CN109255484B - Data-driven discrete manufacturing resource collaborative optimization method and system - Google Patents
Data-driven discrete manufacturing resource collaborative optimization method and system Download PDFInfo
- Publication number
- CN109255484B CN109255484B CN201811062143.XA CN201811062143A CN109255484B CN 109255484 B CN109255484 B CN 109255484B CN 201811062143 A CN201811062143 A CN 201811062143A CN 109255484 B CN109255484 B CN 109255484B
- Authority
- CN
- China
- Prior art keywords
- solution
- optimal solution
- initial
- partial
- local
- 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 80
- 238000000034 method Methods 0.000 title claims abstract description 43
- 238000005457 optimization Methods 0.000 title claims abstract description 23
- 229910052782 aluminium Inorganic materials 0.000 claims abstract description 20
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 claims abstract description 20
- 238000010845 search algorithm Methods 0.000 claims abstract description 19
- 230000008569 process Effects 0.000 claims abstract description 17
- 230000007246 mechanism Effects 0.000 claims abstract description 11
- 238000004458 analytical method Methods 0.000 claims abstract description 10
- 230000008859 change Effects 0.000 claims description 26
- 238000012545 processing Methods 0.000 claims description 14
- 238000000605 extraction Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000009191 jumping Effects 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 4
- 238000003491 array Methods 0.000 claims description 3
- 230000003247 decreasing effect Effects 0.000 claims description 3
- 230000001066 destructive effect Effects 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000005242 forging Methods 0.000 abstract description 10
- 230000006378 damage Effects 0.000 abstract description 3
- 238000000354 decomposition reaction Methods 0.000 description 4
- 230000006866 deterioration Effects 0.000 description 4
- 230000001419 dependent effect Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013468 resource allocation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- Manufacturing & Machinery (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a data-driven discrete manufacturing resource collaborative optimization method and system. The method comprises the following steps: the historical production data is analyzed through a multi-factor variance analysis method, core influence factors influencing the production and manufacturing efficiency are extracted, and then the net benefits of all workpieces, namely the fitness function, are calculated according to the core influence factors. And then, a variable neighborhood search algorithm is introduced by referring to a mechanism of executing destruction and reconstruction on the existing solution by the optimal solution, and the greedy reference iterative algorithm is optimized. The method can improve the robustness and diversity of the solution, can effectively improve the production efficiency of the forging forming process of the aluminum ingot, and can be popularized to the discrete manufacturing process of complex products to reduce the production cost.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a data-driven discrete manufacturing resource collaborative optimization method and system.
Background
With the rapid development of the production and manufacturing industry, a large amount of human resources are required to be invested in the traditional production and manufacturing cooperative scheduling, the optimization degree greatly depends on the experience of personnel, and the trend of large-scale and fine production and manufacturing brings new challenges to the traditional manufacturing mode.
Under the influence of multiple factors such as material characteristics, processing temperature, machine performance and the like, a discrete manufacturing process is very complex, for example, in the actual generation process of metal product forging and workpiece forming, the problem of collaborative optimization of discrete manufacturing resources with two-stage deterioration effect widely exists, for example, an aluminum ingot needs to be preheated to a certain temperature in the process of processing and forming, the temperature is reduced in the process of waiting for processing, and if the temperature of the aluminum ingot during processing is higher than the temperature limit required by processing, the aluminum ingot can be processed; otherwise, the aluminum ingot needs to be heated again to recover to the temperature limit and then processed, so that the aluminum ingot needs to be scheduled and optimized in a coordinated manner, the cost is reduced, and the production efficiency is improved.
In the related art, Barketau et al (2008) consider the problem of single machine continuous batch scheduling with two-stage deteriorated workpieces, proving that the problem in this case aimed at minimizing completion time is a strong NP-hard problem. In addition, Leung et al (2008) extended the single-machine problem to a multi-machine environment, demonstrating that even if the deterioration period of the machine is the same in a multi-machine environment, the problem remains a strong NP-hard problem.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a data-driven discrete manufacturing resource collaborative optimization method and a data-driven discrete manufacturing resource collaborative optimization system, which are used for solving the problem of difficult strong NP in the resource allocation process in the related technology.
The embodiment of the invention provides a data-driven discrete manufacturing resource collaborative optimization method, which comprises the following steps:
step 1, analyzing historical production data by using a multi-factor variance analysis method to obtain core factors representing influences on production and manufacturing;
step 2, encoding all workpieces and machine nodes by adopting a random key encoding mode;
step 3, calculating the fitness value of the initial solution based on the randomly generated initial solution, and setting the initial solution as a global optimal solution and a local optimal solution;
step 4, local search is carried out by adopting a local search algorithm based on a greedy reference iterative algorithm, a local optimal solution is re-determined, and if the re-determined local optimal solution is superior to the global optimal solution, the re-determined local optimal solution is selected to update the global optimal solution;
step 5, let x1＝x2＝1，i＝1；
Step 6, executing neighborhood structure change Ni(xi)；
Step 7, carrying out local search by using a local search algorithm based on a greedy reference iterative algorithm to obtain a local optimal solution; if the local optimal solution is superior to the global optimal solution, replacing the global optimal solution by the local optimal solution;
step 8, let xi＝xi+1, i ═ i + 1; if i is less than or equal to 2, returning to the step 6; otherwise, turning to step 9 if i is equal to 1;
step 9, if x2≤xmaxReturning to the step 6; otherwise, outputting a local optimal solution and a global optimal solution;
and step 10, decoding according to the local optimal solution and the global optimal solution, and then optimizing production scheduling on the aluminum ingot hot working problem.
Optionally, the encoding all the workpieces and the machine nodes by using a random key encoding method includes:
randomly generating a decimal array with the length of n + m-1 and the numerical value of 0-1; n and m respectively represent the number of workpieces and the number of machines;
mapping the decimal array into workpiece number arrays processed on various machines;
converting each decimal in the decimal array into integers from small to large, wherein the integers are 1, 2 and … … respectively;
and taking an integer larger than n as a segmentation mark, and distributing the workpiece behind the segmentation mark to other machines.
Optionally, the local search is performed by using a greedy-based reference iterative algorithm, and obtaining a local optimal solution includes:
1) setting maximum iteration times xi of local search disturbance mechanism based on greedy reference iteration algorithm and minimum number q of extraction positions of reference greedy iteration stageminAnd maximum value qmax；
2) Randomly generating an initial solution or generating a solution by neighborhood structure change; the solution is a decimal array with the length of n + m-1 and the numerical value of 0-1;
3) decoding the solution, batching workpieces on each machine according to an LSPT rule, and calculating the fitness Fit;
4) randomly extracting a value at any position in the solution, and then inserting the value into the optimal position in the solution;
5) let q be qmin；
6) Let η equal to 1;
7) setting the number of values of the further extracted positions as q, randomly extracting the values of the q positions in the solution, and enabling the solution after extraction to be a partial solution pipartialLet δ be 1;
8) reinserting the δ -th decimated value in turn to the optimal position in the partial solution s]；PS[s]Indicates the position [ s ]]A value of (d) above;
9) PS' stands for and partially resolves PS[s]The same value, PS' will solve optimally pibestThe device is divided into a left part and a right part as reference;
10) let δ be δ +1, if δ > q, a perturbation mechanism is completed, a new solution is obtained, go to step 10); otherwise, returning to the step 8);
11) let η ═ η +1, if η ≦ ξ, return to step 7); otherwise go to step 12);
12) if xi times of iteration still does not obtain a better solution, making q be q +1, and going to step 13); otherwise, terminating the iteration;
13) if q is less than or equal to qmaxAnd returning to the step 6); otherwise, the iteration is terminated.
Optionally, the step 9) includes:
9.1) randomly selecting a position in the left half of the optimal solution, the value of which is denoted PSL；
9.1.1) partial decomposition of πpartialIn the presence of PSLValue, then exchange PS in partial solutionLAnd
if not, jumping to step 9.1.3);
9.1.2) partial decomposition of π after exchangepartialIf the fitness is reduced, jumping to step 9.2); otherwise, carrying out the next step;
9.1.3) extraction
Value, then inserted into the optimal position such that the partial solution is pipartialThe fitness of (2) is lowest;
9.2) value of randomly selecting a position in the right half of the optimal solution is denoted PSR；
9.2.1) partial decomposition of πpartialIn the presence of PSRValue, then exchange PS in partial solutionRAnd
if not, go to step 9.2.3);
9.2.2) post-decomposition if exchange πpartialIf the fitness is reduced, jumping to step 10); otherwise, carrying out the next step;
9.2.3) extracting
The value is then inserted into the optimal position such that the partial solution is pipartialThe fitness of (2) is lowest.
Optionally, the performing the local search by using a local search algorithm based on a greedy reference iterative algorithm includes:
1) randomly generating an initial solution or producing a solution pi from neighborhood structure changesinitial；
2) Performing a local search to find a locally optimal solution and then replacing the initial solution by itinitial；
3) If piinitialIs randomly generated, let piincumbent＝πinitial，πbest＝πinitialAnd q ═ qmin(ii) a If piinitialIs generated by the change of the neighborhood structure, let piincumbent＝πinitial，q＝qmin；
4) Let θ be 1;
5) performing a destructive process by applying at piincumbentRandomly extracting q positions and generating partial solution of pi by the rest sequencepartial；
6) Performing a reconstruction process by referencing the optimal solution pibestInserting the values of the q positions into the partial solution pi in turnpartialTo obtain a new solution pinew；
7) If pinewIs less than piincumbentThe fitness of (a) is thenincumbentIs replaced by pinew(ii) a If pinewIs less than pibestThe fitness of (a) is thenbestIs replaced by pinew；
8) Let θ be θ + 1; if theta is less than or equal to xi, returning to the step 5); otherwise go to step 9);
9) if not, the solution is piinitialA better solution, let q be q +1, go to step 10);
10) if q is less than or equal to qmaxAnd returning to the step 4); otherwise, the iteration is terminated.
Optionally, the step of calculating the fitness value comprises:
1) sequencing all workpieces on each machine according to the non-decreasing core parameters of the workpieces;
2) machine MjThe number of upper workpieces is represented by njAnd c, batching the workpieces on each machine, wherein the capacity of each batch of workpieces is c, and the batching method comprises the following steps:
2.1) setting j to 1;
2.2) will be before
The workpieces form a batch;
2.3) if the workpieces which are not batched exist, forming the first c workpieces of the workpieces which are not batched into a batch, and if the workpieces which are not batched still exist, continuing to execute the step until all the workpieces on the machine are batched;
2.4) machine MjThe batches are processed according to the sequence of batch processing;
2.5) if j is less than the number of the devices, making j equal to j +1, and turning to step 2.2); otherwise, performing step 3);
3) calculating the completion time C of the last batch on each machinejNet profit
4) And taking the TNR value calculated in the step 3) as a solution fitness value.
The embodiment of the invention also provides a data-driven discrete manufacturing resource collaborative optimization system, which comprises:
the core factor acquisition module is used for analyzing historical production data by using a multi-factor variance analysis method to obtain core factors representing influences on production and manufacturing;
the node coding module is used for coding all workpieces and machine nodes by adopting a random key coding mode;
the fitness calculation module is used for calculating the fitness value of the initial solution based on random generation and setting the initial solution as a global optimal solution and a local optimal solution;
the optimal solution updating module is used for carrying out local search by adopting a local search algorithm based on a greedy reference iterative algorithm, re-determining a local optimal solution, and if the re-determined local optimal solution is superior to a global optimal solution, selecting the re-determined local optimal solution to update the global optimal solution;
an initialization module for ordering x1＝x2＝1，i＝1；
A domain change module for performing a neighborhood structure change Ni(xi)；
The optimal solution updating module is also used for carrying out local search by using a local search algorithm based on a greedy reference iterative algorithm to obtain a local optimal solution; if the local optimal solution is superior to the global optimal solution, replacing the global optimal solution by the local optimal solution;
the initialization module is further used for enabling xi＝xi+1, i ═ i + 1; if i is less than or equal to 2, triggering the domain change module; otherwise, setting i to be 1, and triggering an optimal solution output module;
an optimal solution output module for outputting the optimal solution at x2≤xmaxIf yes, returning to the domain change module; otherwise, outputting a local optimal solution and a global optimal solution;
and the optimal decoding module is used for decoding according to the local optimal solution and the global optimal solution and then optimizing the production scheduling of the aluminum ingot hot processing problem.
According to the technical scheme, the existing historical production data can be analyzed through the multi-factor variance analysis method, core factors which influence the objective production and manufacturing efficiency can be extracted, and therefore the existing production and manufacturing data can be fully utilized. And then calculating the net benefits of all workpieces, namely the fitness function, according to the core influence factors, so that the establishment of the production model and the calculation of the fitness function are more reasonable and fit with reality.
In the embodiment, a variable neighborhood search algorithm is introduced by referring to a mechanism of executing destruction and reconstruction on the existing solution by the optimal solution, a greedy reference iterative algorithm is optimized, the greedy reference iterative algorithm is prevented from being trapped in the local optimal solution too early, the robustness and diversity of the obtained solution are improved, the production efficiency of the forging and forming process of the aluminum ingot is effectively improved, the production cost is reduced, and meanwhile, the method can be popularized to the discrete manufacturing process of complex products.
In this embodiment, the two-stage deterioration effect is introduced into the model for forging and forming the aluminum ingot by studying the problem of collaborative optimization of discrete resources in the process of forging and forming the aluminum ingot, so that the model better conforms to the actual process of forging and forming the aluminum ingot.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a schematic flow chart of a method for collaborative optimization of a data-driven discrete manufacturing resource according to an embodiment of the present invention;
FIG. 2 is a block diagram of a data-driven discrete manufacturing resource co-optimization system according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a flowchart illustrating a method for collaborative optimization of data-driven discrete manufacturing resources according to an embodiment of the present invention. Referring to fig. 1, the data-driven discrete manufacturing resource collaborative optimization method includes steps 1 to 10, wherein:
step 1, analyzing historical production data by using a multi-factor variance analysis method to obtain core factors influencing production and manufacturing in a table.
In this step, a plurality of factors in the historical production data, such as the length of the workpiece, the thickness of the workpiece, the temperature of the workpiece, the normal processing speed of the machine, and the like, are used as independent variables, and dependent variables in the production data, such as the processing time of the workpiece, are found. And then carrying out variance analysis on the independent variable and the dependent variable, and finding out factors which obviously influence the dependent variable as core factors influencing production and manufacturing. Therefore, the embodiment can filter out more influences with small influence or no influence, and reduce the calculation amount.
And 2, encoding all workpieces and machine nodes by adopting a random key encoding mode.
In the step, 1) a decimal array with the length of n + m-1 and the numerical value of 0-1 is randomly generated, and n and m respectively represent the number of workpieces and the number of machines.
2) Mapping the decimal array into workpiece number arrays processed on various machines;
2.1) converting each decimal in the decimal array from small to large into integers, wherein the integers are 1, 2 and … … respectively;
2.1) taking an integer larger than n as a dividing mark, and distributing the workpiece behind the dividing mark to other machines.
For example, there are 6 workpieces and 3 machines, the decimal array (0.41,0.03,0.76,0.25,0.65,0.89,0.35,0.28) is mapped to the integer array as (5,1,7,2,9,6,8,4,3), indicating that workpiece 5 and workpiece 1 are machined on the first machine, work 2 and workpiece 6 are machined on the second machine, and workpiece 4 and workpiece 3 are machined on the third machine.
Step 3, calculating the fitness value of the initial solution based on the randomly generated initial solution, and setting the initial solution as a global optimal solution and a local optimal solution pibest。
In this step, the step of calculating the fitness value includes:
1) sequencing all workpieces on each machine according to the non-decreasing core parameters of the workpieces;
2) machine MjThe number of upper workpieces is represented by njAnd c, batching the workpieces on each machine, wherein the capacity of each batch of workpieces is c, and the batching method comprises the following steps:
2.1) setting j to 1;
2.2) will be before
The workpieces form a batch;
2.3) if the workpieces which are not batched exist, forming the first c workpieces of the workpieces which are not batched into a batch, and if the workpieces which are not batched still exist, continuing to execute the step until all the workpieces on the machine are batched;
2.4) machine MjThe batches are processed according to the sequence of batch processing;
2.5) if j is less than the number of the devices, making j equal to j +1, and turning to step 2.2); otherwise, performing step 3);
3) calculating the completion time C of the last batch on each machinejNet profit
4) And taking the TNR value calculated in the step 3) as a solution fitness value.
And 4, carrying out local search by adopting a local search algorithm based on a greedy reference iterative algorithm, re-determining a local optimal solution, and selecting the re-determined local optimal solution to update the global optimal solution if the re-determined local optimal solution is superior to the global optimal solution.
In this step, the local search algorithm based on the greedy reference iterative algorithm comprises:
1) setting maximum iteration times xi of local search disturbance mechanism based on greedy reference iteration algorithm and minimum number q of extraction positions of reference greedy iteration stageminAnd maximum value qmax；
2) Randomly generating an initial solution or generating a solution by neighborhood structure change;
3) decoding the solution, batching workpieces on each machine according to an LSPT rule, and calculating the fitness Fit;
4) randomly extracting a value at any position in the solution, and then inserting the value into the optimal position in the solution;
5) let q be qmin；
6) Let η equal to 1;
7) setting the number of values of the further extracted positions as q, randomly extracting the values of the q positions in the solution, and enabling the solution after extraction to be a partial solution pipartialLet δ be 1;
8) reinserting the δ -th decimated value in turn to the optimal position in the partial solution s]；PS[s]Indicates the position [ s ]]A value of (d) above;
9) PS' stands for and partially resolves PS[s]The same value, PS' will solve optimally pibestThe device is divided into a left part and a right part as reference;
10) let δ be δ +1, if δ > q, a perturbation mechanism is completed, a new solution is obtained, go to step 10); otherwise, returning to the step 8);
11) let η ═ η +1, if η ≦ ξ, return to step 7); otherwise go to step 12);
12) if xi times of iteration still does not obtain a better solution, making q be q +1, and going to step 13); otherwise, terminating the iteration;
13) if q is less than or equal to qmaxAnd returning to the step 6); otherwise, the iteration is terminated.
Then, the output better solution is taken as a local optimal solution. And if the re-determined local optimal solution is superior to the global optimal solution, selecting the re-determined local optimal solution to update the global optimal solution.
Step 5, let x1＝x2＝1，i＝1；
Step 6, executing neighborhood structure change Ni(xi)；
Step 7, carrying out local search by using a local search algorithm based on a greedy reference iterative algorithm to obtain a local optimal solution; and if the local optimal solution is superior to the global optimal solution, replacing the global optimal solution by using the local optimal solution.
In this step, the step of obtaining the local optimal solution refers to step 4, and the step of updating the global optimal solution also refers to step 4, which is not described herein again.
Step 8, let xi＝xi+1, i ═ i + 1; if i is less than or equal to 2, returning to the step 6; otherwise, let i equal to 1, go to step 9.
Step 9, if x2≤xmaxReturning to the step 6; otherwise, outputting a local optimal solution and a global optimal solution;
and step 10, decoding according to the local optimal solution and the global optimal solution, and then optimizing production scheduling on the aluminum ingot hot working problem.
According to the technical scheme, the existing historical production data can be analyzed through the multi-factor variance analysis method, core factors which influence the objective production and manufacturing efficiency can be extracted, and therefore the existing production and manufacturing data can be fully utilized. And then calculating the net benefits of all workpieces, namely the fitness function, according to the core influence factors, so that the establishment of the production model and the calculation of the fitness function are more reasonable and fit with reality.
In the embodiment, a variable neighborhood search algorithm is introduced by referring to a mechanism of executing destruction and reconstruction on the existing solution by the optimal solution, a greedy reference iterative algorithm is optimized, the greedy reference iterative algorithm is prevented from being trapped in the local optimal solution too early, the robustness and diversity of the obtained solution are improved, the production efficiency of the forging forming process of the aluminum ingot is effectively improved, and the production cost is reduced.
Therefore, in the embodiment, by researching the problem of the collaborative optimization of the discrete resources in the aluminum ingot forging forming process, the two-stage deterioration effect is introduced into the model for aluminum ingot forging forming, so that the model is more in line with the actual process of aluminum ingot forging forming.
Fig. 2 is a block diagram of a data-driven discrete manufacturing resource co-optimization system according to an embodiment of the present invention. Referring to fig. 2, a data-driven discrete manufacturing resource co-optimization system includes:
a core factor obtaining module 201, configured to analyze historical production data by using a multi-factor variance analysis method to obtain core factors representing influences on production and manufacturing;
the node coding module 202 is configured to code all workpieces and machine nodes in a random key coding manner;
a fitness calculation module 203, configured to calculate a fitness value of a randomly generated initial solution based on the initial solution, and set the initial solution as a global optimal solution and a local optimal solution;
the optimal solution updating module 204 is configured to perform local search by using a local search algorithm based on a greedy reference iterative algorithm, to re-determine a local optimal solution, and if the re-determined local optimal solution is better than a global optimal solution, select the re-determined local optimal solution to update the global optimal solution;
an initialization module 205 for letting x1＝x2＝1，i＝1；
A domain change module 206 for performing a neighborhood structure change Ni(xi)；
The optimal solution updating module 204 is further configured to perform local search by using a local search algorithm based on a greedy reference iterative algorithm to obtain a local optimal solution; if the local optimal solution is superior to the global optimal solution, replacing the global optimal solution by the local optimal solution;
the initialization module 205 is further configured to order xi＝xi+1, i ═ i + 1; if i is less than or equal to 2, triggering the domain change module; otherwise, setting i to be 1, and triggering an optimal solution output module;
an optimal solution output module 207 for outputting the optimal solution at x2≤xmaxIf yes, returning to the domain change module; otherwise, outputting a local optimal solution and a global optimal solution;
and the optimal solution decoding module 208 is used for decoding according to the local optimal solution and the global optimal solution and then optimizing the production scheduling of the aluminum ingot hot working problem.
It should be noted that the data-driven discrete manufacturing resource collaborative optimization system provided by the embodiment of the present invention is in a one-to-one correspondence relationship with the above method, and the implementation details of the above method are also applicable to the above system, and the above system will not be described in detail in the embodiment of the present invention.
In the description of the present invention, numerous specific details are set forth. It is understood, however, that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit and scope of the present invention, and they should be construed as being included in the following claims and description.
Claims (5)
1. A data-driven method for collaborative optimization of discrete manufacturing resources, comprising:
step 1, analyzing historical production data by using a multi-factor variance analysis method to obtain core factors representing influences on production and manufacturing;
step 2, encoding all workpieces and machine nodes by adopting a random key encoding mode;
step 3, calculating the fitness value of the initial solution based on the randomly generated initial solution, and setting the initial solution as a global optimal solution and a local optimal solution;
step 4, local search is carried out by adopting a local search algorithm based on a greedy reference iterative algorithm, a local optimal solution is re-determined, and if the re-determined local optimal solution is superior to the global optimal solution, the re-determined local optimal solution is selected to update the global optimal solution;
step 5, let xi＝1，i＝1；
Step 6, executing neighborhood structure change Ni(xi)；
Step 7, carrying out local search by using a local search algorithm based on a greedy reference iterative algorithm to obtain a local optimal solution; if the local optimal solution is superior to the global optimal solution, replacing the global optimal solution by the local optimal solution;
step 8, let xi＝xi+1, i ═ i + 1; if i is less than or equal to 2, returning to the step 6; otherwise, turning to step 9 if i is equal to 1;
step 9, if x2≤xmaxReturning to the step 6; otherwise, outputting a local optimal solution and a global optimal solution;
step 10, decoding according to the local optimal solution and the global optimal solution, and then optimizing production scheduling on the aluminum ingot hot processing problem;
local search is carried out by using a greedy-based reference iterative algorithm, and a local optimal solution is obtained by the method comprising the following steps:
10) setting maximum iteration times xi of local search disturbance mechanism based on greedy reference iteration algorithm and minimum number q of extraction positions of reference greedy iteration stageminAnd maximum value qmax；
20) Randomly generating an initial solution or generating a solution by neighborhood structure change; the initial solution or a solution generated by the neighborhood structure change is a decimal array with the length of n + m-1 and the numerical value of 0-1;
30) decoding the solution, batching workpieces on each machine according to an LSPT rule, and calculating the fitness Fit;
40) randomly extracting a value at any position in the solution, and then inserting the value into the optimal position in the solution;
50) let q be qmin；
60) Let η equal to 1;
70) setting the number of values of the further extracted positions as q, randomly extracting the values of the q positions in the solution, and enabling the solution after extraction to be a partial solution pipartialLet δ be 1;
80) reinserting the δ -th decimated value in turn to the optimal position in the partial solution s]；PS[s]Indicates the position [ s ]]A value of (d) above;
90) PS' stands for and partially resolves PS[s]The same value, PS' will solve optimally pibestThe device is divided into a left part and a right part as reference;
100) let δ be δ +1, if δ > q, a perturbation mechanism is completed, and a new solution pi is obtainednewGo to step 110); otherwise, returning to the step 80);
110) let η be η +1, if η ≦ ξ, return to step 70); otherwise go to step 120);
120) if xi times of iteration still does not obtain a better solution, making q ═ q +1, go to step 130); otherwise, terminating the iteration;
130) if q is less than or equal to qmaxThen returning to step 60); otherwise, terminating the iteration;
the solution change in the local search algorithm based on the greedy reference iterative algorithm comprises the following steps:
1a) randomly generating an initial solution or producing a solution pi from neighborhood structure changesinitial；
2a) Performing a local search to find a locally optimal solution and then replacing the initial solution by itinitial；
3a) If piinitialIs randomly generated, let piincumbent＝πinitial，πbest＝πinitialAnd q ═ qmin(ii) a If piinitialIs generated by the change of the neighborhood structure, let piincumbent＝πinitial，q＝qmin；
4a) Let θ be 1;
5a) performing a destructive process by applying at piincumbentRandomly extracting the values of q positions, the remaining sequence producing a partial solution of pipartial；
6a) Performing a reconstruction process by referencing the optimal solution pibestInserting the values of the q positions into the partial solution pi in turnpartialTo obtain a new solution pinew；
7a) If pinewIs less than piincumbentThe fitness of (a) is thenincumbentIs replaced by pinew(ii) a If pinewIs less than pibestThe fitness of (a) is thenbestIs replaced by pinew；
8a) Let θ be θ + 1; if theta is less than or equal to xi, returning to the step 5 a); otherwise go to step 9 a);
9a) if not, get pi-less than the initial solutioninitialA better solution, let q be q +1, go to step 10 a);
10a) if q is less than or equal to qmaxAnd returning to the step 4 a); otherwise, the iteration is terminated.
2. The method of claim 1, wherein the encoding all the workpiece and machine nodes using random key encoding comprises:
randomly generating a decimal array with the length of n + m-1 and the numerical value of 0-1; n and m respectively represent the number of workpieces and the number of machines;
mapping the decimal array into workpiece number arrays processed on various machines;
converting each decimal in the decimal array into an integer from small to large, wherein the integers are respectively 1, 2, … and n + m-1;
and taking an integer larger than n as a segmentation mark, and distributing the workpiece behind the segmentation mark to other machines.
3. The method for collaborative optimization of discrete manufacturing resources according to claim 1, wherein the 90) includes:
90.1) randomly selecting a position in the left half of the optimal solution, the value of which is denoted PSL；
90.1.1) partial solution of pipartialIn the presence of PSLValue, then exchange PS in partial solutionLAnd PS[s-1](ii) a If not, go to step 90.1.3);
90.1.2) partial resolution of pi after swappingpartialIf the fitness is reduced, the step goes to step 90.2); otherwise, carrying out the next step;
90.1.3) extracting PS[s-1]Value, then inserted into the optimal position such that the partial solution is pipartialThe fitness of (2) is lowest;
90.2) value of randomly selecting a position in the right half of the optimal solution is denoted PSR；
90.2.1) partial solution of pipartialIn the presence of PSRValue, then exchange PS in partial solutionRAnd PS[s+1](ii) a If not presentGo to step 90.2.3);
90.2.2) post-resolution if swapping πpartialIf the fitness is reduced, jumping to step 100); otherwise, carrying out the next step;
90.2.3) extracting PS[s+1]The value is then inserted into the optimal position such that the partial solution is pipartialThe fitness of (2) is lowest.
4. The method for collaborative optimization of discrete manufacturing resources of claim 1, wherein the step of calculating a fitness value comprises:
1b) sequencing all workpieces on each machine according to the non-decreasing core parameters of the workpieces;
2b) machine MjThe number of upper workpieces is represented by njAnd c, batching the workpieces on each machine, wherein the capacity of each batch of workpieces is c, and the batching method comprises the following steps:
2b.1) setting j to 1;
2b.2) will be before
The workpieces form a batch;
2b.3) if the workpieces which are not batched exist, forming the first c workpieces of the workpieces which are not batched into a batch, and if the workpieces which are not batched still exist, continuing to execute the step until all the workpieces on the machine are batched;
2b.4) machine MjThe batches are processed according to the sequence of batch processing;
2b.5) if j is less than the number of the devices, making j equal to j +1, and turning to the step 2 b.2); otherwise, performing step 3 b);
3b) calculating the completion time C of the last batch on each machinejNet profit
4b) Taking the TNR value calculated in the step 3b) as the adaptability value of the solution.
5. A data-driven discrete manufacturing resource co-optimization system, comprising:
the core factor acquisition module is used for analyzing historical production data by using a multi-factor variance analysis method to obtain core factors representing influences on production and manufacturing;
the node coding module is used for coding all workpieces and machine nodes by adopting a random key coding mode;
the fitness calculation module is used for calculating the fitness value of the initial solution based on random generation and setting the initial solution as a global optimal solution and a local optimal solution;
the optimal solution updating module is used for carrying out local search by adopting a local search algorithm based on a greedy reference iterative algorithm, re-determining a local optimal solution, and if the re-determined local optimal solution is superior to a global optimal solution, selecting the re-determined local optimal solution to update the global optimal solution;
an initialization module for ordering xi＝1，i＝1；
A neighborhood change module to perform a neighborhood structure change Ni(xi)；
The optimal solution updating module is also used for carrying out local search by using a local search algorithm based on a greedy reference iterative algorithm to obtain a local optimal solution; if the local optimal solution is superior to the global optimal solution, replacing the global optimal solution by the local optimal solution;
the initialization module is further used for enabling xi＝xi+1, i ═ i + 1; if i is less than or equal to 2, triggering the neighborhood change module; otherwise, setting i to be 1, and triggering an optimal solution output module;
an optimal solution output module for outputting the optimal solution at x2≤xmaxIf yes, returning to the neighborhood change module; otherwise, outputting a local optimal solution and a global optimal solution;
the optimal decoding module is used for decoding according to the local optimal solution and the global optimal solution and then optimizing the production scheduling of the aluminum ingot hot processing problem;
local search is carried out by using a greedy-based reference iterative algorithm, and a local optimal solution is obtained by the method comprising the following steps:
10) Setting maximum iteration times xi of local search disturbance mechanism based on greedy reference iteration algorithm and minimum number q of extraction positions of reference greedy iteration stageminAnd maximum value qmax；
20) Randomly generating an initial solution or generating a solution by neighborhood structure change; the initial solution or a solution generated by the neighborhood structure change is a decimal array with the length of n + m-1 and the numerical value of 0-1;
30) decoding the solution, batching workpieces on each machine according to an LSPT rule, and calculating the fitness Fit;
40) randomly extracting a value at any position in the solution, and then inserting the value into the optimal position in the solution;
50) let q be qmin；
60) Let η equal to 1;
70) setting the number of values of the further extracted positions as q, randomly extracting the values of the q positions in the solution, and enabling the solution after extraction to be a partial solution pipartialLet δ be 1;
80) reinserting the δ -th decimated value in turn to the optimal position in the partial solution s]；PS[s]Indicates the position [ s ]]A value of (d) above;
90) PS' stands for and partially resolves PS[s]The same value, PS' will solve optimally pibestThe device is divided into a left part and a right part as reference;
100) let δ be δ +1, if δ > q, a perturbation mechanism is completed, and a new solution pi is obtainednewGo to step 110); otherwise, returning to the step 80);
110) let η be η +1, if η ≦ ξ, return to step 70); otherwise go to step 120);
120) if xi times of iteration still does not obtain a better solution, making q ═ q +1, go to step 130); otherwise, terminating the iteration;
130) if q is less than or equal to qmaxThen returning to step 60); otherwise, terminating the iteration;
the solution change in the local search algorithm based on the greedy reference iterative algorithm comprises the following steps:
1a) randomly generating an initial solution or a result of a neighborhood structure changeGenerate a solution of piinitial；
2a) Performing a local search to find a locally optimal solution and then replacing the initial solution by itinitial；
3a) If piinitialIs randomly generated, let piincumbent＝πinitial，πbest＝πinitialAnd q ═ qmin(ii) a If piinitialIs generated by the change of the neighborhood structure, let piincumbent＝πinitial，q＝qmin；
4a) Let θ be 1;
5a) performing a destructive process by applying at piincumbentRandomly extracting the values of q positions, the remaining sequence producing a partial solution of pipartial；
6a) Performing a reconstruction process by referencing the optimal solution pibestInserting the values of the q positions into the partial solution pi in turnpartialTo obtain a new solution pinew；
7a) If pinewIs less than piincumbentThe fitness of (a) is thenincumbentIs replaced by pinew(ii) a If pinewIs less than pibestThe fitness of (a) is thenbestIs replaced by pinew；
8a) Let θ be θ + 1; if theta is less than or equal to xi, returning to the step 5 a); otherwise go to step 9 a);
9a) if not, get pi-less than the initial solutioninitialA better solution, let q be q +1, go to step 10 a);
10a) if q is less than or equal to qmaxAnd returning to the step 4 a); otherwise, the iteration is terminated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811062143.XA CN109255484B (en) | 2018-09-12 | 2018-09-12 | Data-driven discrete manufacturing resource collaborative optimization method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811062143.XA CN109255484B (en) | 2018-09-12 | 2018-09-12 | Data-driven discrete manufacturing resource collaborative optimization method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109255484A CN109255484A (en) | 2019-01-22 |
| CN109255484B true CN109255484B (en) | 2021-09-14 |
Family
ID=65046765
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811062143.XA Active CN109255484B (en) | 2018-09-12 | 2018-09-12 | Data-driven discrete manufacturing resource collaborative optimization method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109255484B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113112130B (en) * | 2021-03-23 | 2022-09-30 | 合肥工业大学 | High-end equipment manufacturing process quality on-line monitoring method and system |
| CN113283785A (en) * | 2021-06-09 | 2021-08-20 | 广东工业大学 | Cooperative scheduling optimization method for multi-task manufacturing resources |
| CN115081755B (en) * | 2022-08-19 | 2022-12-09 | 合肥工业大学 | Production and maintenance cooperative scheduling method and system based on variable neighborhood search algorithm |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105956689A (en) * | 2016-04-21 | 2016-09-21 | 合肥工业大学 | Transportation and production coordinated scheduling method based on improved particle swarm optimization |
| CN107392402A (en) * | 2017-09-11 | 2017-11-24 | 合肥工业大学 | Production and transport coordinated dispatching method and system based on modified Tabu search algorithm |
| CN107590603A (en) * | 2017-09-11 | 2018-01-16 | 合肥工业大学 | Based on the dispatching method and system for improving change neighborhood search and differential evolution algorithm |
| CN107703900A (en) * | 2017-11-13 | 2018-02-16 | 浙江大学 | A kind of efficient Optimization Scheduling |
| CN108053119A (en) * | 2017-12-15 | 2018-05-18 | 兰州理工大学 | A kind of Modified particle swarm optimization algorithm for solving zero-waiting Flow Shop Scheduling |
-
2018
- 2018-09-12 CN CN201811062143.XA patent/CN109255484B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105956689A (en) * | 2016-04-21 | 2016-09-21 | 合肥工业大学 | Transportation and production coordinated scheduling method based on improved particle swarm optimization |
| CN107392402A (en) * | 2017-09-11 | 2017-11-24 | 合肥工业大学 | Production and transport coordinated dispatching method and system based on modified Tabu search algorithm |
| CN107590603A (en) * | 2017-09-11 | 2018-01-16 | 合肥工业大学 | Based on the dispatching method and system for improving change neighborhood search and differential evolution algorithm |
| CN107703900A (en) * | 2017-11-13 | 2018-02-16 | 浙江大学 | A kind of efficient Optimization Scheduling |
| CN108053119A (en) * | 2017-12-15 | 2018-05-18 | 兰州理工大学 | A kind of Modified particle swarm optimization algorithm for solving zero-waiting Flow Shop Scheduling |
Non-Patent Citations (3)
| Title |
|---|
| 两阶段混合流水车间调度问题研究;胡璐璐;《中国优秀硕士学位论文电子期刊网 工程科技II辑》;20150915;全文 * |
| 基于连续批加工的生产运输协同调度研究;裴军;《中国博士学位论文电子期刊网 信息科技辑》;20150715;全文 * |
| 混合变邻域和声搜索的独立任务调度问题研究;姜华，包云，刘彦秀等;《计算机工程与设计》;20131016;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109255484A (en) | 2019-01-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109255484B (en) | Data-driven discrete manufacturing resource collaborative optimization method and system | |
| CN107862411B (en) | Large-scale flexible job shop scheduling optimization method | |
| CN109359768B (en) | Data processing method, server and electronic equipment | |
| CN113838536B (en) | Translation model construction method, product prediction model construction method and prediction method | |
| Meyerhenke et al. | Drawing large graphs by multilevel maxent-stress optimization | |
| CN105373845A (en) | Hybrid intelligent scheduling optimization method of manufacturing enterprise workshop | |
| CN107831740A (en) | A kind of Optimization Scheduling during the distributed manufacturing applied to notebook part | |
| CN111007823B (en) | Flexible job shop dynamic scheduling method and device | |
| CN105976421A (en) | Rendering program online optimization method | |
| CN111126707B (en) | Energy consumption equation construction and energy consumption prediction method and device | |
| CN114936681A (en) | Carbon emission prediction method based on deep learning | |
| CN109214695B (en) | High-end equipment research, development and manufacturing cooperative scheduling method and system based on improved EDA | |
| Mousavi et al. | Bi-objective scheduling for the re-entrant hybrid flow shop with learning effect and setup times | |
| CN110928261B (en) | Distributed estimation scheduling method and system for distributed heterogeneous flow shop | |
| CN106326005A (en) | Automatic parameter tuning method for iterative MapReduce operation | |
| CN111985162B (en) | Deep learning-based replacement flow shop control method and system | |
| Kim et al. | Scheduling 3D printers with multiple printing alternatives | |
| CN115455341B (en) | Solving method for raw material blanking layout | |
| CN113297185A (en) | Feature derivation method and device | |
| CN107678411B (en) | A kind of modeling method of uncorrelated parallel machine hybrid flow shop scheduling | |
| CN115826530A (en) | Job shop batch scheduling method based on D3QN and genetic algorithm | |
| CN107730065B (en) | Based on the production scheduling method and system for improving variable neighborhood search algorithm | |
| CN115456268A (en) | Guide roller manufacturing resource optimal allocation method, device, equipment and medium | |
| CN116468137A (en) | Distributed process planning and workshop scheduling integrated optimization method | |
| CN110705650B (en) | Sheet metal layout method based on deep learning |
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 |