CN116449788A - Fault detection and productivity optimization method for casting blank production process of steel mill - Google Patents
Fault detection and productivity optimization method for casting blank production process of steel mill Download PDFInfo
- Publication number
- CN116449788A CN116449788A CN202310618453.XA CN202310618453A CN116449788A CN 116449788 A CN116449788 A CN 116449788A CN 202310618453 A CN202310618453 A CN 202310618453A CN 116449788 A CN116449788 A CN 116449788A
- Authority
- CN
- China
- Prior art keywords
- input sample
- steel mill
- sample data
- model
- 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
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 95
- 238000005266 casting Methods 0.000 title claims abstract description 94
- 229910000831 Steel Inorganic materials 0.000 title claims abstract description 81
- 239000010959 steel Substances 0.000 title claims abstract description 81
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000005457 optimization Methods 0.000 title claims abstract description 37
- 238000001514 detection method Methods 0.000 title claims abstract description 35
- 238000003070 Statistical process control Methods 0.000 claims abstract description 34
- 238000012545 processing Methods 0.000 claims abstract description 13
- 230000002159 abnormal effect Effects 0.000 claims abstract description 7
- 230000007547 defect Effects 0.000 claims abstract description 6
- 238000012216 screening Methods 0.000 claims abstract description 4
- 239000011159 matrix material Substances 0.000 claims description 27
- 238000012417 linear regression Methods 0.000 claims description 14
- 239000013598 vector Substances 0.000 claims description 11
- 230000001419 dependent effect Effects 0.000 claims description 9
- 238000009826 distribution Methods 0.000 claims description 8
- 230000001360 synchronised effect Effects 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 238000010586 diagram Methods 0.000 claims description 5
- 238000012544 monitoring process Methods 0.000 claims description 5
- 238000004458 analytical method Methods 0.000 claims description 3
- 238000007689 inspection Methods 0.000 claims description 3
- 238000012886 linear function Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 description 6
- 239000002893 slag Substances 0.000 description 6
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 229910052742 iron Inorganic materials 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000009628 steelmaking Methods 0.000 description 1
- 239000000758 substrate Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS], computer integrated manufacturing [CIM]
- G05B19/41875—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS], computer integrated manufacturing [CIM] characterised by quality surveillance of production
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/32—Operator till task planning
- G05B2219/32252—Scheduling production, machining, job shop
-
- 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
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Abstract
The invention relates to a method for fault detection and capacity optimization in a steel mill casting blank production process, which relates to the field of multivariate statistical process control, and comprises four steps of data processing, model establishment, parameter regression, fault detection and capacity optimization, wherein the data processing step is used for screening input sample data, eliminating abnormal data and fault data in the input sample data, the model establishment step is used for modeling the whole casting blank production process, the production flow of a casting blank is expressed in a mathematical mode, the parameter regression step is used for substituting the input sample data into the model established in the step S2, the feasibility of the model is verified, and the fault detection and capacity optimization step is used for judging whether production defects exist in production indexes corresponding to the input sample data and adjusting the value of a control variable in the input sample so as to maximize the steel mill casting blank capacity.
Description
Technical Field
The invention relates to the field of multivariate statistical process control, in particular to a method for fault detection and productivity optimization in the production process of casting blanks in a steel mill.
Background
At present, the casting blank production process of iron and steel enterprises in China is finished manually, the stability and uniformity of a casting blank finished product are difficult to ensure in links such as casting blank production, index adjustment and crystallizer slag adding, dead angles and unstable speed exist in slag adding, the quality of produced steel cannot be effectively controlled, the working environment is severe, high temperature and a large amount of dust exist, and certain harm is caused to the health of workers in long-term working. This is unacceptable to most young people in China at present, and the period of a qualified worker to cultivate is long, which results in the increase of the steelmaking cost.
Some steelworks have developed automatic slag adding equipment for casting production, replacing manual means. Few enterprises adopt the machine to add slag, but work in an open-loop mode, and the open-loop machine adds slag and finally returns to the original state, and the slag adding parameters of the robot are set according to the experience of workers, so that the method has higher subjectivity.
The technical method for optimizing the casting blank of the steel mill in the field of multivariate statistical process control in China is still blank, a reasonable steel mill casting blank production process guiding technology is urgently needed in the existing production background, the defect that the quality stability of steel products is difficult to ensure by manual operation is overcome, the defects that the fault detection is insensitive, the error rate is high and the like in the production process of the existing casting blank production guiding technology are overcome, the quality of the casting blank of the steel mill is further improved, and the productivity of the steel mill is maximized. Thus, the first and second substrates are bonded together, A new method for detecting faults and optimizing productivity in the casting blank production process of a steel mill has important significance for accelerating the intelligent process of the steel mill and promoting the upgrading of the steel production industry.
Disclosure of Invention
Therefore, the invention provides a fault detection and productivity optimization method for a casting blank production process of a steel mill, which is used for solving the problems of unstable casting blank quality, low productivity, untimely detection of casting blank production deviation from a stable state, high false alarm rate and the like in the existing casting blank production technology. The quality of the casting blank of the steel mill is further prompted, and the productivity of the steel mill is maximized.
In order to achieve the above object, the present invention provides a method for fault detection and capacity optimization in a steel mill casting production process, which is characterized by comprising:
s1, data processing, which includes screening input sample data and eliminating abnormal data and fault data in the input sample data;
s2, establishing a model, wherein the model comprises modeling of the whole casting blank production process based on production data;
s3, parameter regression, including substituting input sample data into the model established in the step S2, and verifying feasibility of the model;
and S4, fault detection and productivity optimization, wherein the step comprises the steps of judging whether production defects exist in production indexes corresponding to the input sample data and adjusting the value of a control variable in the input sample so as to maximize the productivity of casting blanks of the steel mill.
Further, in the step S1, the data processing includes: and removing the scattered abnormal data and the fault data in the input sample data set by using a clustering method, and solving the value of k in the k-means cluster by adopting a secondary clustering idea.
Further, the secondary clustering includes: for each interval of input sample data, clustering results obtained by adopting a direct clustering method are { C1, C2, …, ck }, and the clustering center of each cluster is { W1, W2, …, wk }; finding out a cluster center Wmax with the largest median value of { W1, W2, …, wk }, wherein the cluster corresponding to the largest cluster center is Cmax; calculating a distance d i between Wmax and each cluster center except Wmax in { W1, W2, …, wk }, and setting d i =wmax-Wi; comparing each d i with a human preset threshold T, if d i < T, combining Ci and Cmax into a cluster, otherwise, excluding Ci; and finally, the number of the reserved clusters, namely the value of k in the k-means algorithm, wherein the reserved clusters are expected normal data in the input sample data.
Further, in the step S2, the modeling includes modeling a multivariate statistical process control model on input sample data under normal production conditions obtained through data processing.
Further, the multivariate statistical process control model establishment comprises: expanding a three-dimensional data matrix formed by input sample data of a steel mill casting blank according to a variable expansion mode to form a two-dimensional matrix, performing partial least square analysis on the two-dimensional matrix, mapping the input sample data in a high-dimensional space into a low-dimensional space to obtain feature vectors of independent variables and dependent variables in the input sample data, and establishing a multiple linear regression relation between the independent variables and the dependent variable feature vectors.
Further, in the step S3, the parameter regression includes: and performing inverse standardization on the multiple linear regression model through the variance and the mean value of the two-dimensional matrix of the original data to obtain a multiple linear regression model between the independent variable and the dependent variable of the original data, substituting the input sample data into the multiple linear regression model to obtain regression values corresponding to the independent variable of the single input sample data.
Further, in the step S4, the fault detection and capacity optimization includes: calculating statistics of each time point for input sample data, drawing a multivariable statistical process control statistical chart of casting blank production of a steel mill through the statistics, calibrating confidence, and monitoring the degree of deviation of the casting blank production process from a partial least square model in real time; the capacity optimization is used for adjusting control parameter variables in samples input into the multivariate statistical process control model, giving weight to each control variable according to the requirements of a producer, giving a large weight to the control variable which is expected to exert a large influence, otherwise, inputting the control parameter with the weight into the multivariate statistical process control model, carrying out feasible domain optimization according to the obtained multivariate linear function model and the actual production condition of the steel mill, and calculating the control variable value corresponding to the maximum output under the current weight so as to maximize the casting blank capacity of the steel mill; meanwhile, the productivity of the steel mill casting blank can be predicted according to the result of the parameter regression of the multivariate statistical process control model and by combining the investment in the casting blank production line of the steel mill in the current year and the production level of the synchronous casting blank sales market.
Further, plotting the multivariate statistical process control statistical graph comprises: drawing Hotelling T from principal component score vector and eigenvalue of covariance matrix of input sample data 2 Statistical diagram, and determining Hotelling T by the number of samples of principal element model, principal element number and corresponding critical value of F under-distribution test level alpha 2 Control limits for statistical graphsDrawing an SPE statistical graph by an error matrix obtained by a parameter regression value of input sample data, and determining a control limit Q of the SPE statistical graph by a characteristic value of a covariance matrix and a critical value of a test level alpha under normal distribution α According to Hotelling T 2 And performing fault detection on the statistical graphs and the SPE statistical graphs.
Compared with the prior art, the method has the beneficial effects that the production process of the casting blank of the steel mill can be monitored, whether the casting blank is in a stable state or not is analyzed, the steel production process data is used for supporting decision making, and the main index factors influencing the steel quality are found out, so that the overall quality of the steel is improved, the reworking on the production line of the defective steel is reduced, the productivity is improved, and the process capability is matched with the steel performance requirement.
The parameter regression step is used for inputting the series index data of the casting blank production process into the production process model subjected to the inverse standardization treatment, calculating the variable value corresponding to the independent variable of the input sample data, verifying the feasibility and the accuracy of the model, and improving the accurate control of the casting blank production process.
According to the invention, the fault detection and productivity optimization step is used for calculating statistics of each time point of input sample data, drawing a multivariate statistical process control statistical chart of the production of the steel mill casting blank corresponding to the statistics, and monitoring the degree of deviation of the production process of the steel mill casting blank from the multivariate statistical process control model in real time, thereby improving the accurate control of the production process of the steel mill casting blank.
Furthermore, the effectiveness of data sources of model construction in the casting blank production process of the steel mill is guaranteed through a clustering method and a secondary clustering idea.
Furthermore, the invention maps the input sample data of the high-dimensional space into the low-dimensional space by establishing the multivariate statistical process control model, thereby reducing the complexity of model establishment.
Furthermore, the invention improves the accurate control of the input sample data through the anti-standardized data processing and the establishment of the multiple linear regression model.
Further, the invention draws Hotelling T through eigenvalues of principal component score vector and covariance matrix 2 Statistical plot, increase the contrast to Hotelling T 2 The accurate control of the drawing process of the statistical graph draws the SPE statistical graph through an error matrix obtained by inputting the parameter regression value of the sample data, improves the accurate control of the SPE statistical graph, and passes through the Hotelling T 2 The statistical graphs and SPE statistical graphs improve the accurate control of the fault detection process.
Furthermore, the invention adjusts the control parameter variable in the sample input to the multivariate statistical process control model, and performs feasible region optimization according to the actual production condition of the steel mill, so that the casting blank productivity of the steel mill reaches the maximum; meanwhile, the production capacity of the steel mill casting blank can be predicted according to the input of the multi-variable statistical process control model parameter regression on the annual steel mill casting blank production line and the production level of the synchronous casting blank sales market, and the production efficiency of the steel mill casting blank is improved.
Drawings
FIG. 1 is a flow chart of a method for fault detection and capacity optimization in the production process of a steel mill casting blank according to the present invention;
FIG. 2 is a flowchart of a detailed scheme of the fault detection and capacity optimization method for the casting blank production process of the steel mill.
Detailed Description
In order that the objects and advantages of the invention will become more apparent, the invention will be further described with reference to the following examples; it should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are merely for explaining the technical principles of the present invention, and are not intended to limit the scope of the present invention.
Referring to fig. 1-2, fig. 1 is a flowchart of a method for fault detection and capacity optimization in a steel mill casting production process according to the present invention, and fig. 2 is a flowchart of a detailed scheme of a method for capacity optimization in a steel mill casting according to the present invention.
The invention relates to a fault detection and productivity optimization method for a casting blank production process of a steel mill, which comprises the following steps:
s1, data processing, which includes screening input sample data and eliminating abnormal data and fault data in the input sample data;
s2, establishing a model, wherein the model comprises modeling of the whole casting blank production process based on production data;
s3, parameter regression, including substituting input sample data into the model established in the step S2, and verifying feasibility of the model;
and S4, fault detection and productivity optimization, wherein the step comprises the steps of judging whether production defects exist in production indexes corresponding to the input sample data and adjusting the value of a control variable in the input sample so as to maximize the productivity of casting blanks of the steel mill.
In the embodiment of the invention, the model is built to mathematically model the whole casting blank production process of the steel mill, and the complete production process of the casting blank is expressed in a mathematical mode, so that the production process of the casting blank is changed into a production process expression form which can be processed by a computer, and the production process of the casting blank is optimized by the computer.
The parameter regression is used for inputting the series index data of the casting blank production process into the production process model subjected to the inverse standardization treatment, calculating the variable value corresponding to the independent variable of the input sample data, verifying the feasibility and the accuracy of the model, and improving the accurate control of the casting blank production process.
According to the invention, fault detection and productivity optimization are carried out by calculating statistics of each time point of input sample data, drawing a multivariate statistical process control statistical chart of casting blank production of a steel mill, calibrating confidence, and monitoring the deviation degree of the casting blank production process from a partial least square model in real time; adjusting control parameter variables in a sample input into a multivariate statistical process control model, and performing feasible region optimization according to actual production conditions of a steel mill so as to maximize casting blank productivity of the steel mill; meanwhile, the investment in the annual steel mill casting blank production line and the production level of the synchronous casting blank sales market are considered according to the result of the parameter regression of the multivariate statistical process control model to predict the productivity of the steel mill casting blank, so that the production efficiency of the steel mill casting blank is improved.
Specifically, in step S1, the data processing includes: and removing the scattered abnormal data and the fault data in the input sample data set by using a clustering method, and solving the value of k in the k-means cluster by adopting a secondary clustering idea.
The invention ensures the effectiveness of data sources of model construction in the casting blank production process of the steel mill through a clustering method and a secondary clustering idea.
Specifically, the secondary clustering includes: for each input sample data interval, clustering results obtained by adopting a direct clustering method are { C1, C2, …, ck }, and the clustering center of each cluster is { W1, W2, …, wk }; finding out a cluster center Wmax with the largest median value of { W1, W2, …, wk }, wherein the cluster corresponding to the largest cluster center is Cmax; calculating a distance d i between Wmax and each cluster center except Wmax in { W1, W2, …, wk }, and setting d i =wmax-Wi; comparing each d i with a human preset threshold T, if d i < T, combining Ci and Cmax into a cluster, otherwise, excluding Ci; and finally, the number of the reserved clusters, namely the value of k in the k-means algorithm, wherein the reserved clusters are expected normal data in the input sample data.
Specifically, in step S2, modeling includes modeling a multivariate statistical process control model on input sample data under normal production conditions as a result of data processing.
According to the invention, the multivariate statistical process control model is built to map the input sample data in the high-dimensional space into the low-dimensional space, so that the complexity of model building is reduced.
Specifically, how muchThe variable statistical process control model establishment comprises the following steps: the three-dimensional data matrix A (I X J X K) formed by the input sample data of the casting blank of the steel mill is unfolded according to a variable unfolding mode to form a two-dimensional matrix X A (J.I.times.K) for a two-dimensional matrix X A And performing partial least square analysis, mapping input sample data in a high-dimensional space into a low-dimensional space to obtain feature vectors of independent variables and dependent variables in the input sample data, and establishing a multiple linear regression relationship between the independent variables and the dependent variable feature vectors.
Specifically, in step S3, the parameter regression includes: by a two-dimensional matrix X of raw data A And (3) performing inverse standardization on the multiple linear regression model to obtain a multiple linear regression model between the independent variable and the dependent variable of the original data, substituting the input sample data into the multiple linear regression model to obtain a regression value corresponding to the independent variable of the single input sample data.
According to the invention, through inverse standardized data processing and multiple linear regression model establishment, the accurate control of input sample data is improved.
Specifically, in step S4, the fault detection and capacity optimization includes: calculating statistics of each time point for input sample data, drawing a multivariable statistical process control statistical chart of casting blank production of a steel mill through the statistics, calibrating confidence, and monitoring the degree of deviation of the casting blank production process from a partial least square model in real time; the capacity optimization is used for giving a large weight to a control variable which is expected to exert a large influence according to the requirements of a producer, if the control variable is small, inputting the control parameter with the weight into the multivariate statistical process control model to obtain a multivariate linear function model of output and control variable, carrying out feasible domain optimization according to the actual production condition of a steel mill, and calculating a control variable value corresponding to the maximum output under the current weight, namely, maximizing the capacity of the steel mill; meanwhile, the productivity of the steel mill casting blank can be predicted according to the result of the parameter regression of the multivariate statistical process control model and by combining the investment in the casting blank production line of the steel mill in the current year and the production level of the synchronous casting blank sales market.
According to the method, the degree of deviation of the casting blank production process from the partial least square model is monitored in real time, and the control parameter variable in the sample input into the multivariate statistical process control model is adjusted, so that the feasible region optimization is performed according to the actual production condition of the steel mill, and the casting blank productivity of the steel mill is maximized; meanwhile, the production capacity of the steel mill casting blank can be predicted according to the input of the multi-variable statistical process control model parameter regression on the annual steel mill casting blank production line and the production level of the synchronous casting blank sales market, and the production efficiency of the steel mill casting blank is improved.
Specifically, plotting the multivariate statistical process control statistical graph includes: drawing Hotelling T from principal component score vector and eigenvalue of covariance matrix of input sample data 2 Statistical diagram, and determining Hotelling T by the number of samples of principal component model, principal component number and corresponding critical value with test level alpha under F distribution 2 Control limits for statistical graphsDrawing an SPE statistical graph by an error matrix obtained by a parameter regression value of input sample data, and determining a control limit Q of the SPE statistical graph by a characteristic value of a covariance matrix and a critical value of a test level alpha under normal distribution α According to Hotelling T 2 And performing fault detection on the statistical graphs and the SPE statistical graphs, and predicting the productivity according to the production investment and the market production level.
Calculation of Hotelling T in the examples of the invention 2 The specific steps of the control limit of the statistical chart are as follows:
Hotelling T 2 the control limit of the statistical diagram is thatSetting up
Wherein m is the number of samples of the established principal component model, k is the number of principal components in the principal component model, alpha is the inspection level, F k,m-1,α The F distribution critical value is given under the conditions that the corresponding test level is alpha, the degree of freedom is k and m-1.
The specific steps for calculating the control limit of the SPE statistical chart in the embodiment of the invention are as follows:
the control limit of the SPE statistical diagram is Q α Setting up
Wherein (i=1, 2,3 …), λ j For covariance matrix eigenvalue of input data matrix X, ca is critical value of normal distribution when inspection level is alpha, k is number of principal elements in principal element model, n is number of variables in input data.
The invention draws Hotelling T through principal component score vector and eigenvalue of covariance matrix 2 Statistical plot, increase the contrast to Hotelling T 2 The accurate control of the drawing process of the statistical graph draws the SPE statistical graph through an error matrix obtained by inputting the parameter regression value of the sample data, improves the accurate control of the SPE statistical graph, and passes through the Hotelling T 2 The statistical graphs and SPE statistical graphs improve the accurate control of the fault detection process.
Thus far, the technical solution of the present invention has been described in connection with the preferred embodiments shown in the drawings, but it is easily understood by those skilled in the art that the scope of protection of the present invention is not limited to these specific embodiments. Equivalent modifications and substitutions for related technical features may be made by those skilled in the art without departing from the principles of the present invention, and such modifications and substitutions will be within the scope of the present invention.
The foregoing description is only of the preferred embodiments of the invention and is not intended to limit the invention; various modifications and variations of the present invention will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (8)
1. The fault detection and productivity optimization method for the casting blank production process of the steel mill is characterized by comprising the following steps of:
s1, data processing, which includes screening input sample data and eliminating abnormal data and fault data in the input sample data;
s2, establishing a model, wherein the model comprises modeling of the whole casting blank production process based on production data;
s3, parameter regression, including substituting input sample data into the model established in the step S2, and verifying feasibility of the model;
and S4, fault detection and productivity optimization, wherein the step comprises the steps of judging whether production defects exist in production indexes corresponding to the input sample data and adjusting the value of a control variable in the input sample so as to maximize the productivity of casting blanks of the steel mill.
2. The method for fault detection and capacity optimization of a steel mill casting process according to claim 1, wherein in step S1, the data processing includes: and removing the scattered abnormal data and the fault data in the input sample data set by using a clustering method, and solving the value of k in the k-means cluster by adopting a secondary clustering idea.
3. The method for fault detection and capacity optimization of a steel mill casting production process according to claim 2, wherein the secondary clustering comprises: for each interval of input sample data, clustering results obtained by adopting a direct clustering method are { C1, C2, …, ck }, and the clustering center of each cluster is { W1, W2, …, wk }; finding out a cluster center Wmax with the largest median value of { W1, W2, …, wk }, wherein the cluster corresponding to the largest cluster center is Cmax; calculating the distance di of each cluster center except Wmax in Wmax and { W1, W2, …, wk }, and setting di=Wmax-Wi; comparing each di with a human preset threshold T, if di is smaller than T, combining Ci and Cmax into a cluster, otherwise, excluding Ci; and finally, the number of the reserved clusters, namely the value of k in the k-means algorithm, wherein the reserved clusters are expected normal data in the input sample data.
4. A method of fault detection and capacity optimisation of a steel mill casting process according to claim 3, wherein in step S2 the modeling comprises modeling a multivariate statistical process control model on input sample data from normal production conditions obtained by data processing.
5. The method of fault detection and capacity optimization for a steel mill casting process according to claim 4, wherein the multivariate statistical process control model establishment comprises: expanding a three-dimensional data matrix formed by input sample data of a steel mill casting blank according to a variable expansion mode to form a two-dimensional matrix, performing partial least square analysis on the two-dimensional matrix, mapping the input sample data in a high-dimensional space into a low-dimensional space to obtain feature vectors of independent variables and dependent variables in the input sample data, and establishing a multiple linear regression relation between the independent variables and the dependent variable feature vectors.
6. The method for fault detection and capacity optimization of a steel mill casting process according to claim 5, wherein in the step S3, the parameter regression includes: and performing inverse standardization on the multiple linear regression model through the variance and the mean value of the two-dimensional matrix of the original data to obtain a multiple linear regression model between the independent variable and the dependent variable of the original data, substituting the input sample data into the multiple linear regression model to obtain regression values corresponding to the independent variable of the single input sample data.
7. The method for fault detection and capacity optimization of a steel mill casting process according to claim 6, wherein in the step S4, the fault detection and capacity optimization comprises: calculating statistics of each time point for input sample data, drawing a multivariable statistical process control statistical chart of casting blank production of a steel mill through the statistics, calibrating confidence, and monitoring the degree of deviation of the casting blank production process from a partial least square model in real time; the capacity optimization is used for adjusting control parameter variables in samples input into the multivariate statistical process control model, giving weight to each control variable according to the requirements of a producer, giving a large weight to the control variable which is expected to exert a large influence, otherwise, inputting the control parameter with the weight into the multivariate statistical process control model, carrying out feasible domain optimization according to the obtained multivariate linear function model and the actual production condition of the steel mill, and calculating the control variable value corresponding to the maximum output under the current weight so as to maximize the casting blank capacity of the steel mill; meanwhile, the productivity of the steel mill casting blank can be predicted according to the result of the parameter regression of the multivariate statistical process control model and by combining the investment in the casting blank production line of the steel mill in the current year and the production level of the synchronous casting blank sales market.
8. The method of fault detection and capacity optimization of a steel mill casting process of claim 7 wherein plotting a multivariate statistical process control statistical graph comprises: drawing Hotelling T from principal component score vector and eigenvalue of covariance matrix of input sample data 2 Statistical diagram, and determining Hotelling T by the number of samples of principal element model, principal element number and corresponding critical value of F under-distribution test level alpha 2 Control limits for statistical graphsFrom inputDrawing an SPE statistical graph by an error matrix obtained by the parameter regression value of the sample data, and determining a control limit Q of the SPE statistical graph by the characteristic value of the covariance matrix and the critical value of the inspection level alpha under normal distribution α According to Hotelling T 2 And performing fault detection on the statistical graphs and the SPE statistical graphs.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310618453.XA CN116449788A (en) | 2023-05-30 | 2023-05-30 | Fault detection and productivity optimization method for casting blank production process of steel mill |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310618453.XA CN116449788A (en) | 2023-05-30 | 2023-05-30 | Fault detection and productivity optimization method for casting blank production process of steel mill |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116449788A true CN116449788A (en) | 2023-07-18 |
Family
ID=87127554
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310618453.XA Pending CN116449788A (en) | 2023-05-30 | 2023-05-30 | Fault detection and productivity optimization method for casting blank production process of steel mill |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116449788A (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6564119B1 (en) * | 1998-07-21 | 2003-05-13 | Dofasco Inc. | Multivariate statistical model-based system for monitoring the operation of a continuous caster and detecting the onset of impending breakouts |
US20100191361A1 (en) * | 2009-01-23 | 2010-07-29 | Mks Instruments, Inc. | Controlling a Manufacturing Process with a Multivariate Model |
CN105929812A (en) * | 2016-04-18 | 2016-09-07 | 北京科技大学 | Strip steel hot continuous rolling quality fault diagnosis method and device |
CN114329940A (en) * | 2018-02-05 | 2022-04-12 | 北京科技大学 | Continuous casting billet quality prediction method based on extreme learning machine |
CN115860211A (en) * | 2022-11-29 | 2023-03-28 | 北京科技大学 | Casting blank quality prediction method based on local online modeling |
CN115994337A (en) * | 2023-03-24 | 2023-04-21 | 北京科技大学 | Method and device for detecting minor faults in non-stationary process of hot continuous rolling of strip steel |
-
2023
- 2023-05-30 CN CN202310618453.XA patent/CN116449788A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6564119B1 (en) * | 1998-07-21 | 2003-05-13 | Dofasco Inc. | Multivariate statistical model-based system for monitoring the operation of a continuous caster and detecting the onset of impending breakouts |
US20100191361A1 (en) * | 2009-01-23 | 2010-07-29 | Mks Instruments, Inc. | Controlling a Manufacturing Process with a Multivariate Model |
CN105929812A (en) * | 2016-04-18 | 2016-09-07 | 北京科技大学 | Strip steel hot continuous rolling quality fault diagnosis method and device |
CN114329940A (en) * | 2018-02-05 | 2022-04-12 | 北京科技大学 | Continuous casting billet quality prediction method based on extreme learning machine |
CN115860211A (en) * | 2022-11-29 | 2023-03-28 | 北京科技大学 | Casting blank quality prediction method based on local online modeling |
CN115994337A (en) * | 2023-03-24 | 2023-04-21 | 北京科技大学 | Method and device for detecting minor faults in non-stationary process of hot continuous rolling of strip steel |
Non-Patent Citations (1)
Title |
---|
杨尚玉;王旭东;姚曼;李全育;: "基于多变量统计对连铸结晶器过程的监测", 钢铁研究学报, no. 10, pages 1 - 4 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109459993B (en) | Online adaptive fault monitoring and diagnosing method for process industrial process | |
US11079726B2 (en) | Predictive monitoring and diagnostics systems and methods | |
TWI625615B (en) | Prediction model building method and associated predicting method and computer software product | |
US9378183B2 (en) | Monitoring diagnostic device and monitoring diagnostic method | |
KR20040088491A (en) | Manufacturing design and process analysis system | |
CN105159239B (en) | A kind of SPC strategies automatic update method and SPC Strategy Auto system | |
EP3373211A1 (en) | Management device and non-transitory computer-readable medium | |
KR102626984B1 (en) | Managing apparatus and managing system | |
TWI584134B (en) | Method for analyzing variation causes of manufacturing process and system for analyzing variation causes of manufacturing process | |
CN111105082A (en) | Workpiece quality prediction model construction method and prediction method based on machine learning | |
CN110096036A (en) | A kind of determination method, device and equipment of equipment state | |
CN101118422A (en) | Virtual measurement prediction generated by semi-conductor, method for establishing prediction model and system | |
CN116449788A (en) | Fault detection and productivity optimization method for casting blank production process of steel mill | |
Perkasa | Benefit using statistical process control (SPC) for process control in textile manufacturing: A Review | |
JP2011054804A (en) | Method and system for management of semiconductor manufacturing device | |
CN101118423A (en) | Serviceability selecting method and system of virtual measuring prediction model | |
Jain et al. | Quality control based tool condition monitoring | |
Wenzel et al. | Improving the accuracy of cycle time estimation for simulation in volatile manufacturing execution environments | |
CN112990682A (en) | Chemical device control parameter optimization method and system based on big data | |
US20220309431A1 (en) | Analysis apparatus, analysis method, and computer-readable medium | |
Heo et al. | A modified lasso model for yield analysis considering the interaction effect in a multistage manufacturing line | |
JP7218338B2 (en) | Data processing system, data processing method, and data processing program | |
JP2019184575A (en) | Measurement operation parameter adjustment device, machine learning device, and system | |
Nishimura et al. | Anomaly detection and analysis by a gradient boosting trees and neural network ensemble model | |
Wang et al. | On-line modeling and monitoring for multi-operation batch processes with infinite data types |
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 |