CN107831740B - It is a kind of applied to notebook part it is distributed manufacture during Optimization Scheduling - Google Patents
It is a kind of applied to notebook part it is distributed manufacture during Optimization Scheduling Download PDFInfo
- Publication number
- CN107831740B CN107831740B CN201710961541.4A CN201710961541A CN107831740B CN 107831740 B CN107831740 B CN 107831740B CN 201710961541 A CN201710961541 A CN 201710961541A CN 107831740 B CN107831740 B CN 107831740B
- Authority
- CN
- China
- Prior art keywords
- factory
- notebook
- machine
- indicate
- individual
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 29
- 238000005457 optimization Methods 0.000 title claims abstract description 28
- 238000000034 method Methods 0.000 claims abstract description 38
- 239000000872 buffer Substances 0.000 claims abstract description 23
- 238000009826 distribution Methods 0.000 claims abstract description 22
- 238000006073 displacement reaction Methods 0.000 claims abstract description 10
- 238000013507 mapping Methods 0.000 claims abstract description 7
- 238000003754 machining Methods 0.000 claims abstract description 6
- 238000012545 processing Methods 0.000 claims description 14
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000005070 sampling Methods 0.000 claims description 8
- 238000012163 sequencing technique Methods 0.000 claims description 6
- 235000013399 edible fruits Nutrition 0.000 claims 1
- 238000011161 development Methods 0.000 description 4
- 239000004744 fabric Substances 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 101100517651 Caenorhabditis elegans num-1 gene Proteins 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 230000002028 premature Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005728 strengthening Methods 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)
Abstract
The present invention relates to the Optimization Schedulings during a kind of distributed manufacturing applied to notebook part, method are as follows: optimization aim is optimized by determining distributed displacement pipeline schedule model and optimization aim with limited buffer, and using the optimization method of effective Estimation of Distribution Algorithm;Scheduling model therein is established according to each part machining the time on each machine, target is to minimize its total completion date: for the present invention in the distributed production scheduling problems for proposing a kind of effective Estimation of Distribution Algorithm and being used to solve notebook part, this is to be used for the intelligent algorithm based on EDA to solve problems for the first time;It is put forward for the first time according to factory allocation rule ECF and completes anti-factory's mapping ruler MECF earliest, overcome to the problem of part distributed intelligence is disturbed after optimum individual progress local search;Using the local search based on Swap neighborhood and based on Inverse neighborhood, the local search ability of algorithm is further strengthened.
Description
Technical field
The present invention relates to the Optimization Schedulings during a kind of distributed manufacturing applied to notebook part, belong to
In displacement flow shop intelligent optimization dispatching technique field of the distribution with limited buffer.
Background technique
The distributed manufacturing for the notebook part that the present invention studies is the displacement about distribution with limited buffer
The production line of assembly line has pushed computer industry with the rapid development of society with the continuous improvement of people's lives level
It continues to develop, notebook is light and handy as a kind of i.e. convenience, and the representative of fashion, there is very big development potentiality and demand.With
The development of globalization, the purchase between the joint and enterprise of intercompany is more and more common, the production of notebook part be no longer by
Single factory completes, and each part in notebook can be assigned in different factories according to constraint condition and produce, in this way can
The resource of multiple enterprises or factory is adequately used, realizes the reasonable distribution and optimization of resource, it is fast with minimum cost
Speed is completed to manufacture.Especially in increasingly competitive today, the benefit and core of factory is can be improved in distributive knowledge network
Competitiveness, especially as the development of just-in-time (just-in-time) manufacture and billboard control, the distribution with limited buffer
Manufacture the scheduling problem model encountered in closer actual production process.
Notebook part possesses following spy in displacement flow line production manufacturing process of the distribution with limited buffer
Point: first, the dispersibility of notebook part production makes production system increasingly complex, in order to allow co-ordination between each factory,
Need the allocation plan of a reasonable feasibility, second, distribution produces, pen more complicated relative to traditional single plant produced
Remember distribution of this part between factory and the arrangement in inside plants production sequence, these factors intercouple, Yao Shixian
The effect of global optimization, it is necessary to the factor collaboration optimization that these are intercoupled.Third, distribution, which is manufactured, needs a conjunction
The allocation rule of reason makes the produce load of each factory relatively uniform.The following steps are included: first in the production, in allotment
The heart is first assigned to each part of notebook in each factory by allocation rule, second, and each factory obtains to be produced zero
After part, produced according to the displacement pipeline mode with limited buffer.The process is typically with limited buffer
Distribution displacement fluvial incision (Distributed Permutation Flow Shop Scheduling Problem
with Limited Buffers,DPFSSP_LB).DPFSSP_LB belongs to NP-hard problem, most notebook part
Distributed pipeline produce problem can reduction be DPFSSP_LB, therefore have to the research of DPFSSP_LB derivation algorithm very strong
Engineering background and learning value, guidance and help can be provided for the design of associated production optimization system.
Since DPFSSP_LB is NP hardly possible, so that traditional combined optimization method can only solve small-scale problem, and inspire
The effect of formula optimization method is not again fine.Therefore, the present invention designs a kind of Optimized Operation based on effective Estimation of Distribution Algorithm
Scheme can obtain the excellent solution of displacement fluvial incision of the notebook part distribution with limited buffers within a short period of time
Certainly scheme.
Summary of the invention
The technical problem to be solved by the present invention is to obtain the distributed production scheduling of notebook part within a short period of time
The excellent fix of problem, provide it is a kind of applied to notebook part it is distributed manufacture during Optimized Operation
Method.
The technical scheme is that the optimization tune during a kind of distributed manufacturing applied to notebook part
Degree method, by determining distributed displacement pipeline schedule model and optimization aim with limited buffer, and using effective point
The optimization method of cloth algorithm for estimating optimizes optimization aim;Scheduling model therein is according to each part in each machine
On machining the time establish, target is by its total completion date Cmax(π) is minimized:
K=1,2, F, i=1,2, nk, j=1,2, m
Wherein, F indicates F isomorphism factory, and k indicates some specific factory, and n indicates that notebook part to be processed is total
Number, nkIndicate that notebook part sum to be processed, i indicate i-th bit in notebook Job Scheduling to be processed in the factory in factory k
It sets, j indicates that a certain machine in factory, m indicate number of machines total in each factory;π=π (1), π (2), π (n) }
Indicate total sequence when notebook part is not allocated to factory, π (n) indicates the notebook in total sequence in nth position
Part, i.e. notebook part on the last one position, the notebook part π that always sorts are generated respectively after factory's allocation rule
Notebook part Machining Sequencing π in factoryk, πk(i) it indicates in factory k, in Job Scheduling to be processed zero on i-th of position
Part, πk={ πk(1),πk(2),···,πk(nk) indicate factory k in each part of notebook sequence;It indicates in work
Part π in factory kk(i) operation on machine j, in factory k, notebook part πk(i) will pass throughOperation could be processed completion, whereinIndicate the notebook part in factory k
πk(i) operation on machine m, for part after being assigned to some factory, all operations of the part all must be in the work
It is completed the process in factory;Indicate the part π in factory kk(i) start the processing moment on machine j,It indicates
Part π in factory kk(i) process time on machine j, and it is greater than 0,Indicate the part π in factory kk(i) In
The moment is completed the process on machine j,Indicate part π in factory kk(nk) on machine m the moment is completed the process,
It indicates in the buffer area in factory k between machine j-1 and j;In addition, arrange all parts be it is independent, at 0 moment, allow appoint
One part is processed, and on every machine in the factory, the processing sequence of part is not changing after determining, and in the buffer zero
Part obeys first in, first out rule;The same part can only be processed on a machine at the same moment;Synchronization, one
A part can only be processed on machine, some operation of part does not allow midway to stop;Process timeThe buffer area and
It is known that the traveling time between the setting time and operation of machine is ignored.
Specific step is as follows for the optimization method of effective Estimation of Distribution Algorithm:
The coding and decoding mode of Step1, solution: it is encoded according to the sequencing that part is processed, is contained in algorithm population
There are multiple individuals, a solution either part of each individual correspondence problem always sorts;Decoding is exactly that the coding and sorting order is passed through
Factory's allocation rule generation one is feasible to carry into execution a plan;
Step2, probabilistic model initialization: the probability point in algorithm gen generation is indicated using the matrix P (g) of n × n dimension
Cloth model, algorithm are set as in initial phase, each element set in probabilistic model1/n, specifically it is expressed as follows:
Wherein, Pn,1(g) indicate g for when notebook part 1 in always sequence π nth position or before the probability that occurs
Size, Pn,n(g) from the processing priority for numerically indicating notebook difference part;
Step3, sampling generate new population: setting population scale is popsize, and it is competing to carry out championship to probabilistic model P (g)
It strives sampling and generates g for the processing sequence of notebook part to be processed in individual each in new population, enable g=g+1;
Step4, local search: after sampling generates new population, Swap and Inverse are successively implemented to optimum individual in population
Two kinds of local searches, the individual after search, factory's sequence generate the total sequence of individual by anti-factory's mapping ruler, after search
If body individual than before is more excellent, otherwise the individual before replacement is not replaced;
The update of Step5, probabilistic model: T excellent individual is picked out in the population after local search, and to outstanding
Part position in body is recorded accordingly, is generated the probability matrix P (g+1) in g+1 generation, is updated using following formula
Probabilistic model:
Wherein, Py,z(g) indicate g for when notebook part z in always sequence π y-th of position or before the probability that occurs
Size, Py,z(g+1) indicate g+1 for when notebook part z in always sequence π y-th of position or before the probability that occurs it is big
Small, α is the learning rate of P (g) and its value between 0 to 1, and T is excellent individual number,It is defined as follows:
Step6, termination condition: the maximum operation algebra of set algorithm is Max_gen, if current algebra gen is less than most
Big operation algebra Max_gen, goes to Step3, iterates, and when meeting termination condition, exports optimal solution.
Anti- factory's mapping ruler sorts factory πkMap back total sequence π.
The population scale popsize=100, learning rate α=0.01, the maximum value buffer_size=of buffer area
3, the excellent individual number T=10 that probability matrix selects when updating, the maximum algebra Max_gen=1000 of algorithm operation.
The beneficial effects of the present invention are:
1, a kind of effective Estimation of Distribution Algorithm (EEDA) is algorithmically proposed, for solving the distribution of notebook part
Production scheduling problems, this is to be used for the intelligent algorithm based on EDA to solve problems for the first time.
2, the present invention is put forward for the first time according to factory allocation rule ECF completes anti-factory's mapping ruler MECF earliest, overcomes
The problem of part distributed intelligence after optimum individual progress local search is disturbed;
3, using the local search based on Swap neighborhood and based on Inverse neighborhood, the part for further strengthening algorithm is searched
Suo Nengli.
Detailed description of the invention
Fig. 1 is that the coding of solution and factory's distribution illustrate;
Fig. 2 is the pseudocode of MECF rule;
Fig. 3 is Swap operation chart;
Fig. 4 is Inverse operation chart;
Fig. 5 is EEDA algorithm flow chart of the invention.
Specific embodiment
Embodiment 1: as shown in Figs. 1-5, it is a kind of applied to notebook part it is distributed manufacture during optimization
Dispatching method, by determining distributed displacement pipeline schedule model and optimization aim with limited buffer, and using effective
The optimization method of Estimation of Distribution Algorithm optimizes optimization aim;Scheduling model therein is according to each part in each machine
Machining the time on device is established, and target is by its total completion date Cmax(π) is minimized:
K=1,2, F, i=1,2, nk, j=1,2, m
Wherein, F indicates F isomorphism factory, and k indicates some specific factory, and n indicates that notebook part to be processed is total
Number, nkIndicate that notebook part sum to be processed, i indicate i-th bit in notebook Job Scheduling to be processed in the factory in factory k
It sets, j indicates that a certain machine in factory, m indicate number of machines total in each factory;π=π (1), π (2), π (n) }
Indicate total sequence when notebook part is not allocated to factory, π (n) indicates the notebook in total sequence in nth position
Part, i.e. notebook part on the last one position, the notebook part π that always sorts are generated respectively after factory's allocation rule
Notebook part Machining Sequencing π in factoryk, πk(i) it indicates in factory k, in Job Scheduling to be processed zero on i-th of position
Part, πk={ πk(1),πk(2),···,πk(nk) indicate factory k in each part of notebook sequence (such as Fig. 1, always sort π=
{ 6,9,3,5,1,2,4,8,7,10 }, after factory distributes, the sequence π of factory one1={ π1(1),π1(2),···,π1
(n1)={ 6,5, Isosorbide-5-Nitrae, 10 }, π1(1) it represents in factory one, No. 6 parts in Job Scheduling on the 1st position);Indicate the part π in factory kk(i) operation on machine j, in factory k, notebook part πk(i) will
ByOperation could be processed completion, whereinIndicate the notebook in factory k
Part πk(i) operation on machine m, for part after being assigned to some factory, all operations of the part all must be
It is completed the process in the factory;Indicate the part π in factory kk(i) start the processing moment on machine j,Table
Show the part π in factory kk(i) process time on machine j, and it is greater than 0,Indicate the part π in factory kk
(i) moment is completed the process on machine j,Indicate part π in factory kk (nk)On machine m when completing the process
It carves,It indicates in the buffer area in factory k between machine j-1 and j;In addition, arrange all parts be it is independent, at 0 moment,
Any part is allowed to be processed, on every machine in the factory, the processing sequence of part is not changing after determining, in buffer area
In part obey first in, first out rule;The same part can only be processed on a machine at the same moment;With for the moment
It carves, a part can only be processed on a machine, some operation of part does not allow midway to stop;Process timeWith
Buffer areaIt is known that the traveling time between the setting time and operation of machine is ignored.
Further, specific step is as follows for the optimization method of effective Estimation of Distribution Algorithm:
The coding and decoding mode of Step1, solution: it is encoded according to the sequencing that part is processed, is contained in algorithm population
There are multiple individuals, one of each individual correspondence problem solves or part always sorts (such as be n's for notebook number of components
DPFSSP_LB problem, π={ π (1), π (2), π (n) } are exactly a solution in DPFSSP_LB problem.Such as Fig. 1, one
It is π={ 6,9,3,5,1,2,4,8,7,10 } that a solution or part, which always sort, and part 6 is initially treated, followed by part 9,
And so on, it is finally part 10);Decoding is exactly that the coding and sorting order is generated a feasible execution by factory's allocation rule
Scheme;Such as Fig. 1, if the processing sequence of factory one and factory two is generated there are two factory.It is calculated again by the problem model
Final solution out.
Step2, probabilistic model initialization: the probability point in algorithm gen generation is indicated using the matrix P (g) of n × n dimension
Cloth model, algorithm are set as 1/n, are specifically expressed as follows in initial phase, each element set in probabilistic model:
Wherein, Pn,1(g) indicate g for when notebook part 1 in always sequence π nth position or before the probability that occurs
Size, Pn,n(g) from the processing priority for numerically indicating notebook difference part;
Step3, sampling generate new population: setting population scale is popsize, and it is competing to carry out championship to probabilistic model P (g)
It strives sampling and generates g for the processing sequence of notebook part to be processed in individual each in new population, enable g=g+1;In the operation
During, that part z is specifically put on the y of position all by Py,z(g) size determines.If after part z is selected, just will
Z column are arranged to 0 in probabilistic model, then P (g) are carried out row normalized, so that row in probabilistic model and be 1.
Step4, local search: after sampling generates new population, Swap and Inverse are successively implemented to optimum individual in population
Two kinds of local searches, the individual after search, factory's sequence generate the total sequence of individual, search by anti-factory's mapping ruler (MECF)
If the individual than before of individual afterwards is more excellent, otherwise the individual before replacement is not replaced;
Define πk_maxIt is each factory's sequence πkThe middle maximum sequence completed in factory, defines πk_eIt is each factory's sequence πkIn
The sequence in other factories other than maximum completion factory, defines Swap (πk_max,u,πk_e, v) and it is in each factory sequence πk
It is middle πk_maxU-th of position on part and πk_eV-th of position on part exchange.Inverse(πk_max, w, q) be
In each factory sequence πkIt is middle πk_maxIn w-th of position to q-th of position on part reversion.The process of local search is as follows
It is shown:
Step 4.1: disturbance operation, for each factory sequence πk, execute Swap (πk_max,u,πk_e, v) and operation, generate each work
Factory's sequence
Step 4.2: search operation,
Enable loop=1;
It starts the cycle over;
It sorts for each factory
In πk_maxMiddle random selection position w and q, and w and q are unequal;
Execute Inverse (πk_max, w, q) and operation, generate each factory's sequence
It is calculated using Step1Total complete time, ifTotal complete time ratio πkTotal complete time it is small,
ThenReplacementOtherwise it does not replace;
Enable loop++;
Circulation is jumped out as loop > Num* (Num-1);
Step 4.3: ifTotal complete time ratio πkTotal complete time it is small, thenReplace πk, otherwise do not replace
It changes;
Step 4.4: factory being sorted π using MECF rulekMap back total sequence π.
Disturbance operation of second step, is that algorithm Premature Convergence falls into local optimum in order to prevent, is convenient for algorithm during this
Jump to the region there may be optimal solution;Third step is search operation, and search is possible to deposit in the region that disturbance operation is found
Optimal solution.In every generation of algorithm, individual optimal in current population is executed local search 200 times.
The update of Step5, probabilistic model: T excellent individual is picked out in the population after local search, and to outstanding
Part position in body is recorded accordingly, is generated the probability matrix P (g+1) in g+1 generation, is updated using following formula
Probabilistic model:
Wherein, Py,z(g) indicate g for when notebook part z in always sequence π y-th of position or before the probability that occurs
Size, Py,z(g+1) indicate g+1 for when notebook part z in always sequence π y-th of position or before the probability that occurs it is big
Small, α is the learning rate of P (g) and its value between 0 to 1, and T is excellent individual number,It is defined as follows:
Step6, termination condition: the maximum operation algebra of set algorithm is Max_gen, if current algebra gen is less than most
Big operation algebra Max_gen, goes to Step3, iterates, and when meeting termination condition, exports optimal solution.
As follows it is possible to further which parameter is arranged: the population scale popsize=100, learning rate α=0.01 are delayed
Rush the maximum value buffer_size=3 in area, the excellent individual number T=10 that probability matrix selects when updating, the maximum of algorithm operation
Algebra Max_gen=1000.
Specific comparative test is as follows: by EEDA designed by the present invention and current existing mainstream algorithm-EDA (see document
Wang S Y,Wang L,Liu M,Xu Y.An effective estimation of distribution algorithm
for solving the distributed permutation flow-shop scheduling problem[J]
.Int.J.Production Economics, 2013,145 (1): 387-396.) it compares, demonstrate the validity of EEDA.
Algorithm realizes that operating system Win10, processor is Intel (R) Core (TM) i5-4590CPU using Delphi2010 programming
3.40GHz inside saves as 12GB.For each test problem, the identical algebra of two kinds of algorithm each runs, and independent operating 20
It is secondary, then average to 20 calculated result.Test result is as shown in table 1.Wherein, AVG indicates the flat of 20 calculated results
Mean value, Min indicate solution best in 20 calculated results, and Max indicates solution worst in 20 calculated results, and DX indicates 20 meters
Calculate the variance of result.
Seen from table 1, in these indexs considered here, in addition to individual fingers in 2 × 50,2 × 70 this two groups of data
Outside being marked with, EEDA shows superperformance on other problems.This shows the good learning algorithm of EEDA energy in search process
It was found that high-quality solution distributed intelligence, scanned for so that bootstrap algorithm jumps to high-quality solution region, at the same based on Swap neighborhood and
Search based on Interchange neighborhood can carry out careful local search to excellent solution region to find high-quality solution.Therefore,
EEDA is to solve for a kind of effective algorithm of DFSSP_LB.
Above in conjunction with attached drawing, the embodiment of the present invention is explained in detail, but the present invention is not limited to above-mentioned
Embodiment within the knowledge of a person skilled in the art can also be before not departing from present inventive concept
Put that various changes can be made.
Claims (4)
1. the Optimization Scheduling during a kind of distributed manufacturing applied to notebook part, it is characterised in that: logical
It crosses and determines distributed displacement pipeline schedule model and optimization aim with limited buffer, and use effective Estimation of Distribution Algorithm
Optimization method optimization aim is optimized;Scheduling model therein is processing on each machine according to each part
It is established at the time, target is by its total completion date Cmax(π) is minimized:
K=1,2, F, i=1,2, nk, j=1,2, m
Wherein, F indicates F isomorphism factory, and k indicates some specific factory, and n indicates notebook part sum to be processed, nk
Indicate that notebook part sum to be processed in factory k, i indicate that i-th bit is set in notebook Job Scheduling to be processed in the factory, j
Indicate that a certain machine in factory, m indicate number of machines total in each factory;π=π (1), π (2), π (n) } it indicates
Notebook part is not allocated to total sequence when factory, and π (n) indicates the notebook part in total sequence in nth position,
Notebook part i.e. on the last one position, the notebook part π that always sorts generate each factory after factory's allocation rule
Middle notebook part Machining Sequencing πk, πk(i) it indicates in factory k, part, π on i-th of position in Job Scheduling to be processedk
={ πk(1),πk(2),···,πk(nk) indicate factory k in each part of notebook sequence;It indicates in factory k
Part πk(i) operation on machine j, in factory k, notebook part πk(i) will pass throughOperation could be processed completion, whereinIndicate the notebook part in factory k
πk(i) operation on machine m, for part after being assigned to some factory, all operations of the part all must be in the work
It is completed the process in factory;Indicate the part π in factory kk(i) start the processing moment on machine j,It indicates
The part π in factory kk(i) process time on machine j, and it is greater than 0,Indicate the part π in factory kk(i)
The moment is completed the process on machine j,Indicate part π in factory kk(nk) on machine m the moment is completed the process,It indicates in the buffer area in factory k between machine j-1 and j;In addition, arrange all parts be it is independent, at 0 moment, allow
Any part is processed, and on every machine in the factory, the processing sequence of part is not changing after determining, in the buffer
Part obeys first in, first out rule;The same part can only be processed on a machine at the same moment;Synchronization, one
A part can only be processed on platform machine, some operation of part does not allow midway to stop;Process timeThe buffer area andIt is known that the traveling time between the setting time and operation of machine is ignored.
2. the Optimized Operation side during the distributed manufacturing according to claim 1 applied to notebook part
Method, it is characterised in that: specific step is as follows for the optimization method of effective Estimation of Distribution Algorithm:
The coding and decoding mode of Step1, solution: being encoded according to the sequencing that part is processed, containing more in algorithm population
One solution either part of individual, each individual correspondence problem always sorts;Decoding is exactly that the coding and sorting order is passed through factory
Allocation rule generation one is feasible to carry into execution a plan;
Step2, probabilistic model initialization: the probabilistic model in algorithm gen generation is indicated using the matrix P (g) of n × n dimension, is calculated
Method is set as 1/n, is specifically expressed as follows in initial phase, each element set in probabilistic model:
Wherein, Pn,1(g) indicate g for when notebook part 1 in always sequence π nth position or before the probability that occurs it is big
It is small, Pn,n(g) from the processing priority for numerically indicating notebook difference part;
Step3, sampling generate new population: setting population scale is popsize, carries out championship competition to probabilistic model P (g) and adopts
Sample generates g for the processing sequence of notebook part to be processed in individual each in new population, enables g=g+1;
Step4, local search: after sampling generates new population, Swap and two kinds of Inverse are successively implemented to optimum individual in population
Local search, the individual after search, factory's sequence generate the total sequence of individual by anti-factory's mapping ruler, and the individual after search is such as
Fruit individual than before is more excellent, then otherwise the individual before replacing is not replaced;
The update of Step5, probabilistic model: T excellent individual is picked out in the population after local search, and in excellent individual
Part position recorded accordingly, generate g+1 generation probability matrix P (g+1), using following formula come update probability
Model:
Wherein, Py,z(g) indicate g for when notebook part z in always sequence π y-th of position or before the probability that occurs it is big
It is small, Py,z(g+1) indicate g+1 for when notebook part z in always sequence π y-th of position or before the probability size that occurs, α
It is the learning rate of P (g) and its value between 0 to 1, T is excellent individual number,It is defined as follows:
Step6, termination condition: the maximum operation algebra of set algorithm is Max_gen, if current algebra gen is less than maximum fortune
Row algebra Max_gen, goes to Step3, iterates, and when meeting termination condition, exports optimal solution.
3. the Optimized Operation side during the distributed manufacturing according to claim 2 applied to notebook part
Method, it is characterised in that: anti-factory's mapping ruler sorts factory πkMap back total sequence π.
4. the Optimized Operation side during the distributed manufacturing according to claim 2 applied to notebook part
Method, it is characterised in that: the population scale popsize=100, learning rate α=0.01, the maximum value buffer_ of buffer area
Size=3, the excellent individual number T=10 that probability matrix selects when updating, the maximum algebra Max_gen=1000 of algorithm operation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710961541.4A CN107831740B (en) | 2017-10-17 | 2017-10-17 | It is a kind of applied to notebook part it is distributed manufacture during Optimization Scheduling |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710961541.4A CN107831740B (en) | 2017-10-17 | 2017-10-17 | It is a kind of applied to notebook part it is distributed manufacture during Optimization Scheduling |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107831740A CN107831740A (en) | 2018-03-23 |
CN107831740B true CN107831740B (en) | 2019-12-03 |
Family
ID=61648119
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710961541.4A Active CN107831740B (en) | 2017-10-17 | 2017-10-17 | It is a kind of applied to notebook part it is distributed manufacture during Optimization Scheduling |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107831740B (en) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109002017A (en) * | 2018-07-06 | 2018-12-14 | 昆明理工大学 | A kind of Optimization Scheduling of Suresh Kumar hollow glass manufacture |
CN109085804A (en) * | 2018-07-17 | 2018-12-25 | 昆明理工大学 | It is a kind of for electronic product multiplexing factory manufacture process Optimization Scheduling |
CN109213094A (en) * | 2018-07-25 | 2019-01-15 | 昆明理工大学 | A kind of Optimization Scheduling based on multiplexing inter-plant steel smelting-continuous casting production steel billet process |
CN110276481B (en) * | 2019-05-31 | 2021-11-26 | 清华大学 | Distributed hybrid pipeline scheduling optimization method |
CN110928261B (en) * | 2019-12-17 | 2021-05-28 | 清华大学 | Distributed estimation scheduling method and system for distributed heterogeneous flow shop |
CN111290283B (en) * | 2020-04-03 | 2021-07-27 | 福州大学 | Additive manufacturing single machine scheduling method for selective laser melting process |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102346882A (en) * | 2011-09-16 | 2012-02-08 | 中冶南方工程技术有限公司 | Bilayer planning method for intelligent scheduling for steelmaking-continuous welding production |
CN104408528A (en) * | 2014-11-20 | 2015-03-11 | 昆明理工大学 | Optimization scheduling method in raw material leaching process for chemical industry production |
CN106485347A (en) * | 2016-09-21 | 2017-03-08 | 哈尔滨理工大学 | The two workshop flexible comprehensive dispatching methods that buffering operation migrates in advance |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070050070A1 (en) * | 2005-08-05 | 2007-03-01 | Pfizer Inc | Automated batch manufactuirng |
-
2017
- 2017-10-17 CN CN201710961541.4A patent/CN107831740B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102346882A (en) * | 2011-09-16 | 2012-02-08 | 中冶南方工程技术有限公司 | Bilayer planning method for intelligent scheduling for steelmaking-continuous welding production |
CN104408528A (en) * | 2014-11-20 | 2015-03-11 | 昆明理工大学 | Optimization scheduling method in raw material leaching process for chemical industry production |
CN106485347A (en) * | 2016-09-21 | 2017-03-08 | 哈尔滨理工大学 | The two workshop flexible comprehensive dispatching methods that buffering operation migrates in advance |
Also Published As
Publication number | Publication date |
---|---|
CN107831740A (en) | 2018-03-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107831740B (en) | It is a kind of applied to notebook part it is distributed manufacture during Optimization Scheduling | |
CN107590603B (en) | Based on the dispatching method and system for improving change neighborhood search and differential evolution algorithm | |
Zheng et al. | Combining reinforcement learning with Lin-Kernighan-Helsgaun algorithm for the traveling salesman problem | |
Li et al. | A discrete artificial bee colony algorithm for the distributed heterogeneous no-wait flowshop scheduling problem | |
Shao et al. | LS-HH: A learning-based selection hyper-heuristic for distributed heterogeneous hybrid blocking flow-shop scheduling | |
CN101901425A (en) | Flexible job shop scheduling method based on multi-species coevolution | |
CN110533301A (en) | A kind of population dispatching method based on dynamic constrained matrix | |
Shobaki et al. | An exact algorithm for the sequential ordering problem and its application to switching energy minimization in compilers | |
CN109858780B (en) | Steelmaking-continuous casting production scheduling optimization method | |
CN113569483A (en) | Method for solving multi-target flexible job shop scheduling based on artificial bee colony algorithm | |
CN110928261A (en) | Distributed estimation scheduling method and system for distributed heterogeneous flow shop | |
Sajadi et al. | A new fuzzy multi-objective multi-mode resource-constrained project scheduling model | |
Guo et al. | A learning interactive genetic algorithm based on edge selection encoding for assembly job shop scheduling problem | |
Demir et al. | Process planning and due-date assignment with ATC dispatching where earliness, tardiness and due-dates are punished | |
CN102222274A (en) | Immune clone selection job shop scheduling method based on scheduling coding | |
CN106598716A (en) | Task scheduling method based on multiple processors | |
CN109783189A (en) | A kind of quiescent operation stream scheduling method and device | |
Etminaniesfahani et al. | An efficient relax-and-solve method for the multi-mode resource constrained project scheduling problem | |
Ivutin et al. | Optimization Problem for Heterogeneous Computing Systems | |
Zagoruiko et al. | International competition for technological efficiency: Methodology of nonparametric modelling | |
Luh et al. | Architectural design of neural network hardware for job shop scheduling | |
Okada et al. | A genetic algorithm with local search using activity list characteristics for solving resource‐constrained multiproject scheduling problem | |
CN114254721B (en) | Flexible job shop scheduling method and equipment with interval gray processing time | |
Tian et al. | Scheduling Unrelated Parallel Batch Processing Machines Under Time-of-Use Electricity Prices | |
CN114839930B (en) | Integrated scheduling system for distributed assembly blocking flow shop |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |