CN110297852B - Ship coating defect knowledge acquisition method based on PCA-rough set - Google Patents
Ship coating defect knowledge acquisition method based on PCA-rough set Download PDFInfo
- Publication number
- CN110297852B CN110297852B CN201910571065.4A CN201910571065A CN110297852B CN 110297852 B CN110297852 B CN 110297852B CN 201910571065 A CN201910571065 A CN 201910571065A CN 110297852 B CN110297852 B CN 110297852B
- Authority
- CN
- China
- Prior art keywords
- knowledge
- attribute
- data
- ship
- coating
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000576 coating method Methods 0.000 title claims abstract description 80
- 239000011248 coating agent Substances 0.000 title claims abstract description 68
- 230000007547 defect Effects 0.000 title claims abstract description 63
- 238000000034 method Methods 0.000 title claims abstract description 38
- 230000009467 reduction Effects 0.000 claims abstract description 36
- 230000008569 process Effects 0.000 claims abstract description 15
- 238000000513 principal component analysis Methods 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 238000012545 processing Methods 0.000 claims description 23
- 239000003973 paint Substances 0.000 claims description 14
- 238000010422 painting Methods 0.000 claims description 13
- 238000007689 inspection Methods 0.000 claims description 12
- 230000007613 environmental effect Effects 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 10
- JEIPFZHSYJVQDO-UHFFFAOYSA-N iron(III) oxide Inorganic materials O=[Fe]O[Fe]=O JEIPFZHSYJVQDO-UHFFFAOYSA-N 0.000 claims description 7
- 230000003746 surface roughness Effects 0.000 claims description 7
- 238000005507 spraying Methods 0.000 claims description 4
- 238000009825 accumulation Methods 0.000 claims description 3
- 230000015572 biosynthetic process Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 239000007787 solid Substances 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 239000013598 vector Substances 0.000 claims description 3
- 230000002265 prevention Effects 0.000 abstract description 2
- 230000002452 interceptive effect Effects 0.000 description 3
- 238000010276 construction Methods 0.000 description 2
- 230000037303 wrinkles Effects 0.000 description 2
- 230000002411 adverse Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000002547 anomalous effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000005336 cracking Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000003085 diluting agent Substances 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000001035 drying Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000006260 foam Substances 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000012847 principal component analysis method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000007665 sagging Methods 0.000 description 1
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B05—SPRAYING OR ATOMISING IN GENERAL; APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
- B05B—SPRAYING APPARATUS; ATOMISING APPARATUS; NOZZLES
- B05B12/00—Arrangements for controlling delivery; Arrangements for controlling the spray area
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B05—SPRAYING OR ATOMISING IN GENERAL; APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
- B05B—SPRAYING APPARATUS; ATOMISING APPARATUS; NOZZLES
- B05B12/00—Arrangements for controlling delivery; Arrangements for controlling the spray area
- B05B12/08—Arrangements for controlling delivery; Arrangements for controlling the spray area responsive to condition of liquid or other fluent material to be discharged, of ambient medium or of target ; responsive to condition of spray devices or of supply means, e.g. pipes, pumps or their drive means
- B05B12/084—Arrangements for controlling delivery; Arrangements for controlling the spray area responsive to condition of liquid or other fluent material to be discharged, of ambient medium or of target ; responsive to condition of spray devices or of supply means, e.g. pipes, pumps or their drive means responsive to condition of liquid or other fluent material already sprayed on the target, e.g. coating thickness, weight or pattern
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B05—SPRAYING OR ATOMISING IN GENERAL; APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
- B05B—SPRAYING APPARATUS; ATOMISING APPARATUS; NOZZLES
- B05B12/00—Arrangements for controlling delivery; Arrangements for controlling the spray area
- B05B12/08—Arrangements for controlling delivery; Arrangements for controlling the spray area responsive to condition of liquid or other fluent material to be discharged, of ambient medium or of target ; responsive to condition of spray devices or of supply means, e.g. pipes, pumps or their drive means
- B05B12/12—Arrangements for controlling delivery; Arrangements for controlling the spray area responsive to condition of liquid or other fluent material to be discharged, of ambient medium or of target ; responsive to condition of spray devices or of supply means, e.g. pipes, pumps or their drive means responsive to conditions of ambient medium or target, e.g. humidity, temperature position or movement of the target relative to the spray apparatus
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B05—SPRAYING OR ATOMISING IN GENERAL; APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
- B05B—SPRAYING APPARATUS; ATOMISING APPARATUS; NOZZLES
- B05B12/00—Arrangements for controlling delivery; Arrangements for controlling the spray area
- B05B12/08—Arrangements for controlling delivery; Arrangements for controlling the spray area responsive to condition of liquid or other fluent material to be discharged, of ambient medium or of target ; responsive to condition of spray devices or of supply means, e.g. pipes, pumps or their drive means
- B05B12/12—Arrangements for controlling delivery; Arrangements for controlling the spray area responsive to condition of liquid or other fluent material to be discharged, of ambient medium or of target ; responsive to condition of spray devices or of supply means, e.g. pipes, pumps or their drive means responsive to conditions of ambient medium or target, e.g. humidity, temperature position or movement of the target relative to the spray apparatus
- B05B12/124—Arrangements for controlling delivery; Arrangements for controlling the spray area responsive to condition of liquid or other fluent material to be discharged, of ambient medium or of target ; responsive to condition of spray devices or of supply means, e.g. pipes, pumps or their drive means responsive to conditions of ambient medium or target, e.g. humidity, temperature position or movement of the target relative to the spray apparatus responsive to distance between spray apparatus and target
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B05—SPRAYING OR ATOMISING IN GENERAL; APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
- B05B—SPRAYING APPARATUS; ATOMISING APPARATUS; NOZZLES
- B05B12/00—Arrangements for controlling delivery; Arrangements for controlling the spray area
- B05B12/16—Arrangements for controlling delivery; Arrangements for controlling the spray area for controlling the spray area
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/22—Indexing; Data structures therefor; Storage structures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- 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
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Fuzzy Systems (AREA)
- Application Of Or Painting With Fluid Materials (AREA)
Abstract
The invention discloses a ship coating defect knowledge acquisition method based on PCA-rough set, which comprises the following steps: selecting a target data set from a ship coating process database and a ship coating defect case library; checking and preprocessing the selected target data set; performing dimension reduction treatment on the multi-source heterogeneous data of the ship coating defects by using principal component analysis; the rough set theory acquires knowledge of the ship coating defect data; the knowledge is classified and stored. The invention can process various data when knowledge is acquired, such as: continuous and discrete, without considering inconsistent data types; by adopting the PCA-rough set knowledge acquisition method, the ship coating defect cause knowledge can be acquired, the current situation that the ship coating defect can only be detected afterwards is changed, and the advanced prevention and control of the coating defect are realized.
Description
Technical Field
The invention relates to the field of ship coating defect knowledge acquisition, in particular to a ship coating defect knowledge acquisition method based on PCA-rough set.
Background
In the process of ship coating, various defects can be generated due to the influence of improper operation, environmental change during drying and curing, quality of the coating and the like. According to experience deduction and related organization investigation, 80% of coating defects are caused by improper operation of constructors in the construction process, and the reasons for the defects can be clearly understood to guide the coating operation process and reduce the coating defects caused by operation irregularity. The traditional coating process detects the quality of the coating in a specific time after construction is finished, judges and records the type and grade of the generated defects, and cannot prevent and control the generation of the defects in advance.
Knowledge acquisition is the ability to obtain new knowledge from existing process data and information. The current common knowledge acquisition modes can be divided into two types, namely interactive knowledge acquisition and automatic knowledge acquisition, wherein the interactive knowledge acquisition is that a knowledge engineer cooperates with a domain expert, and related domain knowledge and expert knowledge are collected, analyzed, mined, synthesized, arranged and generalized, and stored in a knowledge base after standardized expression of the knowledge. The subjectivity of the knowledge acquired by the interactive method is strong, the accuracy of the knowledge cannot be guaranteed, meanwhile, the acquired knowledge needs to be input into a computer one by one, the knowledge acquisition efficiency is low, and therefore automatic knowledge acquisition research needs to be conducted.
Coating process data obtained from a field product data management system (PDM) is from enterprise design BOM, process BOM and manufacturing BOM, which contain massive amounts of data, so that it is necessary to perform dimension reduction processing on a large amount of multi-source heterogeneous data, and data is reduced while as little lost data information as possible is ensured so as to lay a good foundation for later data analysis and knowledge acquisition. However, not all information is necessary, and many noise and interference terms are often added to the data, so that the data needs to be preprocessed before the dimension is reduced to prevent the data from adversely affecting the final data result.
In the real world, data are often complex and uncertain, and to analyze these uncertainty data in a rapid and collaborative manner and provide more accurate and effective knowledge for users in a hierarchical manner, a new intelligent analysis theory, model and method for complex data must be studied.
Disclosure of Invention
The invention aims to: in order to overcome the defects of the background technology, the invention discloses a ship coating defect knowledge acquisition method based on PCA-rough set, which can accurately and rapidly acquire the ship coating defect knowledge.
The technical scheme is as follows: the invention discloses a ship coating defect knowledge acquisition method based on PCA-rough set, which comprises the following steps:
(1) Selecting a target data set from a ship coating process database and a ship coating defect case library;
(2) Checking and preprocessing the selected target data set;
(3) Performing dimension reduction treatment on the multi-source heterogeneous data of the ship coating defects by using principal component analysis;
(4) The rough set theory acquires knowledge of the ship coating defect data;
(5) The knowledge is classified and stored.
The target data set in the step (1) comprises attribute information, parameter information and environment information, wherein the attribute information comprises a coating area, surface roughness, a rust removal grade, a coating method and coating equipment; the parameter information comprises paint viscosity, aerodynamic force, spraying distance and paint transfer rate; the environmental information includes air flow rate, relative humidity, air temperature.
Further, the data inspection in the step (2) includes data missing value inspection, inconsistent data inspection and data noise value inspection, and it is determined whether the data needs to be preprocessed.
The data preprocessing comprises deletion value processing, rationality checking and noise value processing, and the specific means comprise automatic deleting, filling and correcting processing, judging whether the processing is complete after the processing is finished, returning to recheck and data processing if the processing is incomplete, and otherwise, carrying out the next step.
Further, the step (3) specifically adopts a one-hot coding mode to code the relevant attribute of the painting defect of the discrete ship, and the relevant attribute of the painting defect of the discrete ship comprises painting equipment, paint types, painting methods and operation groups; solving a covariance matrix of the preprocessed data, calculating characteristic values and characteristic vectors of the covariance matrix, sorting according to the characteristic values, calculating a variance contribution rate and a variance accumulation contribution rate, interactively determining the number of principal components, obtaining principal component values, and inputting variables corresponding to the principal components as new data, so that dimension reduction processing of a large amount of multi-source heterogeneous data in the ship coating defect data is realized.
Further, the step (4) is specifically divided into the following steps:
(a) Establishing an initial case decision table, and specifically dividing condition attributes and decision attributes;
(b) Performing knowledge reduction, and calculating the importance of each reduced condition attribute to the decision attribute;
(c) Performing attribute value reduction to obtain a simplified decision data table;
(d) And extracting rule knowledge.
Wherein, in the step (a), a plurality of factors which can influence the final formation of the coating defect are comprehensively considered as condition attributes, wherein the condition attributes comprise process attributes, process parameters and environmental information; the technological properties comprise a coating area, surface roughness and a rust removal grade, the technological parameters comprise coating viscosity, dry film thickness, solid content and nozzle distance, and the environmental information comprises air temperature, relative humidity and air flow rate; the decision attribute serves as the final defect name.
And (b) carrying out attribute reduction by adopting a reduction algorithm based on attribute importance under the condition of not affecting knowledge expression capability, eliminating redundant knowledge, and calculating the importance of each reduced condition attribute to decision attribute.
The attribute importance reduction algorithm comprises the following steps:
I. obtaining a Core attribute set Core by adopting a difference matrix calculation;
II. Calculating the dependence gamma of the whole decision table before reduction c (D) And the importance σ of each condition attribute other than the kernel attribute CD (c i )；
III, calculating the dependency gamma of the core attribute Core (D)；
IV, if gamma Core (D)≥γ C (D) The algorithm ends and the output attribute is about Jian Jige Core; if gamma is Core (D)＜γ C (D)，
Ordering attributes from big to small according to attribute importance c i (i=1, 2, … n) added to the set Core one by one for (i=1:n), core=core ∈ i Calculating gamma Core (D) If gamma is Core (D)≥γ C (D) The algorithm ends and the output attribute is about Jian Jige Core; if gamma is Core (D)＜γ C (D) The cycle continues.
Further, in the step (5), the support degree, coverage degree and credibility of each ship coating defect knowledge are calculated, and the knowledge is divided into deterministic knowledge, strong rule knowledge and weak rule knowledge and stored in a knowledge base.
The beneficial effects are that: compared with the prior art, the invention has the advantages that:
(1) Firstly, the PCA-rough set knowledge acquisition method is adopted, so that the ship coating defect cause knowledge can be acquired, the current situation that the ship coating defect can only be detected afterwards is changed, and the advanced prevention and control of the coating defect are realized;
(2) Secondly, when the invention acquires knowledge, various types of data (such as continuous type and discrete type) can be processed without considering the problem of inconsistent data types;
(3) In addition, the invention adopts principal component analysis to reduce the dimension, applies a large amount of multi-source heterogeneous data to the ship to reduce the dimension, and eliminates the repeated attribute;
(4) Finally, knowledge acquisition is carried out by adopting the rough set, so that hidden rules in a ship coating process database and a defect case library can be found, and powerful guidance is provided for ship coating work.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a flow chart of a reduction algorithm of attribute importance of the present invention;
FIG. 3 is a diagram of a principal component analysis lithotripsy of the present invention.
Detailed Description
The technical scheme of the invention is further described below with reference to the accompanying drawings and examples.
The ship coating defect knowledge acquisition method based on the PCA-rough set as shown in fig. 1 comprises the following steps:
s1, selecting a target data set:
selecting a target data set from a ship coating process database and a ship coating defect case library, wherein the target data set comprises attribute information, parameter information and environment information, and the attribute information comprises a coating area, surface roughness, a rust removal grade, a coating method, coating equipment and the like; the parameter information comprises paint viscosity, aerodynamic force, spraying distance, paint transfer rate and the like; the environmental information includes air flow rate, relative humidity, air temperature, etc.
S2, data checking and data preprocessing:
checking and preprocessing the selected target data set,
(1) Performing data inspection, including data missing value inspection, inconsistent data inspection and data noise value inspection, and judging whether the data needs to be processed or not;
(2) The data to be processed is correspondingly preprocessed, including missing value processing, rationality checking and noise value processing, the specific means include automatic deleting, filling and correcting processing, whether the processing is complete is judged after the processing is finished, if not, the data is returned to be rechecked and processed, otherwise, the next step is carried out.
Taking a defect case with a coating area as an outer plate, a coating area of 200-400 square meters and a rust removal grade of Sa2.0-Sa3.0 as a data set; and directly removing samples with the defect data of the ship coating missing more than 1/3 of the total data, and carrying out data filling on the samples with the defect data not more than 1/3 of the total data by adopting an average number. Anomalies in data include inconsistent data and noisy data, which are often related to inconsistencies in the ship process measured input and output variables or some other factor, anomalous data such as individual samples being too far apart from most samples on some variables, and the probability of this value appearing very low from the statistical distribution of samples, data anomalies can be corrected by short time interval estimation or artificial filling.
S3, carrying out principal component analysis and dimension reduction on the preprocessed data:
multi-source heterogeneous data of ship coating defects are subjected to dimension reduction treatment by utilizing principal component analysis
Specifically, a one-hot coding mode is adopted to code the relevant attribute of the painting defect of the discrete ship, wherein the relevant attribute of the painting defect of the discrete ship comprises painting equipment, paint types, painting methods, operation groups and the like; solving a covariance matrix of the preprocessed data, calculating characteristic values and characteristic vectors of the covariance matrix, sorting according to the characteristic values, calculating a variance contribution rate and a variance accumulation contribution rate, interactively determining the number of principal components, obtaining principal component values, and inputting variables corresponding to the principal components as new data, so that dimension reduction processing of a large amount of multi-source heterogeneous data in the ship coating defect data is realized.
S4: and (3) carrying out knowledge acquisition on the ship coating defect data by utilizing a rough set theory:
the method comprises the following steps:
(a) Establishing an initial case decision table, and specifically dividing condition attributes and decision attributes;
comprehensively considering various factors which can influence the final formation of the coating defect as condition attributes, wherein the condition attributes comprise process attributes, process parameters and environmental information; the process attributes comprise a coating area, surface roughness, a rust removal grade and the like, the process parameters comprise coating viscosity, dry film thickness, solid content, nozzle distance and the like, and the environmental information comprises air temperature, relative humidity, air flow rate and the like; the decision attribute serves as the final defect name.
(b) Performing knowledge reduction;
under the condition of not influencing the knowledge expression capability, adopting a reduction algorithm based on the attribute importance to carry out attribute reduction, eliminating redundant knowledge, and calculating the importance of each reduced condition attribute to the decision attribute.
The above-described attribute importance reduction algorithm as shown in fig. 2 includes the steps of:
I. obtaining a Core attribute set Core by adopting a difference matrix calculation;
II. Calculating the dependence gamma of the whole decision table before reduction c (D) And the importance σ of each condition attribute other than the kernel attribute CD (c i )；
III, calculating the dependency gamma of the core attribute Core (D)；
IV, if gamma Core (D)≥γ C (D) The algorithm ends and the output attribute is about Jian Jige Core; if gamma is Core (D)＜γ C (D) Ordering the attributes from big to small according to the importance of the attributes c i (i=1, 2, … n) added to the set Core one by one for (i=1:n), core=core ∈ i Calculating gamma Core (D) If gamma is Core (D)≥γ C (D) The algorithm ends and the output attribute is about Jian Jige Core; if gamma is Core (D)＜γ C (D) The cycle continues.
(c) Performing attribute value reduction;
and removing redundant attribute values to obtain a simplified decision data table.
(d) Extracting rule knowledge;
and merging relevant rule knowledge according to the simplified attribute decision data table.
S5: classifying and storing knowledge
And calculating the support degree, coverage degree and credibility of each ship coating defect knowledge, dividing the knowledge into deterministic knowledge, strong rule knowledge and weak rule knowledge, and storing the knowledge into a knowledge base.
Credibility definition:
support definition:
coverage definition:
wherein, C (x) ≡D (x) | is the total number of samples satisfying the decision rule C (x) →D (x); the number of samples of the front piece C (x) satisfying the decision rule is C (x). And U is the total number of samples of the whole domain. Where |d (x) | is the total number of samples of the back-piece D (x) that satisfy the decision rule.
(1) Significance knowledge: the confidence level is 1, the support level is 1, and the coverage level is 1.
(2) Strong rule knowledge: the credibility is more than or equal to 0.5 and less than 1, the supportability is more than or equal to 0.5 and less than 1, and the coverage is more than or equal to 0.5 and less than 1.
(3) Weak rule knowledge: the credibility is more than 0 and less than 0.5, the support is more than 0 and less than 0.5, and the coverage is more than or equal to 0.5 and less than 1.
As shown in fig. 3, a principal component analysis lithotripsy chart can be used to help determine the optimal number of principal components, wherein the abscissa in the lithotripsy chart represents the number of principal components, and the ordinate represents the eigenvalue, and the continuous steep part of the principal component eigenvalue is the number of principal components to be taken.
As can be seen intuitively from table 1, the cumulative variance contribution rate of the eigenvalues of the first 3 principal components reaches 85%, so the first 3 components are selected instead of the original variables.
TABLE 1 eigenvalues and variance contribution rates
The factor loading of each factor on the different principal components was obtained using a standardized orthogonal rotation method, and the resulting component matrices are shown in table 2.
TABLE 2 component matrix
The ship coating process data is reduced to 13 variables based on a principal component analysis method, namely coating viscosity, wet film thickness, nozzle distance, air temperature, relative humidity, air flow rate, recoating time, nozzle aperture, spraying breadth, surface roughness, coating equipment, diluent content and operation team, wherein the factors comprise ship coating defect influence factors determined by ship coating field experts through experience.
Firstly, solving a core attribute set based on a reduction algorithm of attribute importance, and then adding the core attribute set one by one from large to small according to an attribute importance value; the dependency value after the reduction of the decision table is not lower than the dependency value before the reduction, and the step IV is specifically described.
Table 2 is processed according to the reduction algorithm described above, and the reduction results are obtained after the 13 attributes in table 2 are eliminated by 6 redundant attributes.
A decision table was established and the paint viscosity, wet film thickness, nozzle distance, air temperature, relative humidity, air flow rate were selected as condition attributes, and the meanings represented by the condition attributes are shown in table 3. Decision attributes include D1 wrinkles, D2 sagging, D3 orange peel, D4 cracking, D5 pinholes, D6 blistering.
TABLE 3 Attribute reduction results
The results of attribute reduction and value reduction result in a decision data table, as shown in table 4.
Table 4 decision data table
From the decision table of table 4, the following rule knowledge can be obtained:
TABLE 5 rule knowledge table
The meaning of the above knowledge is:
X 1 : IF (paint viscosity too low) and (wet film thickness too thin) and (nozzle distance too far) and (air temperature normal) and (relative humidity too high) and (air flow normal) and (recoating time normal) THEN d=orange peel. The knowledge is deterministic knowledge.
X 2 : IF (normal paint viscosity) and (too thick wet film) and (too close nozzle distance) and (too high air temperature) and (normal relative humidity) and (normal air flow) and (too thin recoating time) THEN d=pinhole. The knowledge is a strong rule knowledge.
X 3 : IF (paint viscosity too high) and (wet film thickness normal) and (nozzle distance too close) and (air temperature normal) and (relative humidity too low) and (air flow rate too fast) and (recoating time normal) THEN D = wrinkles. The knowledge is a weak rule knowledge.
X 4 : IF (coating viscosity too high) and (wet film thickness too thin) and (nozzle distance too low) and (relative humidity normal) and (air flow rate normal) and (recoating time too long) THEN d=crack. The knowledge is a weak rule knowledge.
X 5 : IF (normal paint viscosity) and (normal wet film thickness) and (too close nozzle distance) and (too high air temperature) and (too high relative humidity) and (normal air flow) and (too short recoating time) THEN D = sag. This knowledge is significant knowledge.
X 6 : IF (paint viscosity too low) and (wet film thickness too thick) and (nozzle distance normal) and (air temperature too high) and (relative humidity normal) and (air flow rate normal) and (recoating time too long) THEN d=foam. The knowledge is a strong rule knowledge.
Claims (6)
1. The ship coating defect knowledge acquisition method based on the PCA-rough set is characterized by comprising the following steps of:
(1) Selecting a target data set from a ship coating process database and a ship coating defect case library;
(2) Checking and preprocessing the selected target data set;
(3) Performing dimension reduction treatment on the multi-source heterogeneous data of the ship coating defects by using principal component analysis;
(4) Carrying out knowledge acquisition on the ship coating defect data by utilizing a rough set theory;
(5) Classifying and storing the knowledge;
the method comprises the following steps of (3) specifically adopting a one-hot coding mode to code the relevant attribute of the coating defect of the discrete ship, wherein the relevant attribute of the coating defect of the discrete ship comprises coating equipment, coating types, a coating method and a working team; solving a covariance matrix of the preprocessed data, calculating characteristic values and characteristic vectors of the covariance matrix, sorting according to the characteristic values, calculating a variance contribution rate and a variance accumulation contribution rate, interactively determining the number of principal components to obtain principal component values, and inputting variables corresponding to the principal components as new data, so as to realize dimension reduction processing of a large amount of multi-source heterogeneous data in the ship coating defect data;
the step (4) comprises the following steps:
(a) Establishing an initial case decision table, and specifically dividing condition attributes and decision attributes;
(b) Performing knowledge reduction, and calculating the importance of each reduced condition attribute to the decision attribute;
(c) Performing attribute value reduction to obtain a simplified decision data table;
(d) Extracting rule knowledge;
wherein, in the step (a), a plurality of factors which can influence the final formation of the coating defect are comprehensively considered as condition attributes, wherein the condition attributes comprise process attributes, process parameters and environmental information; the technological properties comprise a coating area, surface roughness and a rust removal grade, the technological parameters comprise coating viscosity, dry film thickness, solid content and nozzle distance, and the environmental information comprises air temperature, relative humidity and air flow rate; the decision attribute is used as a defect name formed finally;
and (b) carrying out attribute reduction by adopting a reduction algorithm based on attribute importance under the condition of not affecting knowledge expression capability, eliminating redundant knowledge, and calculating the importance of each reduced condition attribute to decision attribute.
2. The ship painting defect knowledge acquisition method based on PCA-rough set according to claim 1, wherein: the target data set in the step (1) comprises attribute information, parameter information and environment information, wherein the attribute information comprises a coating area, surface roughness, a rust removal grade, a coating method and coating equipment; the parameter information comprises paint viscosity, aerodynamic force, spraying distance and paint transfer rate; the environmental information includes air flow rate, relative humidity, air temperature.
3. The ship painting defect knowledge acquisition method based on the PCA-rough set according to claim 1 or 2, characterized in that: the data inspection in the step (2) comprises data missing value inspection, inconsistent data inspection and data noise value inspection, and whether the data needs to be preprocessed or not is judged.
4. A ship painting defect knowledge acquisition method based on PCA-rough set according to claim 3, characterized in that: the data preprocessing in the step (2) comprises the steps of missing value processing, rationality checking and noise value processing, and the specific means comprise automatic deleting, filling and correcting processing, judging whether the processing is complete after the processing is finished, returning to recheck and data processing if the processing is incomplete, and otherwise, carrying out the next step.
5. The ship painting defect knowledge acquisition method based on PCA-rough set according to claim 1, wherein: the attribute importance reduction algorithm comprises the following steps:
I. obtaining a Core attribute set Core by adopting a difference matrix calculation;
II. Calculating the dependence gamma of the whole decision table before reduction c (D) And the importance σ of each condition attribute other than the kernel attribute CD (c i )；
III, calculating the dependency gamma of the core attribute Core (D)；
IV、If gamma is Core (D)≥γ C (D) The algorithm ends and the output attribute is about Jian Jige Core; if gamma is Core (D)＜γ C (D) Ordering the attributes from big to small according to the importance of the attributes c i (i=1, 2, … n) added to the set Core one by one for (i=1:n), core=core ∈ i Calculating gamma Core (D) If gamma is Core (D)≥γ C (D) The algorithm ends and the output attribute is about Jian Jige Core; if gamma is Core (D)＜γ C (D) The cycle continues.
6. The ship painting defect knowledge acquisition method based on PCA-rough set according to claim 1, wherein: and (5) specifically, calculating the support degree, coverage degree and credibility of each ship coating defect knowledge, dividing the knowledge into deterministic knowledge, strong rule knowledge and weak rule knowledge, and storing the knowledge into a knowledge base.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910571065.4A CN110297852B (en) | 2019-06-28 | 2019-06-28 | Ship coating defect knowledge acquisition method based on PCA-rough set |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910571065.4A CN110297852B (en) | 2019-06-28 | 2019-06-28 | Ship coating defect knowledge acquisition method based on PCA-rough set |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110297852A CN110297852A (en) | 2019-10-01 |
| CN110297852B true CN110297852B (en) | 2023-04-21 |
Family
ID=68029286
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910571065.4A Active CN110297852B (en) | 2019-06-28 | 2019-06-28 | Ship coating defect knowledge acquisition method based on PCA-rough set |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110297852B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111382579A (en) * | 2020-01-13 | 2020-07-07 | 中船第九设计研究院工程有限公司 | Data preprocessing verification platform of ship pipeline manufacturing execution system |
| CN111539645A (en) * | 2020-04-30 | 2020-08-14 | 西安工业大学 | Assembly knowledge discovery method based on rough set |
| CN111767943B (en) * | 2020-05-20 | 2024-06-11 | 北京简巨科技有限公司 | Mulch film identification method and device, electronic equipment and storage medium |
| CN111932074B (en) * | 2020-07-09 | 2023-10-27 | 江苏科技大学 | Knowledge-based intelligent ship coating process recommendation method |
| CN112246469B (en) * | 2020-10-13 | 2022-02-18 | 华晨宝马汽车有限公司 | Method and apparatus for controlling spray coating of vehicle |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104636493A (en) * | 2015-03-04 | 2015-05-20 | 浪潮电子信息产业股份有限公司 | Dynamic data grading method based on multi-classifier fusion |
| CN105718726A (en) * | 2016-01-18 | 2016-06-29 | 沈阳工业大学 | Medical auxiliary examination system knowledge acquisition and inference method based on rough set |
-
2019
- 2019-06-28 CN CN201910571065.4A patent/CN110297852B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104636493A (en) * | 2015-03-04 | 2015-05-20 | 浪潮电子信息产业股份有限公司 | Dynamic data grading method based on multi-classifier fusion |
| CN105718726A (en) * | 2016-01-18 | 2016-06-29 | 沈阳工业大学 | Medical auxiliary examination system knowledge acquisition and inference method based on rough set |
Non-Patent Citations (1)
| Title |
|---|
| 基于PCA的医疗数据特征提取方法研究及应用;王双维;《中国优秀硕士学位论文全文数据库信息科技辑》;20090115;第I138-705页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110297852A (en) | 2019-10-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110297852B (en) | Ship coating defect knowledge acquisition method based on PCA-rough set | |
| CN106206346B (en) | Measurement Sampling Method with Sampling Rate Determination Mechanism | |
| CN110543616B (en) | SMT solder paste printing volume prediction method based on industrial big data | |
| CN113515512B (en) | Quality control and improvement method for industrial Internet platform data | |
| TWI783400B (en) | Device and method for estimating error factor | |
| CN108960334A (en) | A kind of multi-sensor data Weighted Fusion method | |
| CN113642666A (en) | Active enhanced soft measurement method based on sample expansion and screening | |
| CN110705841A (en) | Chemical production safety assessment method based on improved fuzzy analytic hierarchy process | |
| CN108830492A (en) | A method of sampling observation businessman is determined based on big data | |
| WO2024109954A1 (en) | Mineralization coefficient prediction method and apparatus for sandstone-type uranium ore | |
| WO2020220438A1 (en) | Method for predicting concurrent volume of services of different types for virtual machine | |
| Subali et al. | A new model for measuring the complexity of SQL commands | |
| US11829124B2 (en) | Method for predicting occurrence of tool processing event and virtual metrology application and computer program product thereof | |
| CN117892945A (en) | Construction engineering cost measuring and calculating method based on neural network | |
| CN112700050A (en) | Method and system for predicting ultra-short-term 1 st point power of photovoltaic power station | |
| CN116401570A (en) | Intelligent processing system for printing quality monitoring big data | |
| Liu et al. | The man-hour estimation models & its comparison of interim products assembly for shipbuilding | |
| CN115689331A (en) | Power transmission and transformation project quantity rationality analysis method based on MLP | |
| CN115841184A (en) | Working hour prediction method based on process clustering | |
| CN111160464B (en) | Industrial high-order dynamic process soft measurement method based on multi-hidden-layer weighted dynamic model | |
| CN110673470B (en) | Industrial non-stationary process soft measurement modeling method based on local weighting factor model | |
| Schuh et al. | Achieving higher scheduling accuracy in production control by implementing integrity rules for production feedback data | |
| CN114002204A (en) | Laser-induced breakdown spectroscopy analysis method based on spectral jitter | |
| CN111507003A (en) | Method for predicting key variables of debutanizer based on quality-related dynamic characteristic extraction | |
| CN112329535A (en) | CNN-based quick identification method for low-frequency oscillation modal characteristics of power system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |