CN108829052A - Method for optimizing scheduling and device under a kind of semiconductor production line CPS environment - Google Patents
Method for optimizing scheduling and device under a kind of semiconductor production line CPS environment Download PDFInfo
- Publication number
- CN108829052A CN108829052A CN201810548729.0A CN201810548729A CN108829052A CN 108829052 A CN108829052 A CN 108829052A CN 201810548729 A CN201810548729 A CN 201810548729A CN 108829052 A CN108829052 A CN 108829052A
- Authority
- CN
- China
- Prior art keywords
- production line
- scheduling
- production
- cps
- information
- 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 307
- 238000000034 method Methods 0.000 title claims abstract description 109
- 239000004065 semiconductor Substances 0.000 title claims abstract description 75
- 238000004891 communication Methods 0.000 claims abstract description 38
- 230000008569 process Effects 0.000 claims abstract description 37
- 238000013461 design Methods 0.000 claims abstract description 26
- 238000010276 construction Methods 0.000 claims abstract description 16
- 230000008447 perception Effects 0.000 claims abstract description 12
- 238000003066 decision tree Methods 0.000 claims description 31
- 238000012545 processing Methods 0.000 claims description 23
- 238000004422 calculation algorithm Methods 0.000 claims description 15
- 238000005516 engineering process Methods 0.000 claims description 14
- 230000033001 locomotion Effects 0.000 claims description 9
- 230000009471 action Effects 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 5
- 238000003754 machining Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 9
- 230000008901 benefit Effects 0.000 abstract description 5
- 239000010410 layer Substances 0.000 description 17
- 238000005457 optimization Methods 0.000 description 16
- 238000004088 simulation Methods 0.000 description 15
- 238000011160 research Methods 0.000 description 13
- 230000006870 function Effects 0.000 description 12
- 239000000047 product Substances 0.000 description 12
- 238000012549 training Methods 0.000 description 10
- 230000008859 change Effects 0.000 description 8
- 229920000742 Cotton Polymers 0.000 description 7
- 238000009987 spinning Methods 0.000 description 7
- 238000002474 experimental method Methods 0.000 description 6
- 238000013138 pruning Methods 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 5
- 238000009826 distribution Methods 0.000 description 5
- 239000000463 material Substances 0.000 description 5
- 235000012431 wafers Nutrition 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 239000003245 coal Substances 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 230000001976 improved effect Effects 0.000 description 3
- 238000009776 industrial production Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000035508 accumulation Effects 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000002360 preparation method Methods 0.000 description 2
- 239000002994 raw material Substances 0.000 description 2
- 238000009877 rendering Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 238000012384 transportation and delivery Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 239000006227 byproduct Substances 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000004040 coloring Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 230000001066 destructive effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000013401 experimental design Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 239000011229 interlayer Substances 0.000 description 1
- 238000005468 ion implantation Methods 0.000 description 1
- 239000000155 melt Substances 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000013439 planning Methods 0.000 description 1
- 238000013468 resource allocation Methods 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 239000011265 semifinished product Substances 0.000 description 1
- 210000000697 sensory organ Anatomy 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000000007 visual effect Effects 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] or computer integrated manufacturing [CIM]
- G05B19/41865—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] or computer integrated manufacturing [CIM] characterised by job scheduling, process planning, material flow
-
- 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]
Landscapes
- Engineering & Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- General Factory Administration (AREA)
Abstract
The present invention relates to the method for optimizing scheduling and device under a kind of semiconductor production line CPS environment, the method is specially：The semiconductor production line CPS environment towards scheduling is constructed, all kinds of burst disturbances carry out readjustment degree based on production line multidate information when burst disturbance occurs to perception semiconductor production line in process of production；Described device includes：Communication module realizes the communication with actual production line；CPS environment construction module, for establishing digital three-dimemsional model；Optimizing scheduling designs module, is used for real-time perception production information, triggers readjustment degree in production information variation, and generate new production line scheduling scheme according to production line multidate information, realizes readjustment degree.Compared with prior art, the present invention has many advantages, such as to can be applied to that the complicated manufacture systems such as semiconductor production line, timeliness are high, dispatching effect is good.
Description
Technical field
The present invention relates to Semiconductor Wafer Fabrication Scheduling optimisation techniques, more particularly, to a kind of semiconductor production line CPS environment
Under method for optimizing scheduling and device.
Background technique
Semiconductor production line is a kind of typical complicated manufacture system, has complicated production technology, process-cycle length, to machine
The features such as device required precision is high, cost input is big.How by means of the Production planning and scheduling scheme of optimization bigger receipts are obtained
Benefit is one of semiconductor production manufacturing enterprise the most concern.However, all kinds of disturbance events in semiconductor production process
It happens occasionally, such as equipment fault, rush order, equipment process time deviation.System needs the such event of quick response,
Scheduling scheme is adjusted in due course in necessary situation, to guarantee the efficient operation of production line.
It is found after scientific and technical literature by retrieving and analyzing domestic and international Related Research Domain, in Semiconductor Wafer Fabrication Scheduling
In practical application, both production line model building and optimizing scheduling are usually respectively independent.Some researchs are often opened in production process
Corresponding model was established before beginning and carries out analytical calculation, formulated suitable production scheduling scheme according to acquired results.Such as Chinese special
In benefit application " the semiconductor production line scheduling method based on multi-ant colony optimization " (publication number CN102253662A), Li Li etc. is used
A kind of calculation method of multi-ant colony optimization, takes the bottleneck machining area of production line into consideration, to the best of semiconductor production line
Scheduling scheme carries out parallel search, to realize scheduling.Some other research is then by simulation result and production line historical data
After being compared, specific scheduling strategy is selected for a certain optimization aim.Such as " one kind is for partly leading in Chinese patent application
In efficient scheduling rule selection method of the body production line based on simplified simulation model " (publication number CN105843189A), Cao Zhengcai
Deng the scheduling using on-time-delivery rate and quantum of output as regulation goal, using simplified simulation model in a certain scheduling instance to concentration
Rule carries out Fast Evaluation, obtains optimal case.
In addition, in Chinese patent " wafer manufacture dispatching method and scheduling system in semiconductor integrated circuit production " (bulletin
Number CN104423331B) in, Wang Mian proposes one kind by product information unit, moment table unit, product monitoring unit, priority
The IC wafers manufacture and Production Scheduling System that the components such as computing unit, priority criteria unit, display unit are constituted, root
The priority of each product and batch is calculated according to delivery urgency level, to formulate suitable production scheduling scheme.But it is lacked
Point be cannot according to the real-time Adjusted Option of dynamic change of production line state, former design scheme probably due to equipment fault or
The generations of emergency events such as personnel leave the post and fail.
In U.S. Patent application " method and system (the Method and for semiconductor fabrication factory dynamic assignment
System for dynamic dispatching in semiconductor manufacturing plants) " it is (open
Number：US5612886A in), Weng Yi-Cherng is proposed based on billboard thought, while being considered workpiece priority and being lined up
The dispatching method of time.This method is theoretically close to common FIFO (i.e. first in, first out strategy), in the less feelings of system WIP
There is preferably optimization performance, but in the horizontal higher situation of WIP, the performance decline of the strategy is obvious under condition.Especially in reality
In the semiconductor production line of border, since system uncertain factor is more, WIP number will not usually remain at reduced levels.
In United States Patent (USP) " a kind of semiconductor production process dynamic dispatching method and system (Systems and methods
For dynamic semiconductor process scheduling) " (notification number：US9659799B2 in), it is different from biography
It unites in Semiconductor Wafer Fabrication Scheduling research, is used for factors such as process time, workpiece queue length, high latencies
Static scheduling method, Keith R. propose the dynamic dispatching side that resource allocation is carried out based on the working ability variation of production component
Method reduces the accumulation of buffer area workpiece, and then improves system productivity.But this method is only applicable to specific production line condition
Under, and only consider that system compared with the dynamic changes under steady production state, can not cope with processing and fluctuate biggish system
Variation.
To sum up, being directed to the dispatching method of conventional semiconductors production line, the information system of emulation with decision is carried out, usually
It cannot timely respond to the variation of production line state；Meanwhile production system also can not quickly after optimizing application scheduling scheme, influence
Dispatching effect.Under the background that nowadays all kinds of emerging science and technology continue to bring out and develop, CPS (Cyber-Physical
System) proposition with application of concept provide new thinking for the solution of this problem.
CPS is " industry 4.0 " and the core content that intelligence manufacture is realized.It is by Internet of Things, information communication and big
The technological means such as data analysis, the data information of various kinds of equipment in industrial processes is melted by network and information world depth
Close, and collaboration optimization carried out to production process, thus constitute intelligence, digitlization that virtual information and physical equipment interact,
Networking manufacturing system.
In recent years in the manufacturing and optimizing scheduling field, the correlative study of CPS also has numerous scientific and technical literatures, and successively
There are some invention achievements.Lei Xuefeng is proposed by establishing CPS model analysis to coal preparation plant based on Modified particle swarm optimization
Coal preparation plant's task scheduling algorithm of algorithm carries out dynamic dispatching to coal separation process.Liu Chun Yao et al. is directed to the complexity and isomery of CPS
Property, the CPS multipriority Dynamic Scheduling Strategy based on larger sensor network is had studied, normal precedence level scheduling method is improved
Effect.Human relations are forever bright on the basis of proposing the CPS environment framework and context aware model of isomery interconnection, construct scheduling of resource
Model, optimization algorithm dynamic organization and distribution resource using design, realizes the best mode of service with resource distribution.Meanwhile
The combination of production scheduling is also research direction place under some typical intelligent algorithms and CPS environment.Han J et al. calculates ant colony
Method is introduced among the CPS model under a discrete topology environment, while considering that Network Load Balance and service quality are established
Dispatching algorithm frame, the proof of analog result obtained by simulation software the validity of its method.It is carried out in advance using CPS
Scheduling aspect is surveyed, Zhou B H is for unnecessary task switching purpose in production system is avoided, based on CPS Environment Design one
Kind has the production scheduling model of prediction and dynamic adjustment function in real time, to reduce low priority production task for whole
The influence of body production performance improves the real-time performance of system.In addition to this, T Kaihara et al. incorporates CPS correlation theory
Automatic factory's construction is constructed for production line based on CPS principle in the dynamic dispatching side of product number WIP among management
Method, according to the comparison of simulation result and real system WIP, guidance adjustment scheduling strategy, result greatly improves the life in workshop
Produce performance level, it was demonstrated that the validity of method.
It makes a general survey of after domestic and international correlative study and scientific and technical literature it can be found that being directed to the CPS optimizing scheduling of manufacturing system
Study it is still at an early stage, only stop at according to CPS concept propose system establish frame, or the emulation scene in building
Middle research method for optimizing scheduling, more lacks the application in actual production manufacture system.
In Chinese patent application " a kind of CPS framework towards intelligent cotton spinning production " (publication number CN106530111A),
Bao Jingsong et al. devises a kind of CPS frame towards intelligent cotton spinning production, including perception physical layer, connection communication layer, knowledge
Fused layer, Decision Control layer.The invention is served as theme with the data through cotton spinning production cycle overall process, using Intellisense as core
The heart is calculated as relying on network and magnanimity, realizes the efficient outfit of all kinds of resources in cotton spinning production process, optimizes production stream
Journey improves production efficiency.However, the CPS environment that the invention proposes is only in cotton spinning production field, and cotton spinning produced
Journey is single pipeline-type, and researching value is lower.In addition, the process that resource is equipped in its method, the spy without dynamic change
Property, it can not run through the cotton spinning production process complete period.
In a kind of Chinese patent application " CPS control system and its implementation for industry " (publication number
CN106527383A in), Li Qingxin et al. proposes a kind of CPS control system implementation method for industry, for industry
The demand that Flexible Production complication system merges information physical provides the distributed component Dynamic Discovery under a kind of dynamic environment
With the industrial control unit (ICU) of self-organizing, information physical emerging system function package and driving component-based, the scope of application is wider, behaviour
Make simply, more conveniences can be brought to industrial production.But the invention is mainly directed towards controller in industrial production and sets with bottom
Communication perception between standby, lacks the concrete application of upper layer module, belongs in CPS building research field the more part of bottom.Cause
This, the invention is not related to more in terms of production line scheduling optimization, and research directive significance is little.
In United States Patent (USP) " information physical system and its control method (the Cyber-physical system in virtual machine
And method of monitoring virtual machine thereof) " (notification number：US9417904B2 in), Shin
Et al. establish the information physical system containing the communication middleware under several target controllers and different operating system, pass through
Semanteme control realizes the sync response between wherein different virtual machine.The invention realizes a kind of construction method of CPS environment, tool
There is certain theoretical research value.Its shortcoming is that invention is realized mainly for the framework in CPS theory, and in virtual ring
It is practiced in border, lacks verifying in systems in practice.Therefore, application feasibility of the framework under specific practical problem still has
Wait prove.
In conclusion at present in the industrial production, the building of CPS environment and optimizing scheduling application therein are more dilute
It lacks, especially in the production line scheduling field of more complicated semi-conductor manufacturing system, research achievement is still blank.
CPS correlation theory technology is combined with this kind of complicated manufacture system of semiconductor production line, designs the production under corresponding CPS environment
Line dispatching method can reach more preferably optimizing scheduling effect, have certain researching value, and research approach is feasible.
Summary of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide a kind of semiconductor production lines
Method for optimizing scheduling and device under CPS environment.
The purpose of the present invention can be achieved through the following technical solutions：
The present invention provides the method for optimizing scheduling under a kind of semiconductor production line CPS environment, and this method is specially：Building face
To the semiconductor production line CPS environment of scheduling, semiconductor production line all kinds of burst disturbances in process of production are perceived, are occurring to dash forward
When hair disturbance, readjustment degree is carried out based on production line multidate information, scheduling scheme is adjusted according to production line changing condition in due course.
A kind of optimizing scheduling device under the semiconductor production line CPS environment for realizing the method, including：
Communication module, for realizing the communication with actual production line；
CPS environment construction module, for establishing digital three-dimemsional model corresponding with actual target production line；
Optimizing scheduling designs module, is used for real-time perception production information, triggers readjustment degree, and root in production information variation
New production line scheduling scheme is generated according to production line multidate information, realizes readjustment degree.
Further, the data communication network that the communication module is combined with MQTT communication protocol by OPC technology with
Communication is realized between actual production line.
Further, the CPS environment construction module includes：
Information acquisition unit, for acquiring target production line information by the communication module；
Three-dimensional modeling unit, for establishing the corresponding number of actual target production line according to the target production line information
Word threedimensional model；
Movements design unit, for generating the dynamic action parameter of each component in the digital three-dimemsional model.
Further, in the three-dimensional modeling unit, institute is resettled after pre-processing to the target production line information
State digital three-dimemsional model；
Further, it is described pretreatment include delete can not observe in process of production face, merge disconnect vertex and
Remove isolated vertex.
Further, the CPS environment construction module further includes：
Online display module, for showing production line machining state.
Further, the optimizing scheduling design module includes：
Input layer, for obtaining production line multidate information by the communication module；
Readjustment degree decision-making level occurs burst disturbance for judging whether according to the production line multidate information, and is occurring
Production readjustment degree trigger signal is generated when burst disturbance；
Weight dispatch layer, responds the production readjustment degree trigger signal, for being used according to the production line multidate information
Readjustment degree scheme generation method generates new production line scheduling scheme.
Further, the readjustment degree scheme generation method is CART decision Tree algorithms.
Further, the production line feature used in the CART decision Tree algorithms includes system in product number, each processing
Area's buffer length and faulty equipment number.
Further, the optimizing scheduling design module further includes：
Layer is evaluated, for carrying out compliance test result to the new production line scheduling scheme based on digital three-dimemsional model.
Compared with prior art, the related theory of present invention combination CPS, establishes the semiconductor production line towards scheduling
CPS environment, and propose the method for optimizing scheduling under a kind of semiconductor production line CPS environment, this method can effectively be realized to both
Determine the semiconductor production line perception that all kinds of bursts disturb in process of production, and then is adjusted in due course according to production line changing condition
Degree scheme, with Instructing manufacture process.
Specifically, the present invention has the advantages that following several respects：
1) Unity3D and OPC+MQTT technology is taken, the building of semiconductor production line three-dimensional display model, energy are completed
Enough process that is more three-dimensional, intuitive and comprehensively showing production line, while realizing the number of production line in industry manufacture
Change modeling process.
2) by the efficient communication network of foundation, system can respond rapidly to the state change of production line, and quickly under
Up to new control instruction, the interaction of Information Level and physics interlayer is more frequently and efficient, reduces due to message delay and causes
Production performance downside risk, while also greatly improving the timeliness of new scheduling scheme.
3) generation that the decision-tree model based on machine learning method carries out new scheduling scheme is used, calculating speed is fast；
By means of a large amount of operation datas of Discrete Event Simulation Models as training set, the accuracy in computation of decision-tree model is also obtained
It improves.Therefore, the present invention has more preferably production line scheduling effect of optimization.
Detailed description of the invention
Fig. 1 is the Establishing process schematic diagram of digital three-dimemsional model of the present invention；
Fig. 2 is the communication network architecture figure under semiconductor production line CPS environment of the present invention；
Fig. 3 is the optimizing scheduling frame under semiconductor production line CPS environment of the present invention；
Fig. 4 is the generating process schematic diagram of CART decision tree in the present invention；
Fig. 5 is that semiconductor wisdom manufactures exemplary cell layout in embodiment；
Fig. 6 is the online display platform main interface established in embodiment；
Fig. 7 is the optimizing scheduling module interfaces established in embodiment；
Production line characteristic value obtains interface when Fig. 8 is the generation scheduling scheme established in embodiment；
Fig. 9 is that the new scheduling scheme established in embodiment assigns interface；
Figure 10 is that production performance histogram compares under three kinds of scenes in embodiment；
Figure 11 is that the total processing times histogram of equipment compares under three kinds of scenes in embodiment；
Figure 12 is that failure generation front and back Gantt chart compares in embodiment.
Specific embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention
Premised on implemented, the detailed implementation method and specific operation process are given, but protection scope of the present invention is not limited to
Following embodiments.
The present invention provides the method for optimizing scheduling under a kind of semiconductor production line CPS environment, and this method is constructed towards scheduling
Semiconductor production line CPS environment, burst disturbance is occurring for all kinds of bursts disturbance in process of production of perception semiconductor production line
When, readjustment degree is carried out based on production line multidate information, scheduling scheme is adjusted according to production line changing condition in due course.
Realize that the optimizing scheduling device under the semiconductor production line CPS environment of the above method includes communication module, CPS environment
It constructs module and optimizing scheduling designs module, wherein communication module is for realizing the communication with actual production line；CPS environment structure
Block is modeled for establishing digital three-dimemsional model corresponding with actual target production line；Optimizing scheduling designs module for real
When perceive production information, trigger readjustment degree in production information variation, and generate new production line according to production line multidate information
Scheduling scheme realizes readjustment degree.
Communication module is by OPC technology with real between the data communication network that MQTT communication protocol combines and actual production line
Now communicate.
CPS environment construction module includes：Information acquisition unit, for acquiring target production line money by the communication module
Material；Three-dimensional modeling unit, for establishing the corresponding digitlization of actual target production line according to the target production line data
Threedimensional model；Movements design unit, for generating the dynamic action parameter of each component in the digital three-dimemsional model.
In certain embodiments, CPS environment construction module further includes online display module, for showing that production line processes shape
State.
The optimizing scheduling designs module：Input layer is believed for obtaining production line dynamic by the communication module
Breath, the production line multidate information primitive scheduling scheme, production scene information and external environment change information etc.；Weight scheduling decision
Layer occurs burst disturbance for judging whether according to the production line multidate information, and generates generation when burst disturbance occurs
Formula readjustment degree trigger signal；Weight dispatch layer, responds the production readjustment degree trigger signal, for according to the production line dynamic
Information generates new production line scheduling scheme using readjustment degree scheme generation method.Readjustment degree scheme generation method can be CART
Decision Tree algorithms etc..
In certain embodiments, optimizing scheduling design module further includes evaluation layer, for being based on the Digital Three-Dimensional mould
Type carries out compliance test result to the new production line scheduling scheme.
(1) the CPS environment construction of semiconductor production line
The present invention realizes the three-dimensional display modeling under semiconductor production line CPS environment using Unity3D d engine technology,
It shows its production line machining state, and perceives production information variation；The side combined using OPC technology with MQTT communication protocol
Method establishes effective data communication network, completes threedimensional model and acquires to the data of production line；According to semiconductor production line data
The characteristics of and type, design corresponding relation type database structure, use MySQL software realization data management.Thus it establishes and real
Production line corresponding digital three-dimemsional model in border completes CPS environment construction.
The detailed process for establishing the three-dimensional display model of production line is as shown in Figure 1：
1) data collection.Comprehensive collection and the every data for arranging target production line, main includes the equipment group of production line
At, spatial distribution, shape, the size of equipment and workpiece, the process flow of relative motion relation and production line between each component
Etc. information, for subsequent step use.
2) three-dimensional modeling.Using the d-making and Rendering software 3dmax of profession, the three-dimensional of relevant device and workpiece is established
Animation model.Equipment more complicated for structure is ressembled after can modeling respectively to each component.In model foundation
In the process, emphasis extracts the main information that models in initial data, as the shape of object, dimension information (vertex, vertex index,
Face, face index, normal etc.) and texture, material information, ignore some secondary information such as mass of object, inner wall thickness of modeling
Deng.Meanwhile the simplification in order to guarantee model, the face that can not observe part in process of production is deleted, merges and disconnects vertex simultaneously
Remove isolated vertex.It between the identical model of shape type, is obtained as far as possible using duplication, reduces the consumption to system resource.
3) model beautifies.Continue to use three-dimensional modeling Rendering software 3dmax, to the threedimensional model established carry out rendering and
Beautification.The specific implementation of rendering mainly includes that the operations such as texture, addition material, textures baking, Jin Er are arranged for model
Under the premise of computer hardware resource allows, appropriate aesthetic appeal keeps the operation of model more vivid.
4) system layout.The model completed will be established, is imported in Unity3D engine with the .FBX format of software support, and
It carries out building assembly according to the mutual alignment relation between set system structure distribution and each component.Meanwhile correction model exists
The problems such as after importing because of compatibility material texture that may be present, illumination, coloring.After the completion of unitary construction, angle from different directions
Degree checks the form of production line model, if situations such as there are breakpoint section, picture overlapping, model distortions.When necessary, again into
The modification and importing process of row threedimensional model.
5) movements design.On the basis of static models, the movement of each assembly operating of production line is designed.With true
Each device action is reference in production process, in conjunction with data collection data early period, the movement for each component at runtime of designing a model
The parametric variables such as direction, size and rotation angle.Relevant action perform script is write using C#, it is dynamic that sub-module describes each equipment
Make the variation logic of parameter, and carry is in systems.Unity3D instant preview function moving model at any time is utilized simultaneously, is checked
Movement whether running succeeded and fluency, make necessary modification.
6) functional development.Equally in such a way that C# writes script, in conjunction with the feature card that Unity3D is carried, description system
Function logic needed for system.From user's sense organ level, realize that scene walkthrough, action control, data are shown, frame shows hidden etc.
Function reaches intuitive displaying of the simulation model to real system operational process；It is then to realize data sense for dispatching requirement
Know, monitoring alarm, scheduler module call etc. functions exploitation.
7) system is issued.After the completion of total system Functional Design, runs designed simulation model and test its function
Realize that situation is issued as individual executable file program according to the demand of application platform after confirmation is errorless for using.
(2) communication network is established
Communication network architecture under semiconductor production line CPS environment as shown in Fig. 2, be it is a kind of based on OPC interface technology and
The communication network of MQTT instant messaging agreement.Different characteristics and advantage according to OPC and MQTT relative to respective communication-hierarchy level, will
The two is coupled, and then realizes high-performance by actual production equipment to a variety of application programs, reliable and safe communication mode,
Complete the acquisition demand of real time data.MQTT protocol message issue client terminal is merged exploitation with OPC client to write, makes OPC
Client is further issued while obtaining the creation data of OPC server-side in real time as the message that can subscribe to theme
Into the message proxy server of MQTT, then corresponding MQTT is write by different application and subscribes to client script, acquisition disappears
Push of the server for real-time production data is ceased, specific implementation process can substantially be divided into following four step：
1) opc server is established in the industrial personal computer of equipment end, industrial personal computer is connected with the data source of production information, the service
Device obtains all production information data transmitted by PLC, and starts the socket for communication, waits asking for client
It asks；
2) corresponding OPC client and MQTT client are write in exploitation, are established socket and are communicated with opc server,
Production information data needed for selection are transmitted, while by its classified finishing and being issued as the subscription theme of MQTT, locating
LAN segment in broadcasted；
3) the installation configuration MQTT service on publishing side PC, while MQTT proxy server is set and is run, store client
Hold the MQTT theme sent and information and the subscription for waiting other clients；
4) script is write in Unity3D to be communicated with MQTT proxy server, and subscribes to corresponding theme, it will be corresponding
Data are hung up the script after writing on the model of foundation with the formal definition of global aray variable and reference, make its life
Effect.
The management of semiconductor production data then uses MySQL relevant database, in order to promoted data processing speed and
System flexibility, it then follows semiconductor production line data model structure rule, classification storage are carried in MySQL, and using MySQL
Workbench be managed.
(3) the method for optimizing scheduling design and realization under semiconductor production line CPS environment
The present invention mutually melts sensing capability of the CPS to production information with production line emergency event disturbance (such as equipment fault)
It closes, is studied in conjunction with conventional semiconductors production line readjustment degree, design optimizing scheduling frame, establish weight scheduling mechanism and set with perceiving production
Standby state change.Simultaneously using the CART traditional decision-tree in machine learning, training simultaneously establishes computation model, if necessary again
It generates and reaches production line under new scheduling scheme, realize optimizing scheduling.
The perceptibility that CPS environment disturbs emergency event is good, is suitble to the semiconductor being directed under emergency event disturbance raw
Producing line is scheduled research.Meanwhile emergency event disturbance is destructive to original scheduling scheme big, needs to design correlation method again
Generate new scheduling scheme.In solution of emergent event perturbed problem, original production plan does not adapt to production line scheduling
Therefore demand should be based on event-driven mechanism, the readjustment degree of production is carried out to production line.
Based on semiconductor production line CPS environment, designed optimizing scheduling frame is as shown in Figure 3.Input layer contains respectively
The multidate information of class production line, such as primitive scheduling scheme, production scene information and external environment change information, determine for readjustment degree
Plan layer is analyzed and processed.It, can be real by CPS environment when the bursts such as system occurrence of equipment failure, rush order disturb situation
When perception capture, and in the form of monitoring alarm etc. embody, and then cause production readjustment degree.In weight dispatch layer, based on acquisition
Input layer production scene information generates new production line scheduling scheme using corresponding rescheduling method.Finally in evaluation layer, knot
The Discrete Event Simulation Models of symphysis producing line verify the rescheduling method and the actual effect for generating scheme.
Scheduling scheme generation method is CART decision Tree algorithms in the present invention.Algorithm mainly includes both sides content, point
It is not the generation and beta pruning of decision tree.Decision tree generate process as shown in figure 4, wherein CART decision tree nodes branch according to
According to the size for being Gini coefficient, Gini value is smaller, shows that the degree of purity of sample is higher, i.e., the sample is pertaining only to of a sort probability
It is higher.After the Gini coefficient for measuring out all values of some feature of data set, so that it may obtain the Gini Split of this feature
Info, as Gini Gain.In the case of not considering beta pruning, the process of decision tree recurrence creation selects Gini Gain minimum every time
Node do bifurcation, until Sub Data Set belong to it is same classification or all features all judged after terminate.
Assuming that sample data set D shares the classification of K kind, the other probability of kth type is pk, then the expression formula of Gini coefficient be：
According to the calculation formula of Gini coefficient, if there is the classification of K kind in data set D, the Sub Data Set being made of the classification of kth kind
For Ck, then the Gini coefficient of data set D is represented by following formula.Wherein | Ck| it is CkSample size, | D | be D sample
Quantity：
Gini Gain is also referred to as Gini information gain, for the foundation for judging bifurcation, since CART algorithm generates y-bend
Tree, each node is divided into two according to the size of its feature corresponding eigenvalue, therefore the calculation formula of Gini Gain is：
The input of overall flow is total training dataset, exports the decision-tree model for foundation.Algorithm since root node,
CART decision tree is established with the recursive mode of data set, specific step is as follows：
1) Gini Gain of the characteristic value to current data set D of the existing each feature of present node is calculated.It is calculating
Characteristic value Gini Gain in, select Gini Gain minimum value corresponding feature A and characteristic value a to divide as optimal item
Knuckle point and bifurcated value.
2) according to the difference of feature corresponding eigenvalue and best eigenvalue a size, data set D is divided into D1 and D2 two
Divided data establishes the left and right node of present node, and enabling the data set of left sibling is D1, and the data set of right node is D2.
3) enter child node, judge whether it is leaf node, if then stopping recurrence, and decision subtree is returned to, into step
Rapid 4；Otherwise continue recurrence, execute step 1-3.
4) judge whether that the case where all nodes have all formed leaf node or all features has all judged to finish, it can not
Continue bifurcated, recursion cycle terminates and generates final decision-tree model.
Data overfitting is avoided the problem that using predictive pruning mode in Decision Tree Construction, is limited according to some principles
Threshold value in training, the growth for stopping decision tree early prevent over-fitting, such as sample in the depth capacity of setting decision tree, node
This minimum etc..The present invention takes python to call categorised decision tree tree.DecisionTreeCla in the library scikit-learn
The method of ssifier () function realizes CART decision Tree algorithms.When training, by calling the fit method of function to be fitted
And beta pruning, establish corresponding CART decision-tree model.In application, to the predict method of the model calling classification device function, just
The classification results under input feature value can be calculated, are realized to the output forecast function under special characteristic.
Wherein, it is selected in training method comprising multiple parameters, to meet different training requirements.Selection in the present invention is such as
Under：Feature selecting standard parameter criterion is Gini (default Gini coefficient), and feature division points selection criteria splitter is
Best (find global optimum), the maximum characteristic max_features that when division considers is None (considering all features), number
In advance whether according to, sequence presort is False (not sorting), decision tree depth capacity max_depth and maximum leaf segment points max_
Leaf_nodes is not defined (with no restrictions).Further, it is contemplated that beta pruning situation, the classifier take predictive pruning method to decision tree into
Row optimization.Need to define the value of two parameters, be the minimum sample number min_samples_leaf (being defaulted as 1) of leaf node respectively with
And the smallest sample weights of leaf node and min_weight_fraction_leaf (being defaulted as 0).
In practical applications, it is necessary first to target production line related data information is acquired, as training dataset.It is specific next
It says, the training dataset of the CART decision tree scheduling model can be described as：
D={ (Xi,Yi)|Xi∈Rm,Yi∈Rn, i=1,2 ..., N } (4)
Wherein, Xi=(xi,1,xi,2,...,xi,m)TIt is input vector, Yi=(yi,1,yi,2,...,yi,n)TCorrespond to Xi
Output vector.Input vector XiIt is the production line state indicated by associated production feature set, output vector YiExpression is being coped with
Under emergency event disturbance, current manufacturing lines state XiCorresponding optimal scheduling scheme.
Since semiconductor production line process is complicated, the data class that can be used as production feature is numerous, if all data are made
Being characterized collection, then sample input dimension is excessive, is easy to cause the decision tree speed of growth slow.Therefore, it is necessary to combine actual demand feelings
Condition selects suitable production feature set.Scheduling scheme is then the foundation for instructing process of manufacture.It is disturbed towards emergency event
Optimizing scheduling research in, select to be relatively easy to the heuristic rule that dynamic changes and adjusts is application, production line scheduling side
Case is the combination of scheduling rule applied by each production line processing district.
Assuming that under the burst disturbance of occurrence of equipment failure, considering CPS environment optimizing scheduling in semiconductor production line
Real-time, performance under short-term load index are obtained compared with conducive to quick obtaining, select production line state XiBy 3 production character representations, respectively
For：System is in product number WIP, each processing district buffer length Buffer_len, faulty equipment Fail_ID.Current optimal tune
Degree scheme YiThe heuristic rule as applied by 3 production line processing districts is composed.Input vector can be expressed as Xi=
(xi,1,xi,2,xi,3,xi,4,xi,5)T, xi,jIndicate j-th of production feature；Corresponding optimal scheduling scheme is Yi=(yI, 1,yI, 2,
yI, 3)T, yi,jIndicate the scheduling rule selection of j-th of processing district.
Then, the optimal sample data { ((x of N item is obtained from the semiconductor production linei,1,xi,2,xi,3,xi,4,xi,5),(yi,1,
yi,2,yi,3)) | i=1,2 ..., N }, production line sample data set D is formed, using previously mentioned categorised decision tree
DecisionTreeClassifier implementation method, by obtaining CART decision-tree model after training, being embedded in scheduler module and making
For a kind of production rescheduling method.
Finally, obtaining current production feature set X when equipment fault occursi', new dispatching party is calculated in real time
Case Yi', as obtained Optimized Operation scheme.
So far, the design and realization of semiconductor production line CPS environment dispatching optimization method are just completed.
Embodiment
The present embodiment realizes the calculating logic of designed dispatching method using python, and carries out forms circle using C# language
Face programming, calls designed computation model, scheduler module program and CPS environment is integrated, and realizes in semiconductor production mistake
Real-tim scheduling and control in journey.
The present embodiment is realized in a certain wisdom manufacture exemplary cell based on semiconductor Minifab production line.This implementation
The dispatching device of example is realized by technologies such as PROFINET, SafetyBridge, RFID to semiconductor MiniFab production line
The simulation of actual production process.The exemplary cell has a simulation lithographic equipment, two analog spread equipment and two moulds
Quasi- ion implantation device.In addition, the unit further includes a simulation warehouse compartment, it is made of raw material area and finished product area, is responsible for storage
Caching semi-finished product to be processed is responsible in raw material workpiece and finished work-piece and a simulation material Accreditation Waiting Area.The unit
The KUKA robot KR6R900 in center is responsible for carrying workpiece between simulation warehouse compartment, buffer area, process equipment.Fig. 5 is the unit
Layout.
First by Unity3D and OPC+MQTT the relevant technologies, the CPS for semiconductor wisdom manufacture exemplary cell is realized
Environment construction, the online display platform interface of the production line finally established are as shown in Figure 6.
The online display platform interface of production line shows hidden including addition scheduling number, frame, opens schedule file, mould is dispatched in starting
The order buttons such as block.
Addition scheduling number (i.e. scheduling number write-in part) can add the dispatching mechanical hand number of production process manually, main to use
It is used in off-line state test model and emulator processing action.
Frame shows the hidden display that can be switched fast robot work platform peripheral frame and hides, and sees convenient for user
It examines.
The scheduling list of schedule file then available existing simulation model processing flow is opened, is realized to emulation mould
The displaying of type production procedure.
Starting scheduler module is then used to open the optimizing scheduling module that design is completed, when equipment fault, to life
Producing line carries out production readjustment degree, and assigns new scheduling scheme.
After starting semiconductor wisdom manufacture exemplary cell, while the online display platform that operation is established, verify its function
Realization situation：
1) platform realizes the displaying of the production procedure under two kinds of scenes：It on the one hand can be according to the operation of current manufacturing lines
State, real-time exhibition production procedure.On the other hand, by reading schedule file, the scheduling list of simulation model is loaded into, exhibition
Show ideal production and processing process.
2) three-dimensional display platform real-time perception semiconductor production line state, and when it equipment fault occurs, show failure
Situation prompts administrative staff to carry out the optimizing scheduling of production line.
3) the comprehensive observation to production line can be realized by the operation of mouse and keyboard in platform.
4) phenomena such as during system overall operation, model movement is more smooth, fails without Caton, deadlock, crawl
Generation.
Under the semiconductor production line CPS environment constructed, realization and the integrating process for being scheduled optimization method are as follows：
Using C# language, the forms program of exploitation optimizing scheduling module on 2017 platform of Visual Studio, and with
Unity3D model integrated can call opening in system operation at any time.Its main interface is as shown in Figure 7.
Wherein, scheduler module is communicated by OPC agreement with actual production line, obtains required all kinds of production lines letter
Breath.The scheduling scheme of current manufacturing lines is illustrated in interface, and needs the new scheduling scheme applied to production line.Click write-in
Data can will reach the PLC controller of production line under the new rule for generating or being manually entered, and then change current scheduling side
Case.
The data button CART decision-tree model file tree.pkl that be then Selection and call designed is generated, OPC is passed through
Interface obtains the characteristic value of current desired production line state, and with the incoming decision write using python language of array form
In tree prediction perform script, operation obtains the new scheduling rule combination of decision-tree model generation, shows in new scheduling rule
In text box.Scheduling decision personnel can refer to obtained new scheduling scheme, be compared with former scheme, and decide whether to select
It selects and uses the new departure.
Production line characteristic value when generating new scheduling scheme obtains and new departure is assigned, as shown in Figure 8,9.
In same production plan, under three kinds of different scenes production line performance indicator and job order compare point
Analysis, verifies its optimizing scheduling effect.
Three kinds of scenes are respectively：Production line do not occur disturbance, disturb after be not scheduled optimization, disturb after
It is scheduled optimization.In semiconductor wisdom manufacture exemplary cell, selecting equipment fault is sudden disturbance event, will be given birth to by hand
It produces equipment and is set as simulated failure state, it is made to stop working.Then in three-dimensional display model perception failure and prompt readjustment degree
Afterwards, new scheduling scheme is generated by integrated optimizing scheduling module, and is applied to exemplary cell.After production, remember respectively
Record all kinds of production information data under three kinds of scenes.
In an experiment, the order product number for selecting production and processing is 50, includes 2 kinds of products (product A, product B) each 25
A, every 5 one group feeds intake in turn, takes fixed WIP number feeding method (WIP<11).Include in the scheduling rule set of selection
4 kinds of heuristic rules, respectively：FIFO (first in, first out), CR (critical value), SRPT (most short remaining process time), FSVCT
(manufacturing cycle variance smoothing fluctuations).The scheduling as corresponding to 3 production areas (Mab, Mcd, Me) of the scheduling scheme of production line
Rule combination is formed.The production performance index recorded in experiment is that product after processing is completed is averaged process-cycle (AvgCT), most
Greatly/minimum process period (MaxCT/MinCT) and process-cycle standard deviation (CT_sd) and each equipment total processing times.Together
When, obtain the corresponding production and processing job order of record.
In order to guarantee the science and accuracy of experimental verification, two groups of experiment sides under the different conditions of production are separately designed
Case, as shown in table 1.
The experimental designs twice of table 1
According to experiment, available semiconductor production line is three kinds of scenes (non-failure, failure be unscheduled, failure and dispatch)
Under production performance index as shown in table 2 and 3.
Production performance under 2 three kinds of scenes of table compares (scheme one)
Production performance under 3 three kinds of scenes of table compares (scheme two)
By taking scheme two as an example, production performance comparison result is as shown in Figure 10.It can be seen that from the comparison of two groups of data
Although compared to the ideal scene of non-occurrence of equipment failure, the process-cycle related performance indicators of other two kinds of scenes have one
Determine the increase of degree.Illustrate that workpiece to be processed caused by equipment fault is accumulated, has to the process-cycle of product of production line not small
Influence, and increase its degree of fluctuation.However, carrying out the scene of optimizing scheduling after failure occurs, obtained items are raw
Index is produced still better than the production process not being scheduled.Therefore, it was demonstrated that the method for optimizing scheduling is after improve equipment fault
On the problem of process-cycle related performance indicators decline, have the effect of certain.
In addition, Choice two, continues each equipment processing times of system that analysis collects, such as table 4 and Figure 11 institute
Show.
The total processing times of equipment under 4 three kinds of scenes of table compare (scheme two)
It can be found that, in the case where Md equipment breaks down, Mc equipment alternatively always adds after the completion of production
Work number obviously rises, and this is obvious.Simultaneously as the generation of equipment fault, a large amount of workpiece accumulations are but also Mb
The perfect condition that processing times do not break down is substantially increased, and has been even more than Ma.In a short time, the sharply increasing of Mb load
It is high, it is likely that cause equipment that " avalanche effect " occurs in succession and new fault point occurs.Compared to the feelings for not being scheduled optimization
Scape, after having used method for optimizing scheduling, total processing times gap of Ma and Mb equipment room tend to perfect condition, Jin Erwei again
Production system overall stability is held.
From another angle it can also be seen that influence situation of the optimizing scheduling to production line.Choice one collects three kinds
Absolute time therein is converted to relative time on the basis of the time that respective production starts by the production job order under scene,
Then compare.It can be found that after production starts, three kinds of scenes send work information essentially identical, until after failure occurs
Dispatching point when, occur respectively it is different send work to select, and then form different production decisions.
Therefore, a period of time in corresponding job order before and after dispatching point is intercepted, the Gantt chart of production and processing is drawn, compares not
Two kinds of scenes of scheduling and scheduling, as shown in figure 12.
Wherein, different workpieces (Lot) is distinguished using different colours, square indicates its process time, and the number on square is
Lot ID has chosen the Gantt chart that the total 6min in front and back or so the time occurs for failure.There it can be seen that occurring in equipment fault
Afterwards, the new scheduling scheme of production line application.Within next one section of production time, compared with unscheduled production procedure,
The workpieces processing distribution of production line bottleneck device Me is more uniform after scheduling, and within identical a period of time, has more 1 workpiece
Process, utilization rate is improved；Meanwhile workpiece sends the linking of work arrangement more compact in production line, no longer in a jumble
No chapter (such as the Mcd and the processing district Me 13-19 workpiece after dispatching).Therefore, reduce the equipment idle waiting time, and then optimize
Whole production system efficiency.
To sum up, passing through the excavation and analysis to experiment the data obtained, it was demonstrated that designed optimizing scheduling of the invention
Method, happening suddenly caused by coping with system equipment failure has validity when disturbing, and can improve production line performance indicator, and
And system stability is improved to a certain extent.
The preferred embodiment of the present invention has been described in detail above.It should be appreciated that those skilled in the art without
It needs creative work according to the present invention can conceive and makes many modifications and variations.Therefore, all technologies in the art
Personnel are available by logical analysis, reasoning, or a limited experiment on the basis of existing technology under this invention's idea
Technical solution, all should be within the scope of protection determined by the claims.
Claims (10)
1. the method for optimizing scheduling under a kind of semiconductor production line CPS environment, which is characterized in that this method is specially：
The semiconductor production line CPS environment towards scheduling is constructed, perception semiconductor production line in process of production disturb by all kinds of bursts
It is dynamic, when burst disturbance occurs, readjustment degree is carried out based on production line multidate information, is adjusted in due course according to production line changing condition
Degree scheme.
2. the optimizing scheduling device under a kind of semiconductor production line CPS environment for realizing method as described in claim 1, feature
It is, including：
Communication module, for realizing the communication with actual production line；
CPS environment construction module, for establishing digital three-dimemsional model corresponding with actual target production line；
Optimizing scheduling designs module, is used for real-time perception production information, triggers readjustment degree in production information variation, and according to life
Producing line multidate information generates new production line scheduling scheme, realizes readjustment degree.
3. the optimizing scheduling device under semiconductor production line CPS environment according to claim 2, which is characterized in that described
Communication module is communicated by OPC technology with realization between the data communication network that MQTT communication protocol combines and actual production line.
4. the optimizing scheduling device under semiconductor production line CPS environment according to claim 2, which is characterized in that described
CPS environment construction module includes：
Information acquisition unit, for acquiring target production line information by the communication module；
Three-dimensional modeling unit, for establishing the corresponding digitlization of actual target production line according to the target production line information
Threedimensional model；
Movements design unit, for generating the dynamic action parameter of each component in the digital three-dimemsional model.
5. the optimizing scheduling device under semiconductor production line CPS environment according to claim 4, which is characterized in that described
In three-dimensional modeling unit, the digital three-dimemsional model is resettled after pre-processing to the target production line information；
The pretreatment includes deleting the face that can not be observed in process of production, merge disconnection vertex and removing isolated vertex.
6. the optimizing scheduling device under semiconductor production line CPS environment according to claim 4, which is characterized in that described
CPS environment construction module further includes：
Online display module, for showing production line machining state.
7. the optimizing scheduling device under semiconductor production line CPS environment according to claim 2, which is characterized in that described
Optimizing scheduling designs module：
Input layer, for obtaining production line multidate information by the communication module；
Readjustment degree decision-making level occurs burst disturbance for judging whether according to the production line multidate information, and is happening suddenly
Production readjustment degree trigger signal is generated when disturbance；
Weight dispatch layer, responds the production readjustment degree trigger signal, for using readjustment according to the production line multidate information
Degree scheme generation method generates new production line scheduling scheme.
8. the optimizing scheduling device under semiconductor production line CPS environment according to claim 7, which is characterized in that described
Readjustment degree scheme generation method is CART decision Tree algorithms.
9. the optimizing scheduling device under semiconductor production line CPS environment according to claim 8, which is characterized in that described
The production line feature used in CART decision Tree algorithms includes system in product number, each processing district buffer length and faulty equipment
Number.
10. the optimizing scheduling device under semiconductor production line CPS environment according to claim 7, which is characterized in that described
Optimizing scheduling designs module：
Layer is evaluated, for carrying out compliance test result to the new production line scheduling scheme based on digital three-dimemsional model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810548729.0A CN108829052A (en) | 2018-05-31 | 2018-05-31 | Method for optimizing scheduling and device under a kind of semiconductor production line CPS environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810548729.0A CN108829052A (en) | 2018-05-31 | 2018-05-31 | Method for optimizing scheduling and device under a kind of semiconductor production line CPS environment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108829052A true CN108829052A (en) | 2018-11-16 |
Family
ID=64146991
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810548729.0A Pending CN108829052A (en) | 2018-05-31 | 2018-05-31 | Method for optimizing scheduling and device under a kind of semiconductor production line CPS environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108829052A (en) |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109361694A (en) * | 2018-11-22 | 2019-02-19 | 北京未尔锐创科技有限公司 | A kind of network communication method |
| CN109685398A (en) * | 2019-02-02 | 2019-04-26 | 深圳埃克斯工业自动化有限公司 | A kind of failure response method and system of semiconductor process module |
| CN109784570A (en) * | 2019-01-23 | 2019-05-21 | 西北工业大学 | Wisdom workshop flexible production dispatching method based on information physical emerging system |
| CN110389534A (en) * | 2019-07-01 | 2019-10-29 | 东华大学 | A kind of Multi-axis motion control virtual experimental system based on the twin technology of number |
| CN110987865A (en) * | 2019-12-13 | 2020-04-10 | 齐鲁工业大学 | Method for detecting fig quality based on near infrared spectrum |
| CN111077803A (en) * | 2019-12-24 | 2020-04-28 | 国电南瑞科技股份有限公司 | Panoramic information physical fusion system simulation framework and working method thereof |
| CN111356988A (en) * | 2020-02-21 | 2020-06-30 | 长江存储科技有限责任公司 | Method and system for scheduling semiconductor processing |
| CN111553664A (en) * | 2020-05-11 | 2020-08-18 | 重庆金美通信有限责任公司 | Method for realizing intelligent management of design and production of communication equipment based on 5G technology |
| CN111882151A (en) * | 2020-06-16 | 2020-11-03 | 杭州未名信科科技有限公司 | Production scheduling method and system for discrete manufacturing industry based on reinforcement learning |
| CN112230596A (en) * | 2020-10-14 | 2021-01-15 | 衡阳市华毅自动化科技有限公司 | Job scheduling system and scheduling method of automatic industrial equipment |
| CN112650179A (en) * | 2020-12-23 | 2021-04-13 | 同济大学 | Dynamic scheduling method of semiconductor manufacturing system |
| CN113031543A (en) * | 2021-02-24 | 2021-06-25 | 同济大学 | Control scheduling method and device for semiconductor production line |
| CN113050553A (en) * | 2021-02-18 | 2021-06-29 | 同济大学 | Scheduling modeling method of semiconductor production line based on federal learning mechanism |
| CN113064388A (en) * | 2021-02-24 | 2021-07-02 | 同济大学 | Scheduling optimization method and device for semiconductor production line |
| CN113341902A (en) * | 2021-06-25 | 2021-09-03 | 中国科学院重庆绿色智能技术研究院 | Design method and system for balance production line |
| CN113888921A (en) * | 2021-10-21 | 2022-01-04 | 安徽工业大学 | Belt conveyor simulation teaching device based on Unity3D |
| CN114285869A (en) * | 2021-11-10 | 2022-04-05 | 贵阳铝镁设计研究院有限公司 | Perception and prediction method for process parameters of electrolytic aluminum equipment |
| CN114815759A (en) * | 2022-06-27 | 2022-07-29 | 广州力控元海信息科技有限公司 | Virtual-real fusion flexible production line variable control method and system |
| CN115390529A (en) * | 2022-08-31 | 2022-11-25 | 同济大学 | Simulation intelligent device based on unit-level CPS |
| CN115577576A (en) * | 2022-12-08 | 2023-01-06 | 中国电子工程设计院有限公司 | Dynamic virtual wire-grouping simulation system and method for semiconductor factory |
| US12055914B2 (en) | 2020-10-30 | 2024-08-06 | Beijing Zhongxiangying Technology Co., Ltd. | Data processing method, device and system, and electronic device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103439886A (en) * | 2013-07-26 | 2013-12-11 | 同济大学 | Semiconductor production line self-adaptation dynamic dispatching device |
| CN105843189A (en) * | 2016-04-09 | 2016-08-10 | 北京化工大学 | Simplified simulation model based high efficient scheduling rule choosing method for use in semiconductor production lines |
-
2018
- 2018-05-31 CN CN201810548729.0A patent/CN108829052A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103439886A (en) * | 2013-07-26 | 2013-12-11 | 同济大学 | Semiconductor production line self-adaptation dynamic dispatching device |
| CN105843189A (en) * | 2016-04-09 | 2016-08-10 | 北京化工大学 | Simplified simulation model based high efficient scheduling rule choosing method for use in semiconductor production lines |
Non-Patent Citations (7)
| Title |
|---|
| WEIXIN_30808693: "Unit模型要求标准", 《HTTPS://BLOG.CSDN.NET/WEIXIN_30808693/ARTICLE/DETAILS/95786998》 * |
| 曹政才: "半导体生产线建模方法与调度策略研究进展", 《计算机集成制造系统》 * |
| 曹秋实: "半导体生产的CPS构建及调度可行性分析", 《新型工业化》 * |
| 李莉: "半导体制造重调度研究", 《中国机械工程》 * |
| 李莉: "半导体生产线全局修正式重调度方法研究", 《计算机集成制造系统》 * |
| 王娜: "多产品可重入生产系统性能的影响因素探究与性能评估", 《中国优秀硕士学位论文全文数据库经济与管理科学辑》 * |
| 马玉敏: "基于支持向量机的半导体生产线动态调度方法", 《计算机集成制造系统》 * |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109361694A (en) * | 2018-11-22 | 2019-02-19 | 北京未尔锐创科技有限公司 | A kind of network communication method |
| CN109361694B (en) * | 2018-11-22 | 2021-11-30 | 北京未尔锐创科技有限公司 | Network communication method |
| CN109784570A (en) * | 2019-01-23 | 2019-05-21 | 西北工业大学 | Wisdom workshop flexible production dispatching method based on information physical emerging system |
| CN109784570B (en) * | 2019-01-23 | 2024-01-12 | 西北工业大学 | Intelligent workshop flexible production scheduling method based on information physical fusion system |
| CN109685398A (en) * | 2019-02-02 | 2019-04-26 | 深圳埃克斯工业自动化有限公司 | A kind of failure response method and system of semiconductor process module |
| CN110389534A (en) * | 2019-07-01 | 2019-10-29 | 东华大学 | A kind of Multi-axis motion control virtual experimental system based on the twin technology of number |
| CN110987865A (en) * | 2019-12-13 | 2020-04-10 | 齐鲁工业大学 | Method for detecting fig quality based on near infrared spectrum |
| CN111077803A (en) * | 2019-12-24 | 2020-04-28 | 国电南瑞科技股份有限公司 | Panoramic information physical fusion system simulation framework and working method thereof |
| CN111356988A (en) * | 2020-02-21 | 2020-06-30 | 长江存储科技有限责任公司 | Method and system for scheduling semiconductor processing |
| CN111356988B (en) * | 2020-02-21 | 2021-03-12 | 长江存储科技有限责任公司 | Method and system for scheduling semiconductor processing |
| CN111553664A (en) * | 2020-05-11 | 2020-08-18 | 重庆金美通信有限责任公司 | Method for realizing intelligent management of design and production of communication equipment based on 5G technology |
| CN111553664B (en) * | 2020-05-11 | 2022-10-14 | 重庆金美通信有限责任公司 | System for realizing intelligent management of communication equipment design and production based on 5G technology |
| CN111882151A (en) * | 2020-06-16 | 2020-11-03 | 杭州未名信科科技有限公司 | Production scheduling method and system for discrete manufacturing industry based on reinforcement learning |
| CN112230596A (en) * | 2020-10-14 | 2021-01-15 | 衡阳市华毅自动化科技有限公司 | Job scheduling system and scheduling method of automatic industrial equipment |
| US12055914B2 (en) | 2020-10-30 | 2024-08-06 | Beijing Zhongxiangying Technology Co., Ltd. | Data processing method, device and system, and electronic device |
| CN112650179A (en) * | 2020-12-23 | 2021-04-13 | 同济大学 | Dynamic scheduling method of semiconductor manufacturing system |
| CN112650179B (en) * | 2020-12-23 | 2022-05-27 | 同济大学 | Dynamic scheduling method of semiconductor manufacturing system |
| CN113050553A (en) * | 2021-02-18 | 2021-06-29 | 同济大学 | Scheduling modeling method of semiconductor production line based on federal learning mechanism |
| CN113064388A (en) * | 2021-02-24 | 2021-07-02 | 同济大学 | Scheduling optimization method and device for semiconductor production line |
| CN113064388B (en) * | 2021-02-24 | 2022-05-03 | 同济大学 | Scheduling optimization method and device for semiconductor production line |
| CN113031543A (en) * | 2021-02-24 | 2021-06-25 | 同济大学 | Control scheduling method and device for semiconductor production line |
| CN113341902A (en) * | 2021-06-25 | 2021-09-03 | 中国科学院重庆绿色智能技术研究院 | Design method and system for balance production line |
| CN113888921A (en) * | 2021-10-21 | 2022-01-04 | 安徽工业大学 | Belt conveyor simulation teaching device based on Unity3D |
| CN114285869A (en) * | 2021-11-10 | 2022-04-05 | 贵阳铝镁设计研究院有限公司 | Perception and prediction method for process parameters of electrolytic aluminum equipment |
| CN114815759A (en) * | 2022-06-27 | 2022-07-29 | 广州力控元海信息科技有限公司 | Virtual-real fusion flexible production line variable control method and system |
| CN114815759B (en) * | 2022-06-27 | 2022-09-20 | 广州力控元海信息科技有限公司 | Virtual-real fusion flexible production line variable control method and system |
| CN115390529A (en) * | 2022-08-31 | 2022-11-25 | 同济大学 | Simulation intelligent device based on unit-level CPS |
| CN115577576A (en) * | 2022-12-08 | 2023-01-06 | 中国电子工程设计院有限公司 | Dynamic virtual wire-grouping simulation system and method for semiconductor factory |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108829052A (en) | Method for optimizing scheduling and device under a kind of semiconductor production line CPS environment | |
| KR102184182B1 (en) | Project/Task Intelligent Goal Management Method and Platform based on Super Tree | |
| CN113033001B (en) | Three-dimensional digital factory modeling method and system for digital twin application | |
| CN100549879C (en) | Hybrid user interface with basic presentation information of the outstanding side information of belt variable | |
| US9137050B2 (en) | Demand response system incorporating a graphical processing unit | |
| CN102402716B (en) | Intelligent production decision support system | |
| US10416614B2 (en) | Human programming interfaces for machine-human interfaces | |
| CN113176948A (en) | Edge gateway, edge computing system and configuration method thereof | |
| CN112148926A (en) | Graph data stream processing method, processing device and storage medium | |
| CN109492059A (en) | A kind of multi-source heterogeneous data fusion and Modifying model process management and control method | |
| KR20210147594A (en) | Edge computing devices and methods that provide optimization for energy data collection and management | |
| CN109839910A (en) | Control device, control method and computer readable storage medium | |
| WO2024066683A1 (en) | Industrial internet operating system and product processing method | |
| CN113179190A (en) | Edge controller, edge computing system and configuration method thereof | |
| CN110705891A (en) | Data processing method based on high-allocable changeability | |
| KR20240140038A (en) | Device for providing remote collaboration service based on digital twin and method for controlling the same | |
| CN116796504A (en) | Production line simulation method based on digital twin and related equipment thereof | |
| CN116643877A (en) | Computing power resource scheduling method, training method and system of computing power resource scheduling model | |
| CN117608865B (en) | Mathematical model service method and system of take-away meal delivery platform based on cloud computing | |
| CN114819095A (en) | Method and device for generating business data processing model and electronic equipment | |
| Sun et al. | Digital twin for energy-efficient integrated process planning and scheduling | |
| Rui et al. | Context-based intelligent scheduling and knowledge push algorithms for ar-assist communication network maintenance | |
| CN115755775A (en) | Characteristic tool path dynamic generation system and method based on CAM cloud service architecture | |
| CN116012109A (en) | Order generation method and custom production method based on meta universe | |
| CN115587903A (en) | Assembly line digital management and control resource pushing method and device |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181116 |