivity of the process parameter using the gradient relationship between the mean variation of the process parameter and the probability variation of the quality class to obtain the critical process parameter affecting the quality fluctuation of the product comprises:
extracting a quality class probability change result in the process of changing the process parameter curve in each reference sample, and measuring the sensitivity of the process parameter by the gradient of the curve mean change and the quality class probability change in the process of changing the process parameter curve;
synthesizing the technological parameter sensitivity measurement results in the technological parameter curve transformation process of the multiple reference samples to obtain the sensitivity of the corresponding technological parameters;
and normalizing the sensitivity result of the process parameters to obtain key process parameters influencing the quality fluctuation of the product.
8. A sensitivity analysis system based on changes in curve morphology, comprising:
the process parameter and quality index data acquisition module is used for acquiring process parameters and quality index data to be analyzed and constructing an original sample data set; wherein, in the original sample data set, the process parameters correspond to the quality index data; the quality category of the sample data is divided into two categories of qualified quality and unqualified quality according to a preset quality index standard;
the relational model construction module is used for constructing a relational model of the process parameters and the quality index data;
a reference sample selection and data change module for selecting a reference sample in the original sample data set based on the relationship model; converting the process parameters in the reference sample, and respectively predicting the quality categories of the reference sample before and after conversion by using the relational model to obtain the probability variation result of the corresponding quality category in the process of process parameter variation;
and the sensitivity measurement module is used for measuring the sensitivity of the process parameters by utilizing the gradient relation between the process parameter mean value change and the quality category probability change to obtain the key process parameters influencing the product quality fluctuation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210693186.8A CN115169832A (en) | 2022-06-17 | 2022-06-17 | Sensitivity analysis method and system based on curve form change |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210693186.8A CN115169832A (en) | 2022-06-17 | 2022-06-17 | Sensitivity analysis method and system based on curve form change |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115169832A true CN115169832A (en) | 2022-10-11 |
Family
ID=83484733
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210693186.8A Pending CN115169832A (en) | 2022-06-17 | 2022-06-17 | Sensitivity analysis method and system based on curve form change |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115169832A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115439021A (en) * | 2022-10-26 | 2022-12-06 | 江苏新恒基特种装备股份有限公司 | Metal strengthening treatment quality analysis method and system |
-
2022
- 2022-06-17 CN CN202210693186.8A patent/CN115169832A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115439021A (en) * | 2022-10-26 | 2022-12-06 | 江苏新恒基特种装备股份有限公司 | Metal strengthening treatment quality analysis method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20230333533A1 (en) | System and method for generating machine learning model with trace data | |
JP4394728B2 (en) | Influence factor identification device | |
CN114444986B (en) | Product analysis method, system, device and medium | |
CN116188475A (en) | Intelligent control method, system and medium for automatic optical detection of appearance defects | |
CN113609790B (en) | Product virtual measuring method, system, device and medium | |
CN110969600A (en) | Product defect detection method and device, electronic equipment and storage medium | |
CN113642666B (en) | Active enhancement soft measurement method based on sample expansion and screening | |
JP2023511464A (en) | SYSTEM, METHOD, AND COMPUTER-READABLE MEDIUM FOR ANALYSIS OF PRODUCT FAILURE FACTORS | |
CN116416884B (en) | Testing device and testing method for display module | |
CN115169832A (en) | Sensitivity analysis method and system based on curve form change | |
Saadallah et al. | Early quality prediction using deep learning on time series sensor data | |
CN116653346A (en) | Online monitoring method and tool for forming quality of stamping workpiece | |
CN115270861A (en) | Product composition data monitoring method and device, electronic equipment and storage medium | |
CN115436342A (en) | Method and device for reducing LIBS detection uncertainty among multiple batches of samples | |
CN115686995A (en) | Data monitoring processing method and device | |
CN114972273A (en) | Method, system, device and storage medium for enhancing data set of streamlined product | |
CN114334033A (en) | Screening method, system and terminal for molecular descriptors of anti-breast cancer candidate drugs | |
CN115667915A (en) | Root cause analysis based on wafer bin maps | |
Jacobs et al. | Performance, successes and limitations of deep learning semantic segmentation of multiple defects in transmission electron micrographs | |
CN113128734A (en) | Method and device for predicting oil field yield | |
CN111400644A (en) | Calculation processing method for laboratory analysis sample | |
CN118037137B (en) | Method for determining product quality accident number based on convolutional neural network | |
TWI777169B (en) | Quality control method, device, computer device and storage media | |
CN116773544B (en) | Steel structure plate fault detection method and system based on deep learning | |
CN117076454B (en) | Engineering quality acceptance form data structured storage 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 |