CN107506565A - Method for building up based on neutral net alloy cast steel roll Expert System for Materials Design - Google Patents
Method for building up based on neutral net alloy cast steel roll Expert System for Materials Design Download PDFInfo
- Publication number
- CN107506565A CN107506565A CN201710958750.3A CN201710958750A CN107506565A CN 107506565 A CN107506565 A CN 107506565A CN 201710958750 A CN201710958750 A CN 201710958750A CN 107506565 A CN107506565 A CN 107506565A
- Authority
- CN
- China
- Prior art keywords
- cast steel
- alloy cast
- neutral net
- expert system
- steel roll
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/06—Power analysis or power optimisation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to alloy cast steel roll technical field, is related to the method for building up based on neutral net alloy cast steel roll Expert System for Materials Design, including the module such as the optimization of integrated database, inference machine, knowledge base, performance and human-computer interaction interface.It is stored in for training and testing the sample set of neutral net in the database of expert system；The knowledge rule of the composition, technique and the performance that are obtained during training neutral net is stored in knowledge base with a matrix type, and the deposit of knowledge is provided for the reasoning of system；Performance optimization module is judged performance prediction result automatically automatically, and is returned in the integrated database of system；Each module of expert system has the human-computer interaction interface of close friend.The advantages that present invention applies the expert system of artificial neural network, and knowledge maintenance is convenient, inference speed is fast, while reduce the blindness during alloy cast steel roll design of material, save time and cost.
Description
Technical field
Present invention relates particularly to alloy cast steel roll technical field, more particularly to based on neutral net alloy cast steel roll material
Expect the method for building up of design specialist's system.
Background technology
Alloy cast steel occupies highly important status in roll material development.With Modern Rolling Mill to high pressure and at a high speed roll
The direction of system is developed, and the requirement to roll becomes increasingly harsher, and further requirement increases the quality and combination property of roll.Cause
This, certainly will will develop high performance high-alloy steel roll.In order to reach the target for reducing development cost and improving roller performance, just
It is necessary to carry out performance prediction to the material of roll, further according to the result of prediction, its composition and Technology for Heating Processing is designed,
Production cost can be reduced, and shortens the development time.
During new roll material is researched and developed, alloying element is added to roll material and improves its alloying journey
Degree, roll material alloy design is carried out, be an important measures for improving roller performance.Divided using mathematical statistics method as returned
It is more commonly used method that analysis, which carries out Optimization of Material Property, but due to the uncertain and non-linear of many problems be present, because
This is difficult with mathematical description before recurrence and establishes accurate mathematical modeling.Therefore, opening for new material is carried out using computer
Hair and the inexorable trend that research is design of material development.Neutral net is realized based on true and example by numerical computations
Shallow hierarchy experience inference, and the rule-based profound symbolic logic reasoning of expert system.The present invention is according to the respective of the two
Feature, learn from other's strong points to offset one's weaknesses, both are scientifically combined to form to the expert system based on neural network model, by the expert system application
In alloy cast steel roll investigation of materials and development, tool is of great significance.
The content of the invention
It is an object of the invention to overcome the above-mentioned problems in the prior art, there is provided based on neutral net alloy cast steel
The method for building up of roll material design specialist's system；
Method for building up based on neutral net alloy cast steel roll Expert System for Materials Design, it is characterised in that including such as
Lower step：
Step S1：The foundation of integrated database, gather the mechanical property test and contact fatigue of alloy cast steel roll material
Performance test results, the static database of alloy cast steel alloying component, technological parameter and performance is established, passes through data handling system
Feature extraction is carried out, establishes caused dynamic data base during neural computing；
Step S2：The foundation of base module, according to the example sample of offer, by BP learning algorithms to sample
Practise, the heuristic knowledge for expert being solved practical problem is preserved with a matrix type, and implicitly scattered is stored in nerve
In the every weights and threshold file of network connection, so as to constitute the knowledge base of expert system；
Step S3：The foundation of inference machine；
Step S4：The foundation of performance optimization module, according to the result of system prediction, research and analysis main alloy element and
Heat treatment process parameter makes the optimization of composition, technique and performance to mechanical property and the affecting laws of contact fatigue property；
Step S5：Integrated and human-computer interaction interface the realization of system.
Further, in the step S1, database can realize the inquiry of learning sample data, increase, reduction, clear
Look at, normalize, the adjustment operation of renormalization, network architecture parameters.
Further, in the step S2, unified digital number is carried out to the knowledge in knowledge base.
Further, in the step S3, the reasoning of system is completed by BP neural network system, is comprised the following steps：
The principal element specification for influenceing alloy cast steel roll performance to respective bins, is formed input value, step S3.2 exists by step S3.1
System calculates hidden layer and the output of output layer neuron automatically on the basis of this, and step S3.3 is obtained during calling neural network learning
The weights and threshold file obtained, complete the inference function of the expert system.
Further, in the step S3, inference machine takes the mode of parallel inference, and replaces symbol using numerical value reasoning
The process of number reasoning, the forward calculation for passing through neutral net produce the output of neutral net.
Further, in the step S4, performance optimization module can be judged performance prediction result automatically.
Further, in the step S5, the program of neural network prediction mechanical property and contact fatigue life model is compiled
Write and use MATLAB programming languages, man-machine interface is write using VB6.0, and VB is realized using the method for ActiveX parts
To MATLAB calling, so as to realize the connection of the calculating of numerical value and interface.
Further, in the step S5, not only there is user interface initialization display, environmental variance to set, dynamically adjust
Enter to link the functions such as library file, menu call submodule can also be utilized.
The present invention income effect be：
It can be facilitated using the present invention, accurate and the quick mechanical property for predicting alloy cast steel roll and contact
Fatigue behaviour, and prediction result can be optimized, it is possible to reduce the blindness during alloy cast steel roll design of material, section
Substantial amounts of time and cost are saved, there is obvious practical value.
Brief description of the drawings
In order to illustrate the technical solution of the embodiments of the present invention more clearly, used required for being described below to embodiment
Accompanying drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for ability
For the those of ordinary skill of domain, on the premise of not paying creative work, it can also be obtained according to these accompanying drawings other attached
Figure.
Accompanying drawing 1 is the flow chart of method for building up of the present invention；
Accompanying drawing 2 is the main window of expert system；
Accompanying drawing 3 is the data base querying interface in expert system；
Accompanying drawing 4 is importing sample data interface；
Accompanying drawing 5 is that network parameter sets and trained interface；
Accompanying drawing 6 is output training result interface；
Accompanying drawing 7 is expert system prediction interface；
In accompanying drawing, the list of parts representated by each label is as follows：
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained all other under the premise of creative work is not made
Embodiment, belong to the scope of protection of the invention.
As shown in figs. 1-7, the present invention is a kind of building based on neutral net alloy cast steel roll Expert System for Materials Design
Cube method, it is characterised in that comprise the following steps：
Step S1：The foundation of integrated database, gather the mechanical property test and contact fatigue of alloy cast steel roll material
Performance test results, the static database of alloy cast steel alloying component, technological parameter and performance is established, passes through data handling system
Feature extraction is carried out, establishes caused dynamic data base during neural computing；
Step S2：The foundation of base module, according to the example sample of offer, by BP learning algorithms to sample
Practise, the heuristic knowledge for expert being solved practical problem is preserved with a matrix type, and implicitly scattered is stored in nerve
In the every weights and threshold file of network connection, so as to constitute the knowledge base of expert system；
Step S3：The foundation of inference machine；
Step S4：The foundation of performance optimization module, according to the result of system prediction, research and analysis main alloy element and
Heat treatment process parameter makes the optimization of composition, technique and performance to mechanical property and the affecting laws of contact fatigue property；
Step S5：Integrated and human-computer interaction interface the realization of system.
Wherein, in step S1, database can realize the inquiries of learning sample data, increase, reduce, browse, normalizing,
Renormalization, network architecture parameters adjustment operation.
Wherein, in step S2, unified digital number is carried out to the knowledge in knowledge base.
Wherein, in step S3, the reasoning of system is completed by BP neural network system, is comprised the following steps：Step S3.1
By the principal element specification for influenceing alloy cast steel roll performance to respective bins, input value is formed, step S3.2 is on this basis
System calculates hidden layer and the output of output layer neuron automatically, and step S3.3 calls the weights obtained during neural network learning
And threshold file, complete the inference function of the expert system.
Wherein, in step S3, inference machine takes the mode of parallel inference, and replaces symbolic reasoning using numerical value reasoning
Process, the output of neutral net is produced by the forward calculation of neutral net.
Wherein, in step S4, performance optimization module can be judged performance prediction result automatically.
Wherein, in step S5, the programming of neural network prediction mechanical property and contact fatigue life model uses
MATLAB programming languages, man-machine interface are write using VB6.0, and VB pairs is realized using the method for ActiveX parts
MATLAB calling, so as to realize the connection of the calculating of numerical value and interface.
Wherein, in step S5, not only there is user interface initialization display, environmental variance to set, dynamically call in chained library
The functions such as file, menu call submodule can also be utilized.
One of the present embodiment has particular application as：
First after user's activation system, user logs into main window interface, as shown in Figure 2.After logging in main interface, user
It can be operated accordingly into the module with module control in click function block region.User is also an option that tree
List items are as clicked on " neural network learning training ", and user can sign in neural metwork training interface, and this is the core of inference machine
Heart part.User can by the browsing of the database realizing data, inquire about, add, delete, the operation such as edit, and
Neural network prediction result will be also preserved in corresponding database, can be updated and be expanded learning sample storehouse, be later nerve
E-learning and training are prepared.Now user can be preferentially inquired about database according to required performance indications, inquires about boundary
Face is as shown in figure 3, if the material for meeting design requirement can be inquired, then it is in optimized selection using optimization system；
If inquiry is unsuccessful, user calls neutral net to make inferences prediction as needed.Neural metwork training and study are that this is
The core of system reasoning, the weights and threshold value obtained by neural network learning and training are stored in knowledge with a matrix type
In storehouse, information is provided for Inference Forecast.Neutral net could carry out the prediction of performance only after study and training.Nerve
The function that network training interface is realized includes：The importing of sample data, the setting of network parameter and training, training result it is defeated
Go out, their interface difference is as shown in Figure 4,5, 6.When carrying out performance prediction, system then utilize in knowledge base relevant material into
Point, relation knowledge rule makes the prediction of performance between Technology for Heating Processing, performance, finally shown in the prediction interface of the system
Go out prediction result, prediction interface is as shown in Figure 7.System is optimized and analyzed to the combination property of material according to prediction result,
Therefrom filter out the material for best suiting condition requirement.
In the description of this specification, the description of reference term " one embodiment ", " example ", " specific example " etc. means
At least one embodiment of the present invention is contained in reference to the features such as embodiment or the specific features of example description, structure, material
Or in example.In this manual, identical embodiment or example are not necessarily referring to the schematic representation of above-mentioned term.And
And specific features, structure, material or the feature of description can be in any one or more embodiments or example with suitable
Mode combine.
Present invention disclosed above preferred embodiment is only intended to help and illustrates the present invention.Preferred embodiment is not detailed
All details are described, it is only described embodiment also not limit the invention.Obviously, according to the content of this specification,
It can make many modifications and variations.This specification is chosen and specifically describes these embodiments, is to preferably explain the present invention
Principle and practical application so that skilled artisan can be best understood by and utilize the present invention.The present invention is only
Limited by claims and its four corner and equivalent.
Claims (8)
1. the method for building up based on neutral net alloy cast steel roll Expert System for Materials Design, it is characterised in that including as follows
Step：
Step S1：The foundation of integrated database, gather the mechanical property test and contact fatigue property of alloy cast steel roll material
Result of the test, the static database of alloy cast steel alloying component, technological parameter and performance is established, carried out by data handling system
Feature extraction, establish caused dynamic data base during neural computing；
Step S2：The foundation of base module, according to the example sample of offer, sample is learnt by BP learning algorithms,
The heuristic knowledge for expert being solved practical problem is preserved with a matrix type, and implicitly scattered is stored in neutral net
In the every weights and threshold file of connection, so as to constitute the knowledge base of expert system；
Step S3：The foundation of inference machine；
Step S4：The foundation of performance optimization module, according to the result of system prediction, research and analysis main alloy element and Re Chu
The affecting laws of science and engineering skill Parameters On Mechanical and contact fatigue property, and make the optimization of composition, technique and performance；
Step S5：Integrated and human-computer interaction interface the realization of system.
2. according to the method for building up based on neutral net alloy cast steel roll Expert System for Materials Design described in claim 1,
It is characterized in that：In the step S1, database can realize the inquiries of learning sample data, increase, reduce, browses, normalizing
Change, renormalization, network architecture parameters adjustment operation.
3. according to the method for building up based on neutral net alloy cast steel roll Expert System for Materials Design described in claim 1,
It is characterized in that：In the step S2, unified digital number is carried out to the knowledge in knowledge base.
4. according to the method for building up based on neutral net alloy cast steel roll Expert System for Materials Design described in claim 1,
It is characterized in that：In the step S3, the reasoning of system is completed by BP neural network system, is comprised the following steps：Step
The principal element specification for influenceing alloy cast steel roll performance to respective bins, is formed input value, step S3.2 is in this base by S3.1
System calculates hidden layer and the output of output layer neuron automatically on plinth, is obtained during step S3.3 calling neural network learnings
Weights and threshold file, complete the inference function of the expert system.
5. according to the foundation side based on neutral net alloy cast steel roll Expert System for Materials Design described in claim 1 or 4
Method, it is characterised in that：In the step S3, inference machine takes the mode of parallel inference, and replaces symbol using numerical value reasoning
The process of reasoning, the output of neutral net is produced by the forward calculation of neutral net.
6. according to the method for building up based on neutral net alloy cast steel roll Expert System for Materials Design described in claim 1,
It is characterized in that：In the step S4, performance optimization module can be judged performance prediction result automatically.
7. according to the method for building up based on neutral net alloy cast steel roll Expert System for Materials Design described in claim 1,
It is characterized in that：In the step S5, the programming of neural network prediction mechanical property and contact fatigue life model uses
MATLAB programming languages, man-machine interface are write using VB6.0, and VB pairs is realized using the method for ActiveX parts
MATLAB calling, so as to realize the connection of the calculating of numerical value and interface.
8. according to the foundation side based on neutral net alloy cast steel roll Expert System for Materials Design described in claim 1 or 7
Method, it is characterised in that：In the step S5, not only there is user interface initialization display, environmental variance to set, dynamically call in chain
The functions such as library file are connect, menu call submodule can also be utilized.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710958750.3A CN107506565A (en) | 2017-10-16 | 2017-10-16 | Method for building up based on neutral net alloy cast steel roll Expert System for Materials Design |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710958750.3A CN107506565A (en) | 2017-10-16 | 2017-10-16 | Method for building up based on neutral net alloy cast steel roll Expert System for Materials Design |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107506565A true CN107506565A (en) | 2017-12-22 |
Family
ID=60701028
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710958750.3A Pending CN107506565A (en) | 2017-10-16 | 2017-10-16 | Method for building up based on neutral net alloy cast steel roll Expert System for Materials Design |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107506565A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110010210A (en) * | 2019-03-29 | 2019-07-12 | 北京科技大学 | Multicomponent alloy composition design method based on machine learning and performance oriented requirement |
| CN110501983A (en) * | 2019-07-31 | 2019-11-26 | 农业农村部南京农业机械化研究所 | Expert control system and control method based on batch seed-coating machine |
| CN111008738A (en) * | 2019-12-04 | 2020-04-14 | 云南锡业集团（控股）有限责任公司研发中心 | Sn-Bi alloy elongation and tensile strength prediction method based on multi-mode deep learning |
| CN111222623A (en) * | 2018-11-26 | 2020-06-02 | 沈阳高精数控智能技术股份有限公司 | Ceramic glaze spraying robot glaze spraying technological parameter debugging method |
| CN111353255A (en) * | 2018-12-05 | 2020-06-30 | 财团法人工业技术研究院 | Automatic generation system of processing parameters |
| CN111831808A (en) * | 2020-07-16 | 2020-10-27 | 中国科学院计算机网络信息中心 | Data-driven artificial intelligent material prediction system |
| CN113486587A (en) * | 2021-07-06 | 2021-10-08 | 太原科技大学 | Seamless steel pipe perforation process model parameter prediction system based on Matlab |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101782771A (en) * | 2010-03-17 | 2010-07-21 | 东华大学 | Spinning process intelligent optimized design method based on immune neural network |
| CN104331543A (en) * | 2014-10-17 | 2015-02-04 | 中国船舶重工集团公司第七一二研究所 | Fault diagnostic expert system for marine electrical propulsion system and establishing method thereof |
-
2017
- 2017-10-16 CN CN201710958750.3A patent/CN107506565A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101782771A (en) * | 2010-03-17 | 2010-07-21 | 东华大学 | Spinning process intelligent optimized design method based on immune neural network |
| CN104331543A (en) * | 2014-10-17 | 2015-02-04 | 中国船舶重工集团公司第七一二研究所 | Fault diagnostic expert system for marine electrical propulsion system and establishing method thereof |
Non-Patent Citations (1)
| Title |
|---|
| 晋会锦等: "基于神经网络的材料设计专家系统的研究与开发", 《河北省机械工程学会铸造专业委员会、河北省铸造行业协会第十六届铸造年会》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111222623A (en) * | 2018-11-26 | 2020-06-02 | 沈阳高精数控智能技术股份有限公司 | Ceramic glaze spraying robot glaze spraying technological parameter debugging method |
| CN111353255A (en) * | 2018-12-05 | 2020-06-30 | 财团法人工业技术研究院 | Automatic generation system of processing parameters |
| CN110010210A (en) * | 2019-03-29 | 2019-07-12 | 北京科技大学 | Multicomponent alloy composition design method based on machine learning and performance oriented requirement |
| CN110501983A (en) * | 2019-07-31 | 2019-11-26 | 农业农村部南京农业机械化研究所 | Expert control system and control method based on batch seed-coating machine |
| CN110501983B (en) * | 2019-07-31 | 2021-03-26 | 农业农村部南京农业机械化研究所 | Expert control system and control method based on batch-type coating machine |
| CN111008738A (en) * | 2019-12-04 | 2020-04-14 | 云南锡业集团（控股）有限责任公司研发中心 | Sn-Bi alloy elongation and tensile strength prediction method based on multi-mode deep learning |
| CN111831808A (en) * | 2020-07-16 | 2020-10-27 | 中国科学院计算机网络信息中心 | Data-driven artificial intelligent material prediction system |
| CN111831808B (en) * | 2020-07-16 | 2022-04-22 | 中国科学院计算机网络信息中心 | Data-driven artificial intelligent material prediction system |
| CN113486587A (en) * | 2021-07-06 | 2021-10-08 | 太原科技大学 | Seamless steel pipe perforation process model parameter prediction system based on Matlab |
| CN113486587B (en) * | 2021-07-06 | 2022-08-05 | 太原科技大学 | Seamless steel pipe perforation process model parameter prediction system based on Matlab |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107506565A (en) | Method for building up based on neutral net alloy cast steel roll Expert System for Materials Design | |
| Atkeson | Using locally weighted regression for robot learning | |
| CN106709017B (en) | A kind of aid decision-making method based on big data | |
| Lechevalier et al. | A neural network meta-model and its application for manufacturing | |
| Bertoni et al. | Mining data to design value: A demonstrator in early design | |
| CN114118673A (en) | Workshop intelligent fault diagnosis early warning method based on digital twin technology | |
| Garg et al. | Reliability analysis of the engineering systems using intuitionistic fuzzy set theory | |
| CN113627032A (en) | Intelligent decision method for equipment design/maintenance scheme based on digital twin | |
| Abellana | A proposed hybrid root cause analysis technique for quality management | |
| Duan et al. | Transforming industry: Investigating 4.0 technologies for sustainable product evolution in china through a novel fuzzy three-way decision-making process | |
| Mourtzis et al. | A framework for adaptive scheduling in cellular manufacturing systems | |
| CN105911965A (en) | Method for realizing data analysis in data monitoring system | |
| WO2021186338A1 (en) | System and method for determining solution for problem in organization | |
| WO2020085114A1 (en) | Information processing device, information processing method, and program | |
| Kumar et al. | A framework for assessing the evolvability characteristics along with sub-characteristics in AOSQ model using fuzzy logic tool | |
| CN109783876A (en) | Time series models method for building up, device, computer equipment and storage medium | |
| CN109298686A (en) | System and method for using business intelligence for rule-based design and manufacture technology | |
| CN115392113A (en) | Cross-working condition complex electromechanical system residual life prediction system and method | |
| Ma et al. | Application of BP neural networks in garment pattern design system | |
| CN110866607B (en) | Permeation behavior prediction algorithm based on machine learning | |
| Horváth et al. | New product model representation for decisions in engineering systems | |
| JP6898479B2 (en) | Machine learning device, numerical control device, machine tool system, manufacturing system and machine learning method to learn the display of operation menu | |
| CN111047011B (en) | Scene variable automatic deep mining engine system based on machine learning model | |
| Chan et al. | Simulation‐aided design of production and assembly cells in an automotive company | |
| CN111080121B (en) | Control chart pattern recognition method for different chain lengths |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171222 |
|
| RJ01 | Rejection of invention patent application after publication |