CN112561225B - Flexible job shop scheduling method based on marker post co-evolution algorithm - Google Patents
Flexible job shop scheduling method based on marker post co-evolution algorithm Download PDFInfo
- Publication number
- CN112561225B CN112561225B CN202011056374.7A CN202011056374A CN112561225B CN 112561225 B CN112561225 B CN 112561225B CN 202011056374 A CN202011056374 A CN 202011056374A CN 112561225 B CN112561225 B CN 112561225B
- Authority
- CN
- China
- Prior art keywords
- population
- marker post
- equipment
- individuals
- 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
- 238000000034 method Methods 0.000 title claims abstract description 147
- 239000003550 marker Substances 0.000 title claims abstract description 60
- 230000035772 mutation Effects 0.000 claims description 12
- 238000011423 initialization method Methods 0.000 claims description 8
- 238000012163 sequencing technique Methods 0.000 claims description 6
- 108090000623 proteins and genes Proteins 0.000 claims description 5
- 238000009826 distribution Methods 0.000 claims description 4
- 238000005070 sampling Methods 0.000 claims description 4
- 230000002860 competitive effect Effects 0.000 claims description 3
- 108091081062 Repeated sequence (DNA) Proteins 0.000 claims description 2
- 230000002068 genetic effect Effects 0.000 abstract 1
- 238000004519 manufacturing process Methods 0.000 description 13
- 238000010586 diagram Methods 0.000 description 9
- 210000000349 chromosome Anatomy 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002759 chromosomal effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06316—Sequencing of tasks or work
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- 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—Systems or methods specially adapted for 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
Abstract
The invention discloses a flexible job shop scheduling method based on a marker post co-evolution algorithm, which can further improve the solving precision and the calculating efficiency of a flexible job shop scheduling problem and obtain a scheduling scheme with higher quality. In the invention, in the framework of a general genetic algorithm, a 'marker post' individual is introduced, the 'marker post' and a population are relatively independent and co-develop and evolve, the parallelism of global search and local search of the algorithm is realized, the population is advanced into global search, the problem of large-scale search is solved, and the goal is to search the current global better solution; the evolution of the marker post solves the problem of local searching of the current optimal extremum, and the small-range searching is processed, so that the global optimal solution is positioned in the global optimal solution.
Description
Technical Field
The invention relates to a job shop scheduling technology, in particular to a flexible shop job scheduling method, and specifically relates to a flexible shop scheduling method based on a marker post co-evolution algorithm.
Background
Planning production scheduling is an important task in the job shop production scheduling preparation link. Based on the existing manufacturing resources and mature technology, the processing equipment of the working procedures is selected, the processing sequence of the working procedures on the equipment is arranged, the workshop production operation plan of the equipment level is obtained, subsequent materials are guided to be prepared, distributed, processed and other works, and the method is the basis of workshop production activities. The reasonable operation plan can orderly arrange operation activities, so as to minimize resource conflict or resource waste caused by limited resources. However, in the current large number of manufacturing workshops, a more traditional manual scheduling method is still adopted, and scheduling is performed based on personal experience of a planner. Because the scheduling problem of workshops is a typical combination optimization problem, a better scheduling scheme is difficult to obtain by manual scheduling depending on experience, equipment conflict and equipment idle conditions are quite easy to occur, a great deal of waiting waste is caused, the production efficiency is reduced, and the production capacity of a manufacturing system cannot be fully exerted. Therefore, the advanced production scheduling technology is imperative to replace manual production scheduling.
In recent years, industrial intelligence has wide application in various links of various levels of a production system, such as equipment operation optimization, auxiliary quality analysis, irregular object sorting and the like. Among these, the application of industrial intelligence to production scheduling is also an important direction. The advanced heuristic algorithm is adopted to replace manual scheduling, and the optimal or relatively optimal scheduling plan is obtained under the acceptable time cost, so that the method has very important practical significance for improving the equipment utilization rate, reducing conflict and idle, further releasing the productivity, reducing the cost and improving the enterprise competitiveness. The quality of the scheduling scheme is obtained by adopting a heuristic algorithm and depends on the performance of the algorithm. Therefore, by developing an advanced heuristic algorithm with better performance and higher solving efficiency, the method is a key for obtaining a better workshop scheduling scheme.
Disclosure of Invention
The invention discloses a marker post co-evolution algorithm for solving a flexible workshop scheduling problem. According to the method, the 'marker post' individual is introduced to represent the optimal individual obtained by selection, crossover and variation evolution of each generation of population, and the optimal individual is finely searched around the 'marker post' individual through local search of the 'marker post'. The method can improve the solving precision and efficiency of the scheduling problem of the flexible job shop.
The technical scheme and the steps of the invention are as follows:
step 1: the problem data is entered, including the total number of workpieces, the total number of equipment, the set of process selectable equipment, and the processing time.
Step 2: and setting algorithm parameters including population scale, iteration times, crossover probability, variation probability and the like.
Step 3: the individual initialization of the marker post is responsible for the local search of the global extremum, so that the quality of the initial marker post individual needs to be ensured, and the convergence speed can be increased. And a high-quality initial target individual is obtained by adopting a mixed initialization method of two heuristic rules.
1) Determining machine selection using global selection policy
Flexible job shop scheduling problems require first determining the machine arrangement of the process. And a global selection strategy is adopted, so that the task load of each device is reasonably balanced in the whole scheduling period, the task distribution balance is ensured, and a relatively high-quality machine distribution scheme is obtained.
2) Heuristic alignment procedure using Shortest Processing Time (SPT)
When the process is arranged, the arrangement is performed in the order from the first process to the last process of each workpiece. The arrangement sequence of the same working procedure of different workpieces is determined by adopting SPT heuristic rules. The determination method comprises the following steps: the processing time of each working procedure to be arranged is calculated, the arrangement order is determined according to the processing time, and the shorter the processing time is, the more the arrangement is preferentially. The procedure sequence acquisition flow is as follows:
sub-step 1: the first working procedure of taking out all the workpieces is put into a working procedure set to be scheduled;
sub-step 2: if the number of the working procedures in the working procedure set to be scheduled is more than 1, a rotor step 3 is carried out, otherwise, a rotor step 4 is carried out;
sub-step 3: scheduling priority ordering is carried out on the procedures in the procedure set to be scheduled by adopting SPT rules;
sub-step 4: sequentially taking out the working procedures from the working procedure set to be scheduled one by one to be scheduled on corresponding equipment, firstly checking whether insertable idle time exists during the working procedure scheduling, and preferentially inserting the idle time, otherwise, scheduling the last working procedure of the current equipment; so as to finish the arrangement of each working procedure in the working procedure set to be scheduled;
sub-step 5: taking the next working procedure of each workpiece, putting the next working procedure into a working procedure set to be scheduled, if each working procedure of each workpiece is scheduled, performing a rotor step 6, otherwise, performing a rotor step 2 to continue the scheduling;
sub-step 6: and finishing the sequence of the working procedures to obtain the arrangement of the working procedures.
Step 4: population initialization, wherein population individuals are responsible for global search, and population diversity and coverage solution space uniformity are required to be ensured. Aiming at equipment arrangement and procedure ordering of initial population, two random initialization methods are respectively provided:
1) Method for initializing sequence of non-repeated sequence arrangement
Generating an initial sequence of steps arranged in a first step of a first workpiece to a last step of a last workpiece; then generating an index string group which is equal to the initial procedure string part in length and equal to the initial population in number, wherein the positions of all elements of each index string in the index string group are not repeated; and finally, rearranging the initial process strings by adopting each index string, so that initial individuals with non-repeated process arrangement can be obtained, and population diversity is ensured.
2) Machine selection initialization method for loop machine non-return sampling
When an initial individual device string is generated, randomly selecting a device from the optional device set of each process as the processing device of the current process, temporarily removing the device from the optional device set, extracting the processing device in the same way when the next initial individual device string is arranged in the process, and restoring the optional device set to the initial device set state when the optional device set is empty, so that the cycle is performed, and ensuring that the arranged times of each device are uniform.
Step 5: and decoding the scheduling solutions of the marker post individuals and the population individuals, and calculating an objective function value of each scheduling solution, wherein the objective function value is the maximum finishing time of the scheduling solution.
Step 6: judging whether the exit condition is satisfied, namely: the maximum iteration number setting is satisfied, and if the condition is satisfied, the optimal scheduling solution is output, namely: the solution with the smallest maximum finishing time; otherwise, continuing to execute the step 7.
Step 7: and (3) judging the disturbance of the marker post, namely judging whether the marker post falls into local optimum, and if so, turning to the step (8), otherwise turning to the step (10).
Step 8: to avoid losing the global optimum benchmarking target by disturbance, the current benchmarking individuals and targets are recorded through a benchmarking baseline prior to performing benchmarking.
Step 9: and performing a target disturbance operation, and selecting a global preferred solution which is different from the current target from the population to replace the current target individual.
Step 10: and (3) carrying out local search on the individual benchmarks, and searching better scheduling solutions around the global better solutions. The local search flow of the marker post is as follows:
sub-step 1: selecting the fastest equipment to replace a search strategy, randomly selecting a certain number of working procedures, and for each selected working procedure, if the equipment is not fastest processing equipment, replacing the current processing equipment with the fastest processing equipment;
sub-step 2: the process sequencing adopts a sequencing combination neighborhood searching strategy, under the condition that the selection of the equipment is kept unchanged, a plurality of processes are randomly selected, then all the individuals formed by sequencing and combining are generated, finally the target value of each individual is evaluated, and the best individual is selected as the target marker post individual for searching.
Step 11: and starting population evolution, firstly executing population selection operation, and adopting a multi-element competitive bidding competition selection mode with better performance. And randomly selecting a certain number of individuals from the population, then selecting the optimal individuals, and repeatedly executing the operation until the set population scale is met.
Step 12: and executing cross operation, randomly selecting two parents in the population, and crossing when the cross probability is met, wherein the process string is crossed in a linear order, and the equipment string is uniformly crossed based on workpieces in order to ensure that the infeasible solution does not appear in the cross.
Step 13: and executing mutation operation, namely sequentially selecting each individual in the population, executing mutation operation when the individual meets mutation probability, wherein the sequence variation adopts reverse sequence variation, and the equipment string variation adopts random equipment to replace variation.
Step 14: co-evolution operations, namely: population and benchmarking individuals share evolutionary effort. Co-evolution comprising two aspects:
1) Post collaborative update
Comparing the current generation of the marker post individuals with the current population optimal individuals, and if the population optimal individuals are better than the marker post individuals, replacing the marker post individuals by the population optimal individuals;
2) Marker post co-population evolution
The marker post individuals and the local optimal values searched in the local search are added into the population (whether the marker post individuals are repeated with the population individuals or not is needed to be judged before the marker post individuals are added), and the relatively good individual genes are introduced, so that the population quality is improved, and the evolution of the population is accelerated.
Step 15: turning to step 5.
The invention has the beneficial effects that:
the algorithm can effectively exert the global searching capability of the common evolutionary algorithm through the co-evolution strategy of the 'benchmarking' individuals and the population, and searches the global better solution in the whole solution space; and then, the local search of a 'marker post' individual is adopted to realize the fine search of possible optimal solutions around the global optimal solution, so that the problem of low solving precision caused by poor local searching capability of a common evolutionary algorithm is solved. The convergence can be accelerated through the cooperation of global searching of the population and local searching of the marker post, the solving precision is improved, and a higher-quality scheduling scheme is obtained.
Drawings
FIG. 1 is a block diagram of an algorithm framework and overall flow chart of the present invention
FIG. 2 is a diagram showing an example of the problem of the present invention
FIG. 3 is a diagram showing chromosome expression patterns in the present invention
FIG. 4 is a diagram showing the method for initializing the process string of the winning-pole unit of the present invention
FIG. 5 is a diagram of a method for initializing a string of individual devices of a winning pole in the present invention
FIG. 6 is a diagram showing the operation of searching for part of the string of individual process steps of the present invention
FIG. 7 is a cross-operating chart of a chromosome process string in the present invention
FIG. 8 is a cross-operating diagram of a chromosome apparatus string according to the present invention
FIG. 9 is a diagram showing variation of a chromosome sequence in the present invention
FIG. 10 is a diagram showing variation operation of chromosome apparatus string according to the present invention
FIG. 11 is a scheduling DeGantt chart of an example of a problem of the present invention
Detailed Description
The invention will be further described with reference to the drawings and examples.
The invention discloses a flexible job shop scheduling method based on a marker post co-evolution algorithm, which solves the problem of flexible job shop scheduling by co-evolution of marker post individuals and populations, wherein the algorithm flow is shown in figure 1. An example of the problem shown in fig. 2 will now be described.
Step 1: the basic data of the problem is input, including 5 workpieces, 6 devices and the processing time of each device for the corresponding working procedure, see in particular fig. 2.
Step 2: setting algorithm parameters: population scale 100, crossover probability 0.8, mutation probability 0.1, and iteration times 200.
Step 3: an initialization benchmarking individual is generated. Chromosomal expression of the marker individuals and the population individuals is shown in FIG. 3. Initializing a target individual, respectively obtaining a machine string and a process string by adopting a global selection strategy and an SPT rule, and then combining the machine string and the process string to obtain the initial target individual.
Step 4: the group individual initialization adopts a procedure ordering initialization method of non-repeated procedure arrangement and a machine selection initialization method of sampling without returning a circulating machine to initialize procedure strings and equipment strings respectively, and the steps are as follows:
1) Generating an initial process string according to the workpiece sequence;
2) Generating index string groups with the same length as the process string, different element positions and 100 number;
3) Arranging the initial process strings by adopting each index string to obtain 100 initial individual process strings, as shown in fig. 4;
4) Initializing a first individual machine string, starting from a first position (a first procedure) machine, collecting one machine from a selectable machine set of the procedure by adopting random non-return sampling, and sequentially completing machine extraction of each procedure;
5) Judging whether the optional machine set of each procedure is empty or not, and if so, resetting the machine set to an initial machine set state;
6) Initializing a machine string of a second individual according to the method of 4), judging according to the requirement of 5), and sequentially completing equipment strings of 100 initial individuals;
7) The 100 initial equipment strings and the process strings were combined to obtain 100 initial individuals, as shown in fig. 5.
Step 5: and respectively decoding the marker post individuals and 100 population individuals, and calculating objective function values of the marker post individuals and the population individuals.
Step 6: judging whether the iteration times are equal to the set iteration times, if so, turning to the step 11; otherwise, continuing to execute the step 7.
Step 7: judging whether the objective function value of the current substitute marker post is equal to the objective function value of the previous substitute, if so, adding 1 to the counter; if the counter is equal to 2, the marker post is continuously unchanged for 3 generations, and marker post disturbance is needed, and the step 8 is executed, otherwise, the step 10 is executed.
Step 8: the current benchmarking individual and target values are recorded in the benchmarking baseline.
Step 9: the optimal individual is taken out from 100 individuals in the current population, and if the individual is different from the current target individual, the individual is replaced with the current target individual; otherwise, continuously taking out the suboptimal individuals from the population until the current marker post individuals are replaced, and then resetting the counter.
Step 10: the local search is carried out on the current marker post individual, and the steps are as follows:
1) Randomly selecting 3-5 working procedures, replacing the processing equipment of the selected working procedures with equipment with the fastest processing time in the selectable equipment set, and replacing the equipment number of the corresponding position of the corresponding working procedure of the machine string as the equipment number with the fastest processing time;
2) Keeping the machine string unchanged, randomly selecting 3 processes, and generating 6 process strings formed by 6 kinds of arrangement combinations based on the 3 selected processes, as shown in fig. 6;
3) The 6 process strings are respectively combined with the same machine string to form 6 different individuals;
4) Comparing target values of 6 different individuals, and selecting the optimal individual as a result of the local search of the target rod, namely: the next generation of individuals with benchmarks.
Step 11: executing competitive race selection, randomly selecting 3 individuals from the population, comparing objective function values of the 3 individuals, and selecting better individuals until 100 individuals are selected;
step 12: randomly selecting 2 individuals in the population to generate random decimal numbers, if the random decimal numbers are smaller than the crossover probability, carrying out crossover operation on two parents, and crossing a process string and a device string by adopting linear order crossover and uniform crossover based on workpieces respectively:
1) Linear order crossover
As shown in fig. 7, two random positions, 4 and 6, are generated; then exchanging chromosome strings between random positions (4, 5,6 positions) of the two parent individuals, and deleting genes exchanged from the other parent individual in the primary parent individuals; finally, filling the residual genes outside the two crossing positions from the first gene position.
2) Uniform intersection based on workpieces
As shown in fig. 8, 1 work piece, J2, was randomly selected, and the machine numbers of all the work piece J2 steps of the two work pieces of the parent individual were interchanged.
Step 13: and selecting individuals from the population in turn, generating random decimal numbers, and if the random decimal numbers are smaller than the mutation probability, performing mutation operation on the selected individuals. The reverse sequence mutation and the random equipment replacement mutation are adopted to carry out mutation operation on the process string and the equipment string respectively:
1) Reverse sequence variation of process sequence
As shown in fig. 9, two positions, 4 and 7, are randomly selected, and elements between the two positions (4, 5,6,7 positions) are reordered in reverse order;
2) Device string random device replacement variation
As shown in fig. 10, a process is randomly selected and the process is performed by the current processing equipment M 1 Replaced by dividing by the currentAny one device except the device, M in the figure 3 。
Step 14: and co-evolution operation of the marker post and the population. The method comprises the following steps:
1) The current population optimal individuals are taken out and compared with target values of the target individuals, if the current population optimal individuals are due to the target individuals, the current population optimal individuals are adopted to replace the current target individuals, otherwise, the target is maintained unchanged;
2) Comparing target values of 6 individuals generated in the neighborhood search of the winning pole procedure string in the step 10 with the individuals 10% in front of the current population, and adding the individuals generated in the neighborhood search into the population if the individuals are better than the individuals 10% in front of the current population;
3) The populations were ranked according to the target value, taking the first 100 individuals as the next generation population.
Step 15 goes to step 5.
Step 16, comparing the current-generation marker post individual with the marker post historical individual recorded by the marker post baseline, outputting the individual with the smallest objective function value as a scheduling solution, and drawing a scheduling Gantt chart to represent a scheduling scheme, as shown in fig. 11.
Claims (6)
1. A flexible job shop scheduling method based on a marker post co-evolution algorithm adopts the co-evolution of marker post individuals and common populations to respectively perform local and global searches; the method is characterized in that: the method comprises the following steps:
step 1: inputting problem data, including the total number of workpieces, the total number of equipment, a procedure selectable equipment set and processing time;
step 2: setting algorithm parameters including population scale, iteration times, cross probability and variation probability;
step 3: the individual marker post is initialized, and the individual marker post is responsible for local searching of the global extremum, so that the quality of the individual marker post needs to be ensured initially, and the convergence speed can be increased;
step 4: initializing a population, wherein the population individuals are responsible for global searching, and the diversity of the population and the uniformity of coverage solution space are required to be ensured;
step 5: decoding the scheduling solutions of the marker post individuals and the population individuals, and calculating an objective function value of each scheduling solution, wherein the objective function value is the maximum finishing time of the scheduling solution;
step 6: judging whether the exit condition is satisfied, namely: the maximum iteration number setting is satisfied, and if the condition is satisfied, the optimal scheduling solution is output, namely: the solution with the smallest maximum finishing time; otherwise, continuing to execute the step 7;
step 7: judging whether the target is in local optimum or not by the disturbance of the target, wherein the judging condition is whether the target function value of the target is changed after the target is iterated for 3 times continuously, if not, the target disturbance is needed, and the step 8 is executed, otherwise, the step 10 is executed;
step 8: in order to avoid losing the global optimal target value due to disturbance, recording the current target individual and the target value through a target base line before performing the target disturbance;
step 9: performing a target disturbance operation, and selecting a global preferred solution different from the current target from the population to replace the current target individual;
step 10: performing local search on the individual benchmarks, and searching higher-quality scheduling solutions around the global higher-quality solutions;
step 11: starting population evolution, firstly executing population selection operation, and adopting a multi-element competitive bidding competition selection mode with better performance; randomly selecting a certain number of individuals from the population, then selecting the optimal individuals, and repeatedly executing the operation until the set population scale is met;
step 12: executing cross operation, randomly selecting two parents in a population, and crossing when meeting cross probability, wherein in order to ensure that no infeasible solution appears in the cross, the process string cross adopts linear order cross, and the equipment string adopts uniform cross based on workpieces;
step 13: performing mutation operation, namely sequentially selecting each individual in the population, and performing mutation operation when the individual meets mutation probability, wherein the sequence variation adopts reverse sequence variation, and the equipment string variation adopts random equipment to replace variation;
step 14: co-evolution operations, namely: the population and the marker post individuals share the evolution result;
step 15: turning to step 5.
2. The flexible job shop scheduling method based on the marker post co-evolution algorithm according to claim 1, wherein the method comprises the following steps: adopting a mixed initialization method of two heuristic rules to obtain a high-quality initial target individual;
1) Determining machine selection using global selection policy
The flexible job shop scheduling problem requires first determining the machine arrangement of the process; the overall selection strategy is adopted, so that the task load of each device is reasonably balanced in the whole scheduling period, the task distribution balance is ensured, and a relatively high-quality machine distribution scheme is obtained;
2) Adopts SPT heuristic rule arrangement procedure with shortest processing time
When the working procedures are arranged, the working procedures are arranged according to the sequence from the first working procedure to the last working procedure of each workpiece; the arrangement sequence of the same working procedure of different workpieces is determined by adopting SPT heuristic rules; the determination method comprises the following steps: the processing time of each working procedure to be arranged is calculated, the arrangement order is determined according to the processing time, and the shorter the processing time is, the more the arrangement is preferentially.
3. The flexible job shop scheduling method based on the marker post co-evolution algorithm according to claim 2, wherein the method comprises the following steps: the SPT heuristic arrangement procedure flow is as follows:
sub-step 1: the first working procedure of taking out all the workpieces is put into a working procedure set to be scheduled;
sub-step 2: if the number of the working procedures in the working procedure set to be scheduled is more than 1, a rotor step 3 is carried out, otherwise, a rotor step 4 is carried out;
sub-step 3: scheduling priority ordering is carried out on the procedures in the procedure set to be scheduled by adopting SPT rules;
sub-step 4: sequentially taking out the working procedures from the working procedure set to be scheduled one by one to be scheduled on corresponding equipment, firstly checking whether insertable idle time exists during the working procedure scheduling, and preferentially inserting the idle time, otherwise, scheduling the last working procedure of the current equipment; so as to finish the arrangement of each working procedure in the working procedure set to be scheduled;
sub-step 5: taking the next working procedure of each workpiece, putting the next working procedure into a working procedure set to be scheduled, if each working procedure of each workpiece is scheduled, performing a rotor step 6, otherwise, performing a rotor step 2 to continue the scheduling;
sub-step 6: and finishing the sequence of the working procedures to obtain the arrangement of the working procedures.
4. The flexible job shop scheduling method based on the marker post co-evolution algorithm according to claim 1, wherein the method comprises the following steps: aiming at equipment arrangement and procedure ordering of initial population, two random initialization methods are respectively provided:
1) Method for initializing sequence of non-repeated sequence arrangement
Generating an initial sequence of steps arranged in a first step of a first workpiece to a last step of a last workpiece; then generating an index string group which is equal to the initial procedure string part in length and equal to the initial population in number, wherein the positions of all elements of each index string in the index string group are not repeated; finally, rearranging the initial process strings by adopting each index string to obtain initial individuals with non-repeated process arrangement, thereby ensuring population diversity;
2) Machine selection initialization method for loop machine non-return sampling
The process is as follows: when a certain initial individual equipment string is generated, randomly selecting one piece of equipment from the optional equipment set of each procedure as processing equipment of the current procedure, temporarily removing the equipment from the optional equipment set, extracting the processing equipment in the same way when the next initial individual equipment string is arranged in the procedure, and restoring the optional equipment set to an initial equipment set state when the optional equipment set is empty; this process is repeated until it is ensured that each device is arranged a uniform number of times.
5. The flexible job shop scheduling method based on the marker post co-evolution algorithm according to claim 1, wherein the method comprises the following steps: the local search flow of the marker post is as follows:
sub-step 1: selecting the fastest equipment to replace a search strategy, randomly selecting a certain number of working procedures, and for each selected working procedure, if the equipment is not fastest processing equipment, replacing the current processing equipment with the fastest processing equipment;
sub-step 2: the process sequencing adopts a sequencing combination neighborhood searching strategy, under the condition that the selection of the equipment is kept unchanged, a plurality of processes are randomly selected, then all the individuals formed by sequencing and combining are generated, finally the target value of each individual is evaluated, and the best individual is selected as the target marker post individual for searching.
6. The flexible job shop scheduling method based on the marker post co-evolution algorithm according to claim 1, wherein the method comprises the following steps: step 14 comprises co-evolution of two aspects:
1) Post collaborative update
Comparing the current generation of the marker post individuals with the current population optimal individuals, and if the population optimal individuals are better than the marker post individuals, replacing the marker post individuals by the population optimal individuals;
2) Marker post co-population evolution
The marker post individuals and the local optimal values searched in the local search are added into the population, and the relatively good individual genes are introduced, so that the population quality is improved, and the evolution of the population is accelerated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011056374.7A CN112561225B (en) | 2020-09-30 | 2020-09-30 | Flexible job shop scheduling method based on marker post co-evolution algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011056374.7A CN112561225B (en) | 2020-09-30 | 2020-09-30 | Flexible job shop scheduling method based on marker post co-evolution algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112561225A CN112561225A (en) | 2021-03-26 |
| CN112561225B true CN112561225B (en) | 2023-11-07 |
Family
ID=75042690
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011056374.7A Active CN112561225B (en) | 2020-09-30 | 2020-09-30 | Flexible job shop scheduling method based on marker post co-evolution algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112561225B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113222450B (en) * | 2021-05-27 | 2022-09-13 | 中车青岛四方机车车辆股份有限公司 | Scheduling method and device based on deformation of operation unit, electronic equipment and medium |
| CN113627716A (en) * | 2021-06-30 | 2021-11-09 | 上海大学 | Anti-interference vehicle scheduling method for multiple distribution centers |
| CN115032959A (en) * | 2022-07-29 | 2022-09-09 | 京东方科技集团股份有限公司 | Production scheduling method and device for flexible job shop |
| CN115330283B (en) * | 2022-10-17 | 2023-03-24 | 中汽数据(天津)有限公司 | Workshop-based and production line-based vehicle production sequence optimization method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101668348A (en) * | 2009-10-10 | 2010-03-10 | 曾超宁 | Wind-power wireless sensor network information management system for mine |
| CN106611379A (en) * | 2016-02-22 | 2017-05-03 | 四川用联信息技术有限公司 | Improved culture gene algorithm for solving multi-objective flexible job shop scheduling problem |
| CN110225096A (en) * | 2019-05-20 | 2019-09-10 | 电子科技大学 | Server load balancing method based on genetic algorithm |
| CN111311077A (en) * | 2020-01-21 | 2020-06-19 | 南京航空航天大学 | Method for solving flexible job shop scheduling by particle swarm algorithm based on longicorn stigma search |
| CN111325443A (en) * | 2020-01-21 | 2020-06-23 | 南京航空航天大学 | Method for solving flexible job shop scheduling by improved genetic algorithm based on catastrophe mechanism |
-
2020
- 2020-09-30 CN CN202011056374.7A patent/CN112561225B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101668348A (en) * | 2009-10-10 | 2010-03-10 | 曾超宁 | Wind-power wireless sensor network information management system for mine |
| CN106611379A (en) * | 2016-02-22 | 2017-05-03 | 四川用联信息技术有限公司 | Improved culture gene algorithm for solving multi-objective flexible job shop scheduling problem |
| CN110225096A (en) * | 2019-05-20 | 2019-09-10 | 电子科技大学 | Server load balancing method based on genetic algorithm |
| CN111311077A (en) * | 2020-01-21 | 2020-06-19 | 南京航空航天大学 | Method for solving flexible job shop scheduling by particle swarm algorithm based on longicorn stigma search |
| CN111325443A (en) * | 2020-01-21 | 2020-06-23 | 南京航空航天大学 | Method for solving flexible job shop scheduling by improved genetic algorithm based on catastrophe mechanism |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112561225A (en) | 2021-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112561225B (en) | Flexible job shop scheduling method based on marker post co-evolution algorithm | |
| CN107301473B (en) | Similar parallel machine based on improved adaptive GA-IAGA batch dispatching method and system | |
| CN111966050A (en) | AMMAS-GA nesting algorithm-based dual-resource mold job shop scheduling optimization method | |
| CN112580922A (en) | Flexible job shop scheduling method based on multilevel neighborhood structure and hybrid genetic algorithm | |
| CN107451747B (en) | Workshop scheduling system based on self-adaptive non-dominated genetic algorithm and working method thereof | |
| CN101901425A (en) | Flexible job shop scheduling method based on multi-species coevolution | |
| CN113805545B (en) | Flexible flow shop combined scheduling rule generation method considering batch processing | |
| CN111798120A (en) | Flexible job shop scheduling method based on improved artificial bee colony algorithm | |
| CN106611230A (en) | Critical process-combined genetic local search algorithm for solving flexible job-shop scheduling | |
| CN115471133A (en) | Workshop comprehensive scheduling system based on order management and rolling optimization scheduling | |
| CN112381343A (en) | Flexible job shop scheduling method based on genetic-backbone particle swarm hybrid algorithm | |
| CN116466659A (en) | Distributed assembly flow shop scheduling method based on deep reinforcement learning | |
| CN110135752B (en) | Scheduling method for complete orders with switching time | |
| CN113139720B (en) | Casting shop production scheduling optimization method with learning effect | |
| Mhasawade et al. | A survey of hybrid metaheuristics to minimize makespan of job shop scheduling problem | |
| CN116757411A (en) | Scheduling method of dual-resource flexible job shop for aerospace complex components | |
| CN116011723A (en) | Intelligent dispatching method and application of coking and coking mixed flow shop based on Harris eagle algorithm | |
| CN111105164B (en) | Workshop scheduling method, device and equipment | |
| Liu et al. | An improved nondominated sorting genetic algorithm-ii for multi-objective flexible job-shop scheduling problem considering worker assignments | |
| CN113792494A (en) | Multi-target flexible job shop scheduling method based on migrating bird group algorithm and cross fusion | |
| CN113011767A (en) | Hybrid genetic method for scheduling multi-target flexible job shop | |
| CN115249113A (en) | Distributed zero-waiting flow shop scheduling method and system with preparation time | |
| CN114237166A (en) | Method for solving multi-rotating-speed energy-saving scheduling problem based on improved SPEA2 algorithm | |
| CN115700647A (en) | Workshop flexible operation scheduling method based on tabu search genetic algorithm | |
| CN117314078B (en) | Deadlock-free scheduling method of flexible manufacturing system based on Petri network and neural network |
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 |