CN115270861A - Product composition data monitoring method and device, electronic equipment and storage medium - Google Patents
Product composition data monitoring method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN115270861A CN115270861A CN202210857179.7A CN202210857179A CN115270861A CN 115270861 A CN115270861 A CN 115270861A CN 202210857179 A CN202210857179 A CN 202210857179A CN 115270861 A CN115270861 A CN 115270861A
- Authority
- CN
- China
- Prior art keywords
- sample set
- component data
- product
- control
- data
- 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 42
- 238000012544 monitoring process Methods 0.000 title claims abstract description 27
- 238000003860 storage Methods 0.000 title claims abstract description 16
- 230000002159 abnormal effect Effects 0.000 claims abstract description 54
- 239000013598 vector Substances 0.000 claims abstract description 49
- 230000003044 adaptive effect Effects 0.000 claims abstract description 18
- 238000012549 training Methods 0.000 claims abstract description 18
- 238000012216 screening Methods 0.000 claims abstract description 7
- 238000004422 calculation algorithm Methods 0.000 claims description 40
- 238000004519 manufacturing process Methods 0.000 claims description 16
- 238000012706 support-vector machine Methods 0.000 claims description 11
- 230000009466 transformation Effects 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 6
- 238000012806 monitoring device Methods 0.000 claims description 6
- 238000000342 Monte Carlo simulation Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 2
- 238000004088 simulation Methods 0.000 claims 1
- 230000008569 process Effects 0.000 description 12
- 238000010586 diagram Methods 0.000 description 9
- 238000009826 distribution Methods 0.000 description 5
- 238000003070 Statistical process control Methods 0.000 description 3
- 239000004615 ingredient Substances 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 2
- 239000000654 additive Substances 0.000 description 2
- 230000000996 additive effect Effects 0.000 description 2
- 238000012804 iterative process Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000007704 transition Effects 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
- G06Q10/06393—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)
- Economics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Educational Administration (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Manufacturing & Machinery (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention provides a method and a device for monitoring product component data, electronic equipment and a storage medium, wherein the method comprises the following steps: firstly, screening out an out-of-control component data sample set in a control chart sample set based on a preset condition; then training an initial adaptive enhancement-based vector machine model according to the uncontrolled component data sample set to obtain a corresponding abnormal offset mode; and finally, determining the abnormal offset mode of the product to be tested according to a vector machine model which is complete in training and is based on self-adaption enhancement. The product component data is converted into the control chart sample set for judgment, so that the result is concise and clear, and the traceability is facilitated; after the out-of-control component statistic is obtained, the abnormal offset mode can be determined based on the vector machine model with the adaptive enhancement, so that the abnormal data can be automatically processed, and the abnormal mode of the product can be determined.
Description
Technical Field
The invention relates to the field of data monitoring, in particular to a product composition data monitoring method and device, electronic equipment and a storage medium.
Background
In the quality management of the manufacturing production Process, a Control chart is often used as a tool for Statistical Process Control (SPC) to monitor abnormal fluctuation in the Process, and the method mainly comprises the steps of performing online real-time monitoring on the production Process of a product by applying mathematical statistical analysis, judging whether the production Process is in a controlled or out-of-Control state by monitoring the quality characteristic change of the product, and giving an alarm once the production Process is out of Control, so that a producer can timely find abnormality and adjust the abnormality so as to achieve the purpose of controlling the production quality.
In actual manufacturing scenes, such as chemical and food industries, it is often necessary to control the proportion, concentration, and the like of various quality characteristics of products, and such quality characteristics are ingredient data. The quality of the finished product is greatly influenced by the distribution of the component data, so that the distribution state of the component data is often monitored by using a multivariate control chart, and the production quality is controlled.
However, the conventional statistical process control can only monitor the state in the production process, cannot further identify the abnormal mode according to the monitoring data, and needs to manually check the abnormal source. Therefore, the prior art has the problems that abnormal data cannot be automatically processed in the process of monitoring the product components, and an abnormal mode cannot be determined.
Disclosure of Invention
In view of the above, it is desirable to provide a method, an apparatus, an electronic device and a storage medium for monitoring product composition data, which can effectively monitor a specific abnormal pattern type corresponding to out-of-control data in the product composition data.
In order to solve the above problems, the present invention provides a product composition data monitoring method, including:
acquiring a control chart sample set corresponding to product component data, and screening out an out-of-control component data sample set based on a preset condition;
inputting the out-of-control component data sample set into an initial vector machine model based on self-adaptive enhancement, taking an abnormal offset mode corresponding to the out-of-control component data sample set as output, performing iterative training, and determining a vector machine model based on self-adaptive enhancement with complete training;
and acquiring real-time out-of-control component data, inputting the real-time out-of-control component data into a vector machine model which is completely trained and based on self-adaption enhancement, and determining the abnormal offset mode of the product to be detected.
Further, the initial vector machine model based on the adaptive enhancement comprises a plurality of weak classifiers, and the weighting operation is carried out on the weak classifiers through the adaptive enhancement algorithm to form a strong classifier; and mapping the data of the strong classifier into a high-dimensional feature space through a support vector machine algorithm for classification.
Further, the support vector machine algorithm comprises a nuclear parameter and a penalty factor, and the nuclear parameter and the penalty factor are optimized through a differential evolution algorithm.
Further, the performance indicators controlling the sample set of maps include an average run length under a controlled condition and an average run length under an uncontrolled condition.
Further, solving the average running length under the controlled state through a Markov chain algorithm; and solving the average running length in an out-of-control state by a Monte Carlo method.
Further, the step of obtaining a sample set of control charts includes:
acquiring product component data;
carrying out equidistant logarithmic ratio transformation on the product component data samples to determine a mean value coordinate vector sample;
and determining a statistic sample according to the mean coordinate vector sample, and constructing a control chart sample set.
Further, obtaining product composition data, comprising:
and performing wavelet transformation denoising processing on the collected product production data to obtain product composition data.
In order to solve the above problems, the present invention also provides a product composition data monitoring device, including:
the system comprises an uncontrolled component data sample set acquisition module, a control image data sample set generation module and a control component data sample set selection module, wherein the uncontrolled component data sample set acquisition module is used for acquiring a control image sample set corresponding to product component data and screening out an uncontrolled component data sample set based on preset conditions;
the model training module is used for inputting the out-of-control component data sample set into the initial vector machine model based on the self-adaptive enhancement, taking the abnormal offset mode corresponding to the out-of-control component data sample set as output, performing iterative training, and determining the vector machine model based on the self-adaptive enhancement with complete training;
and the abnormal offset mode determining module is used for acquiring real-time out-of-control component data, inputting the real-time out-of-control component data into a vector machine model which is completely trained and based on self-adaptive enhancement, and determining the abnormal offset mode of the product to be detected.
In order to solve the above problem, the present invention further provides an electronic device, which includes a processor and a memory, where the memory stores a computer program, and when the computer program is executed by the processor, the method for monitoring product composition data as described above is implemented.
In order to solve the above problems, the present invention also provides a computer-readable storage medium storing computer program instructions, which, when executed by a computer, cause the computer to execute the product composition data monitoring method as described above.
The beneficial effect of adopting above-mentioned technical scheme is: the invention provides a method and a device for monitoring product component data, electronic equipment and a storage medium, wherein the method comprises the following steps: firstly, screening out an out-of-control component data sample set in a control chart sample set based on a preset condition; then, training an initial adaptive enhancement-based vector machine model according to the uncontrolled component data sample set to obtain a corresponding abnormal offset mode; and finally, determining the abnormal offset mode of the product to be detected according to the vector machine model which is completely trained and is based on the self-adaptive enhancement. The product component data is converted into the control chart sample set for judgment, so that the result is concise and clear, and the traceability is facilitated; after the out-of-control component statistic is obtained, the abnormal offset mode can be determined based on the vector machine model with the adaptive enhancement, so that the abnormal data can be automatically processed, and the abnormal mode of the product can be determined.
Drawings
FIG. 1 is a schematic flow chart diagram illustrating an embodiment of a method for monitoring product composition data according to the present invention;
FIG. 2 is a schematic flow diagram of one embodiment of a control diagram configured to control score data provided by the present invention;
FIG. 3 is a schematic structural diagram of an embodiment of an initial adaptive enhancement-based vector machine model provided in the present invention;
FIG. 4 is a schematic structural diagram of an embodiment of a product composition data monitoring device according to the present invention;
fig. 5 is a block diagram of an embodiment of an electronic device provided in the present invention.
Detailed Description
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate preferred embodiments of the invention and together with the description, serve to explain the principles of the invention and not to limit the scope of the invention.
Before the examples are set forth, the abnormal pattern of the product ingredients is explained:
taking the ternary process as an example, because the composition data of a product has a definite sum limit, where a shift in one variable causes a shift in the other variable, there is no case where only a single variable is shifted.
For the ternary process, i.e. for products involving only three components, there are mainly several abnormal patterns: the case where both the first and second variables are shifted, both the first and third variables are shifted, both the second and third variables are shifted, and both the three variables are shifted. The 12 possible abnormal patterns are: (1, -1, 0), (1, 0, -1), (0, 1, -1), (-1, 0), (-1, 0, 1), (0, -1, 1); (1, -1, -1), (-1, -1), (1, -1, 1), (1, -1), (1, -1, 1), (-1, 1). Where normal and abnormal states are represented by 0 and + -1, respectively, + represents a shift of the variable up and-represents a shift of the variable down.
The specific steps are as follows:
abnormal mode | Abnormal offset pattern |
(1,-1,0) | Component 1 offset up and component 2 offset down |
(1,0,-1) | Component 1 offset up and component 3 offset down |
(0,1,-1) | Component 2 offset up and component 3 offset down |
(-1,1,0) | Offset below component 1 and offset above component 2 |
(-1,0,1) | Offset below component 1 and offset above component 3 |
(0,-1,1) | Offset down for component 2 and offset up for component 3 |
(1,-1,-1) | Component 1 offset up, component 2 offset down, and component 3 offset down |
(-1,1,-1) | Offset down for component 1, offset up for component 2, and offset down for component 3 |
(1,-1,1) | Component 1 offset up, component 2 offset down, and component 3 offset up |
(1,1,-1) | Component 1 offset up, component 2 offset up, and component 3 offset down |
(1,-1,1) | Offset up component 1, offset down component 2, and offset up component 3 |
(-1,1,1) | Offset down for component 1, offset up for component 2, and offset up for component 3 |
Currently, in order to control the ratio, concentration, etc. of various quality characteristics of a product, the ingredient data of the product is generally monitored. However, the existing multivariate control chart has low recognition accuracy, and is difficult to automatically process abnormal data according to production data, let alone determine the abnormal mode of a product according to the abnormal production data.
Therefore, in the process of monitoring the product components, the prior art has the problem that the corresponding specific abnormal mode cannot be determined according to the out-of-control data in the product component data.
In order to solve the above problems, the present invention provides a method and an apparatus for monitoring product composition data, an electronic device, and a storage medium, which will be described in detail below.
As shown in fig. 1, fig. 1 is a schematic flow chart of an embodiment of a product composition data monitoring method provided by the present invention, including:
step S101: and acquiring a control chart sample set corresponding to the product component data, and screening out an out-of-control component data sample set based on a preset condition.
Step S102: inputting the out-of-control component data sample set into the initial vector machine model based on the adaptive enhancement, taking the abnormal offset mode corresponding to the out-of-control component data sample set as output, performing iterative training, and determining the vector machine model based on the adaptive enhancement with complete training.
Step S103: and acquiring real-time out-of-control component data, inputting the real-time out-of-control component data into a vector machine model which is completely trained and based on self-adaption enhancement, and determining the abnormal offset mode of the product to be detected.
In this embodiment, first, based on a preset condition, a runaway component data sample set of a product is obtained according to a control chart sample set, and then, an abnormal offset mode corresponding to the runaway component data sample set is determined through a vector machine model based on adaptive enhancement. The product component data is judged through the control chart to obtain an out-of-control component data sample set, so that the result is concise and clear, and the traceability is facilitated; after the uncontrolled component data sample set is obtained, the abnormal offset mode can be determined based on the vector machine model which is enhanced in a self-adaptive mode, the abnormal data are automatically processed, and the abnormal offset mode of the product is automatically determined according to the abnormal data.
As a preferred embodiment, in step S101, in order to obtain multiple sets of product composition data and ensure the accuracy of the product composition data, the collected product production data needs to be preprocessed to remove abnormal values.
In a specific embodiment, wavelet transformation denoising processing is performed on the collected product production data, so that product composition data is obtained.
When the signals and the noise in the production data of the product are subjected to wavelet decomposition under different scales through the wavelet transformation denoising processing, the transfer characteristics shown by the signals and the noise are opposite, namely the modulus maximum of the noise is reduced along with the increase of the wavelet scale, and the modulus maximum of the signals is increased along with the increase of the scale. By the characteristic, the noise part in the signal can be removed, and then the original signal is reconstructed from the denoised module maximum value, so that the purpose of removing the noise is achieved, and the product composition data is more accurate.
Further, after the product composition data is obtained, a sample set of control charts is acquired.
In an embodiment, as shown in fig. 2, fig. 2 is a schematic flowchart of an embodiment of obtaining a control chart sample set provided in the present application, including:
step S111: product composition data is obtained.
In one embodiment, random sampling is used to extract p-dimensional component data with a sample size of n, denoted as Xi,j={Xi,1,Xi,2,L,Xi,n}; wherein each X isi,jAre subject to a normal distribution.
Step S112: and (5) carrying out equidistant logarithmic ratio transformation on the product component data samples to determine a mean value coordinate vector sample.
In a specific embodiment, each component data is subjected to equidistant logarithmic ratio transformation to obtain a mean coordinate vector, wherein a formula for calculating the mean coordinate vector is as follows:
wherein, Zi,jObeying an additive logistic normal distribution.
Step S113: and determining a statistic sample according to the mean coordinate vector sample, and constructing a control chart sample set.
In a specific embodiment, processing the mean coordinate vector through a multivariate cumulative sum algorithm to obtain a statistic sample; and constructing a control chart sample set according to the statistic sample.
Wherein, the formula for calculating the statistic sample is as follows:
Ci={(Yi-1+Zi-μ0)'Σ0 -1(Yi-1+Zi-μ0)}1/2
Qi=[Yi'Σ0 -1Yi]1/2
wherein QiFor the statistical sample, k is the process offset coefficient, ZiIs a mean coordinate vector of the samples, Zi∈Rp-1,i=1,2,L。
In the embodiment, the product component data is subjected to equidistant logarithmic ratio transformation, so that the product component data obeys additive logic normalization, and the characteristic of the component data can be reflected by the mean coordinate vector; and obtaining a statistic sample of the product through multivariate accumulation and algorithm, and constructing a control chart sample set so as to conveniently and intuitively obtain a monitoring result.
Further, controlling the performance indicators of the chart sample set includes: average run length under controlled conditions and average run length under out-of-control conditions. The larger the average running length in the out-of-control state is, the more effective the false alarm rate of the control chart sample set can be reduced, and the smaller the average running length in the controlled state is, the more effective the false alarm rate of the control chart sample set can be reduced.
In order to obtain the optimal value of the average run length under the controlled state, a Markov chain is used for solving.
In one embodiment, the average run length under controlled conditions is given by the formula:
ARL0=h'(I-P)-1E
wherein, ARL0Is the average run length under control, h = (1, 0, \8230;) is the initial vector of dimension m +1, I is the identity matrix of (m +1 ), P is the probability transition matrix of (m +1 ), and E is the column vector of m + 1.
In order to obtain the optimal value of the average running length in the out-of-control state, a Monte Carlo method is used for solving.
In one embodiment, the average run chain length ARL under controlled conditions is set0Selecting a set of parameters k and H of a variable accumulation and control graph based on the component data; generating a random number which follows normal distribution and has an offset coefficient delta; calculating statistic sample Q according to given parameter ki(ii) a Recording the running chain length of the out-of-control state; repeat the above step 104Then, 10 is obtained4Secondary RL, calculating the expected value of RL, namely the average running length in the out-of-control state, and recording as ARL1。
In the embodiment, the optimal value of the average running length in the controlled state and the optimal value of the average running length in the uncontrolled state are obtained through calculation, so that the performance of the control chart is ensured, and the reliability of the result is improved.
Further, in step S101, after the control chart sample set is obtained, the uncontrolled component data sample set is screened out based on a preset condition.
In one embodiment, the initial controlled average chain length is set based on experience or historical records; then, the actual value of the controlled average chain length is solved by using the markov chain, the offset reference coefficient k =1/2 × δ is taken, and the non-linear equation of the actual value of the controlled average chain length in the markov chain is solved by using the fsolve function in Matlab software, so as to obtain the control limit at this time, which is the preset condition in the embodiment.
Further, after the preset condition is determined, the control chart sample set and the preset condition are subjected to traversal comparison, and if the control chart sample set is smaller than the preset condition, the control chart sample set is judged to be controlled; if the control chart is larger than the preset condition, the control chart sample set is judged to be out of control, recorded as an out-of-control component data sample set and recorded.
By the mode, the component data are distinguished, the abnormal component data are extracted, the abnormal component data are combed, and a component data control chart is generated, so that the component data control chart is convenient to use later.
Further, in step S102, after the uncontrolled component data sample set is determined, the uncontrolled component data sample set needs to be analyzed to determine an abnormal offset pattern corresponding to the uncontrolled component data sample set.
As a preferred embodiment, in step S102, as shown in fig. 3, fig. 3 is a schematic structural diagram of an embodiment of an initial adaptive enhancement based vector machine model provided by the present invention, where the initial adaptive enhancement based vector machine model includes a plurality of weak classifiers (A1, A2, A3, A4) and a strong classifier (a) composed of the weak classifiers.
In this embodiment, a plurality of weak classifiers (A1, A2, A3, and A4) are subjected to weighting operation by an adaptive enhancement algorithm to form a strong classifier (a), and then, data of the strong classifier is mapped to a high-dimensional feature space by a support vector machine algorithm for classification.
In one embodiment, the adaptive boosting algorithm weights the weak classifiers to form a strong classifier, and thus, the performance of the strong classifier depends on the classification effect of the weak classifier. The support vector machine algorithm maps low-dimensional data to a high-dimensional feature space for classification, and the problem that the error of the self-adaptive enhancement algorithm is increased along with the increase of iteration times is solved to a certain extent, so that the classification accuracy of the classification algorithm can be improved by selecting the support vector machine algorithm as a weak classifier of the self-adaptive enhancement algorithm.
That is to say, an integrated algorithm is formed by fusing the adaptive enhancement algorithm and the support vector machine algorithm, and the two algorithms are made up for the deficiency, so that the performance of the algorithm is improved.
In a specific embodiment, the support vector machine algorithm comprises kernel parameters and penalty factors, and if the kernel parameter values are not properly selected, an 'over-learning' or 'under-learning' phenomenon occurs; if the selected sub-penalty factors are not appropriate, an "over-fit" or "under-fit" result will occur. In order to improve the performance of the support vector machine algorithm and ensure the reliability of the result, the optimal kernel parameter and penalty factor need to be selected.
Further, in order to optimize the nuclear parameters and the penalty factors, a differential evolution algorithm is used for processing.
The differential evolution algorithm is a heuristic algorithm for simulating the natural biological evolution process, continuously approaches the optimal solution in the iterative process by randomly searching the difference in the population, and finally achieves the global convergence. The method is mainly applied to solving the global optimal solution or the problem of forming a hybrid algorithm by combining with other algorithms at present. The main idea is to take the vector difference of two individuals in a randomly selected group and the third individual as a variant individual, cross the variant individual and a target individual to form a new individual and compare the new individual with the target individual, and select a better individual to enter a next generation group. And continuously improving the quality of the population to be close to the optimal solution in the iterative process.
The differential evolution algorithm mainly solves a global optimal solution through three steps of variation, intersection and selection. The variation is to obtain a new individual vector by three individual vectors, and if the difference between the first individual and the second individual is smaller, the influence on the new individual is smaller. In the initial stage of algorithm iteration, the larger the random search range is due to the overlarge difference between every two algorithms; in the middle and later stages of the algorithm iteration, the difference between every two algorithms is reduced, the searching range is reduced, and the algorithm tends to an optimal value. The crossing is to improve the diversity of population individuals, and the variant individuals and the target individuals are crossed to form new test individuals in the crossing process, so that the test individuals can obtain at least one value in the variant individuals, and invalid crossing is avoided. The selection aims to retain the best individual, and the fitness value is compared based on a greedy selection idea, so that the better individual in the test individual and the target individual is retained, and the fitness value of the offspring individual is always better than that of the parent individual.
The optimal kernel parameters and the optimal punishment factors are determined through the differential evolution algorithm, the reliability of the support vector machine algorithm is improved, the judgment precision of the vector machine model based on self-adaption enhancement determined according to the support vector machine algorithm is guaranteed, and therefore the reliability of the obtained abnormal offset mode is effectively guaranteed.
By the mode, firstly, the product component data is converted into the control chart sample set, and the component data is graphed, so that the intuitiveness of the result is improved; then, based on preset conditions, judging whether the product components are controllable according to the control chart sample set, so as to realize primary monitoring on the product components; further, for the uncontrolled component data sample set, the corresponding abnormal offset mode is obtained according to the vector machine model based on the self-adaptive enhancement, so that the abnormal data are automatically processed according to the component data of the product, and the abnormal offset mode is determined.
In order to solve the above problem, the present invention further provides a product composition data monitoring device, as shown in fig. 4, fig. 4 is a schematic structural diagram of an embodiment of the product composition data monitoring device provided by the present invention, and the product composition data monitoring device 400 includes:
the uncontrolled component data sample set acquisition module 401 is configured to acquire a control chart sample set corresponding to product component data, and screen out an uncontrolled component data sample set based on a preset condition;
a model training module 402, configured to input the uncontrolled component data sample set to an initial adaptive enhancement-based vector machine model, perform iterative training with an abnormal offset mode corresponding to the uncontrolled component data sample set as an output, and determine a self-adaptively enhancement-based vector machine model with complete training;
and the abnormal offset mode determining module 403 is configured to obtain real-time out-of-control component data, input the real-time out-of-control component data into a vector machine model based on adaptive enhancement, which is completely trained, and determine an abnormal offset mode of the product to be tested.
The present invention also provides an electronic device, as shown in fig. 5, fig. 5 is a block diagram of an embodiment of the electronic device provided in the present invention. The electronic device 500 may be a computing device such as a mobile terminal, a desktop computer, a notebook, a palmtop, and a server. The electronic device 500 includes a processor 501 and a memory 502, wherein the memory 502 has a product composition data monitoring program 503 stored thereon.
The memory 502 may be an internal storage unit of the computer device, such as a hard disk or a memory of the computer device, in some embodiments. The memory 502 may also be an external storage device of the computer device in other embodiments, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), etc. provided on the computer device. Further, the memory 502 may also include both internal storage units and external storage devices of the computer device. The memory 502 is used for storing application software installed on the computer device and various data, such as program codes for installing the computer device. The memory 502 may also be used to temporarily store data that has been output or is to be output. In one embodiment, the product composition data monitoring program 503 may be executed by the processor 501 to implement the product composition data monitoring method according to the embodiments of the present invention.
The embodiment further provides a computer-readable storage medium, on which a product composition data monitoring program is stored, and when the program is executed by a processor, the computer implements the product composition data monitoring method according to any one of the above technical solutions.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), rambus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
While the invention has been described with reference to specific preferred embodiments, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the following claims.
Claims (10)
1. A method for monitoring product composition data, comprising:
acquiring a control chart sample set corresponding to product component data, and screening out an out-of-control component data sample set based on a preset condition;
inputting the out-of-control component data sample set into an initial vector machine model based on self-adaptive enhancement, taking an abnormal offset mode corresponding to the out-of-control component data sample set as output, performing iterative training, and determining a vector machine model based on self-adaptive enhancement with complete training;
and acquiring real-time out-of-control component data, inputting the real-time out-of-control component data into the vector machine model which is completely trained and based on self-adaptive enhancement, and determining the abnormal offset mode of the product to be detected.
2. The method according to claim 1, wherein the initial adaptively enhanced-based vector machine model comprises a plurality of weak classifiers, and the weak classifiers are weighted by an adaptive enhancement algorithm to form a strong classifier; and mapping the data of the strong classifier to a high-dimensional feature space for classification through a support vector machine algorithm.
3. The method of claim 2, wherein the SVM algorithm includes a kernel parameter and a penalty factor, and wherein the kernel parameter and the penalty factor are optimized by a differential evolution algorithm.
4. The method of claim 1, wherein the performance indicators of the control chart sample set include an average run length under controlled conditions and an average run length under out-of-control conditions.
5. The method of claim 4, wherein the average run length under controlled conditions is solved by a Markov chain algorithm; and solving the average running length under the out-of-control state by a Monte Carlo method simulation method.
6. The method for monitoring product composition data of claim 1, wherein the step of obtaining said control chart sample set comprises:
obtaining the product composition data;
carrying out equidistant logarithmic ratio transformation on the product component data samples to determine a mean coordinate vector sample;
and determining a statistic sample according to the mean coordinate vector sample, and constructing a control chart sample set.
7. The method of monitoring product composition data of claim 6, wherein said obtaining said product composition data comprises:
and performing wavelet transformation denoising processing on the collected product production data to obtain the product component data.
8. A product composition data monitoring device, comprising:
the system comprises an uncontrolled component data sample set acquisition module, a control image data sample set generation module and a control component data sample set selection module, wherein the uncontrolled component data sample set acquisition module is used for acquiring a control image sample set corresponding to product component data and screening out an uncontrolled component data sample set based on preset conditions;
the model training module is used for inputting the out-of-control component data sample set into an initial vector machine model based on self-adaptive enhancement, taking an abnormal offset mode corresponding to the out-of-control component data sample set as output, performing iterative training, and determining a vector machine model based on self-adaptive enhancement with complete training;
and the abnormal offset mode determining module is used for acquiring real-time out-of-control component data, inputting the real-time out-of-control component data into the vector machine model which is completely trained and based on self-adaptive enhancement, and determining the abnormal offset mode of the product to be detected.
9. An electronic device comprising a processor and a memory, the memory having stored thereon a computer program that, when executed by the processor, implements a product composition data monitoring method as claimed in any one of claims 1-7.
10. A storage medium storing computer program instructions which, when executed by a computer, cause the computer to perform the product composition data monitoring method according to any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210857179.7A CN115270861A (en) | 2022-07-20 | 2022-07-20 | Product composition data monitoring method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210857179.7A CN115270861A (en) | 2022-07-20 | 2022-07-20 | Product composition data monitoring method and device, electronic equipment and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115270861A true CN115270861A (en) | 2022-11-01 |
Family
ID=83767298
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210857179.7A Pending CN115270861A (en) | 2022-07-20 | 2022-07-20 | Product composition data monitoring method and device, electronic equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115270861A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116595399A (en) * | 2023-06-14 | 2023-08-15 | 中国矿业大学(北京) | Analysis method for inconsistent element correlation problem in coal |
-
2022
- 2022-07-20 CN CN202210857179.7A patent/CN115270861A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116595399A (en) * | 2023-06-14 | 2023-08-15 | 中国矿业大学(北京) | Analysis method for inconsistent element correlation problem in coal |
CN116595399B (en) * | 2023-06-14 | 2024-01-05 | 中国矿业大学(北京) | Analysis method for inconsistent element correlation problem in coal |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110060144B (en) | Method for training credit model, method, device, equipment and medium for evaluating credit | |
CN111625516B (en) | Method, apparatus, computer device and storage medium for detecting data state | |
CN112149909A (en) | Ship oil consumption prediction method and device, computer equipment and storage medium | |
Yun | Prediction model of algal blooms using logistic regression and confusion matrix | |
CN112527604A (en) | Deep learning-based operation and maintenance detection method and system, electronic equipment and medium | |
Yu et al. | Control chart recognition based on the parallel model of CNN and LSTM with GA optimization | |
CN115587543A (en) | Federal learning and LSTM-based tool residual life prediction method and system | |
CN115049019B (en) | Method and device for evaluating arsenic adsorption performance of metal organic framework and related equipment | |
CN115394358A (en) | Single cell sequencing gene expression data interpolation method and system based on deep learning | |
CN115270861A (en) | Product composition data monitoring method and device, electronic equipment and storage medium | |
US20210166362A1 (en) | Wafer map identification method and computer-readable recording medium | |
CN110956331A (en) | Method, system and device for predicting operation state of digital factory | |
CN114169460A (en) | Sample screening method, sample screening device, computer equipment and storage medium | |
CN113110961A (en) | Equipment abnormality detection method and device, computer equipment and readable storage medium | |
CN117369418A (en) | Fault diagnosis method, system, storage medium and equipment for feeding system of numerical control machine tool | |
CN112506907A (en) | Engineering machinery marketing strategy pushing method, system and device based on big data | |
EP4287083A1 (en) | Determination program, determination apparatus, and method of determining | |
CN115861305A (en) | Flexible circuit board detection method and device, computer equipment and storage medium | |
CN115169832A (en) | Sensitivity analysis method and system based on curve form change | |
CN114418207A (en) | Time series trend prediction method and device | |
CN113836826A (en) | Key parameter determination method and device, electronic device and storage medium | |
CN114637620A (en) | Database system abnormity classification prediction method based on SVM algorithm | |
WO2022183019A1 (en) | Methods for mitigation of algorithmic bias discrimination, proxy discrimination and disparate impact | |
CN113780334A (en) | High-dimensional data classification method based on two-stage mixed feature selection | |
CN113255927A (en) | Logistic regression model training method and device, computer equipment and storage medium |
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 |