CN105045223B - A kind of multi-products ordering production scheduling method under condition of uncertainty - Google Patents
A kind of multi-products ordering production scheduling method under condition of uncertainty Download PDFInfo
- Publication number
- CN105045223B CN105045223B CN201510279760.5A CN201510279760A CN105045223B CN 105045223 B CN105045223 B CN 105045223B CN 201510279760 A CN201510279760 A CN 201510279760A CN 105045223 B CN105045223 B CN 105045223B
- Authority
- CN
- China
- Prior art keywords
- solution
- processing time
- scheduling
- uncertainty
- product
- 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.)
- Expired - Fee Related
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- General Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
Abstract
The invention discloses a kind of multi-products ordering production scheduling methods under processing time condition of uncertainty.Include that the product treatment time in production process is described using the method for probability distribution；It is desired for target with Maximal Makespan and establishes the Random Expected Value Model based on Monte Carlo simulation；Stochastic model is effectively solved using the group searching algorithm of discrete codes, obtains the batch process production scheduling scheme under condition of uncertainty.The present invention provides effective method, adaptable feature wide, validity is good, reliability is high for the uncertain scheduling of batch process processing time：Using the desired value of stochastic simulation, the problem that can be coped under various random distributions solves；Using the group searching algorithm of discrete codes, it is ensured that the high efficiency that scheduling solution is searched；To unlimited intermediate storage tank and the batch process without intermediate storage tank type, this method can all obtain preferable production scheduling scheme in a short time.
Description
Technical field
The invention belongs to production scheduling technical fields, and in particular between the multi-product under a kind of processing time condition of uncertainty
Process of having a rest production scheduling method.
Background technology
Since batch process produces multiple products in a batch mode, equipment and the utilization rate of time are relatively low, simultaneously
The different batches of different product or identical product may propose processing request to certain equipment simultaneously in production, to occur
Conflict.Therefore process of how effectively arranging production, it is right properly according to the market demand or production task distributing equipment and time
It plays a very important role in the production efficiency and economic benefit for improving batch process, wherein the target that enterprise is concerned about very much
It is Maximal Makespan.In the actual production process, it is frequently present of the processing time uncertain situation of product so that by various
The Optimized Operation scheme performance indicator that deterministic models obtain reduces even no longer feasible.Therefore, it is necessary to when fully considering processing
Between uncertain factor, to ensure the feasibility and optimality of scheme.However, existing scheduling system can only generally cope with determination
Property model, also handles processing time uncertain situation usually as certainty.With the increase and scheduling of problem scale
The raising of scheme validity and requirement of real-time has higher to multi-products ordering scheduling solution technique under condition of uncertainty
Requirement.
As a kind of novel colony intelligence optimization algorithm, group searching algorithm is that He etc. is hunted behavior inspiration by animal collective and carried
Go out.In group searching algorithm, the individual in population plays the part of different roles according to the different division of labor of " it was found that-participation ".It is planting
In group, it is found that the individual of current optimal location is referred to as finder, participate in and the individual for sharing finder's achievement is referred to as participating in
Person.Algorithm is absorbed in local optimum in order to prevent, and group searching algorithm introduces " going around " strategy, participant is divided into follower and is patrolled
The person of patrolling, once finder is found that prey trace, follower can be quickly close to finder, while patrolman can be near group
It goes around into row stochastic.In view of group searching algorithm has higher validity and robustness when solving complicated optimum problem,
The present invention is introduced under condition of uncertainty in multi-products ordering scheduling solution, devises each ring of group searching algorithm
Section improves enterprises production efficiency to obtain preferably scheduling scheme.
Invention content
The purpose of the present invention is asked for the multi-products ordering production scheduling under product treatment time condition of uncertainty
Topic proposes a kind of dispatching method based on group searching algorithm with the target that is desired for of minimizes Maximal Makespan.
The present invention takes following technical solution：
A kind of multi-products ordering production scheduling method under condition of uncertainty, includes the following steps.
Step 1：Preparation stage：It is loaded into the processing time information and multi-products ordering of each process of each product in real time
Storage class.
Step 2：The stochastic simulation stage：According to the temporal information of each process of loading, NS are generated using inverse transformation method
It obeys the processing time of corresponding random distribution, NS is stochastic simulation frequency in sampling.
Step 3：The algorithm initialization stage：To multi-products ordering, handled according to the case where displacement, i.e., every sets
Standby upper product process sequence is the same.By the arrangement that each individual UVR exposure of algorithm is product serial number, to each equipment：Row
The product for being listed in front is first handled, and arranges posterior product post-processing.Then group searching algorithm population is initialized as follows：
One solution is generated by NEH rules, other popsize-1 solution is random to be generated（Popsize is Population Size）, utilize stochastic simulation
Each solution of data assessment.
Step 4：The phylogenetic scale of group searching algorithm：The optimization thought for continuing to use group searching algorithm, in conjunction with the neighbour of scheduling problem
The cross and variation operator of domain concept and intelligent algorithm carries out the operational design of finder, follower and patrolman（In the step,
The stochastic simulation data generated using step 2 if being related to assessment scheduling solution）.
Step 4.1：Finder acts in population preferably solution, is disturbed using optimal insertion to individual, current to obtain
Some disturbance solution around solution, is then based on this solution, carries out insertion local search, until obtaining local optimum.Optimal insertion is disturbed
Dynamic process selects several unduplicated product serial numbers, corresponding product is inserted into other positions on the basis of current solution
In optimal location on；Be inserted into local search each product is subjected to insertion neighborhood search successively, it is known that solution cannot improve again for
Only.
Step 4.2：Follower intersects corresponding individual and finder's individual, to obtain and finder's individual phase
As solve.Current solution and population preferably solution are intersected, selectable crossover operator has partially matched crossover, order to intersect,
Location-based intersection, the intersection based on order, recycling cross, linear precedence intersection etc..
Step 4.3：Patrolman primarily serves the effect gone around at random, is realized using two ways：It randomly generates and meets life
Produce the scheduling solution of constraint；Or primary larger disturbance is done on the basis of currently preferably solution.The individual that patrolman obtains will be replaced
Fall the worst person in population, that is, uses " last-one-out system ".
Step 4.4：Judge whether end condition meets, otherwise return to step 4 if being unsatisfactory for are terminated and exported current
Best scheduling scheme.
The beneficial effects of the present invention are：
（1）Using the desired value of stochastic simulation, can complete to be uniformly distributed, the problem under the various random distributions such as exponential distribution
It solves；
（2）Using the group searching algorithm of discrete codes, the solution space of continuous domain search is reduced, it is ensured that scheduling solution is searched
High efficiency；
（3）To unlimited intermediate storage tank and the batch process without intermediate storage tank type, random expected value mould that this method is established
Type is all feasible, and the algorithm solution time is shorter.
Description of the drawings
Fig. 1 is the flow chart of the uncertain batch processes scheduling method of the present invention.
Fig. 2 is that group searching algorithm coding of the present invention decodes schematic diagram.
Fig. 3 is optimal scheduling scheme Gantt chart of the present invention.
In Fig. 2 and Fig. 3, abscissa is the time, and ordinate is equipment, the digital representation product code numbering in rectangular block, square
The completion date of the digital representation corresponding product in the shape square lower right corner, makespan are the Maximal Makespan of all products.
Specific implementation mode
The present invention will be further described below in conjunction with the accompanying drawings.
The dispatching method is applied in certain interval chemical plant multi-product process by present embodiment.It considers in two kinds
Between storage mode：Unlimited intermediate storage tank and without intermediate storage tank.Unlimited intermediate storage tank type has enough when product leaves equipment
Intermediate storage tank stores, and without intermediate storage tank type completely without intermediate storage tank, therefore can only be trapped in equipment, so as to cause life
Produce the increase in period.The batch process has 5 equipment, 10 products.Uncertain processing time such as table 1, wherein N indicate normal state point
Cloth.
Table 1 does not know treatment schedule
The LPT device scheduling strategy of present embodiment includes the following steps if Fig. 1 shows.
Step 1：Preparation stage：It is loaded into the processing time information and multi-products ordering of each process of each product in real time
Storage class.
Step 2：The stochastic simulation stage：According to the temporal information of each process of loading, NS are generated using inverse transformation method
Meet the processing time of corresponding random distribution, NS is stochastic simulation frequency in sampling.
If normal distribution is, generate obey this distribution random numbers process be：
By being uniformly distributedU(0,1) it generates；
By being uniformly distributedU(0,1) it generates；
；
；
Return to x.
The process is executed NS times to the processing time of each process, you can NS sample is obtained, to obey corresponding normal state point
NS matrix of cloth.In view of problem scale and Riming time of algorithm, 50 occasion is less than for product number, takes NS=5000, if
Product number is more than 50, then suitably reduces NS to meet the requirement of problem solving real-time.
When needing to assess the effect of some scheduling solution, according to law of great number, Maximal Makespan is soughtPhase
It can be estimated with following formula when prestige：
Wherein,It is arranged for productMaximal Makespan stochastic variable,For stochastic variable
Ith sample value, NS are frequency in sampling.
If processing time is other random distributions, similar the estimation of desired value can be obtained.
Step 3：The algorithm initialization stage：To multi-products ordering, handled according to the case where displacement, i.e., every sets
Standby upper product process sequence is the same.By the arrangement that each individual UVR exposure of algorithm is product serial number, to each equipment：Row
The product for being listed in front is first handled, and arranges posterior product post-processing.Then group searching algorithm population is initialized as follows：
One solution is generated by NEH rules, other popsize-1 solution is random to be generated（Popsize is Population Size）, utilize stochastic simulation
Each solution of data assessment.Fig. 2 is coding and decoding schematic diagram, illustrates that product coding is（1, 2, 3, 4, 5, 6, 7, 8,
9, 10）Maximal Makespan of processing time corresponding operation plan Gantt chart when taking mean value, this scheduling solution is 525.
Step 4：The phylogenetic scale of group searching algorithm：The optimization thought for continuing to use group searching algorithm, in conjunction with the neighbour of scheduling problem
The cross and variation operator of domain concept and intelligent algorithm carries out the operational design of finder, follower and patrolman（In the step,
The stochastic simulation data generated using step 2 if being related to assessment scheduling solution）.
Step 4.1：Finder acts in current population preferably solution, is disturbed using optimal insertion to individual, to obtain
Some disturbance solution around current solution, is then based on this solution, carries out insertion local search, until obtaining local optimum.
In present embodiment, optimal insertion perturbation process selects three unduplicated product sequences on the basis of current solution
Number, corresponding product is inserted on the optimal location in other positions, realization process is：
Step（1）：It enables and is inserted into listILFor sky, 3 unduplicated products are randomly choosed, and be put intoIL；
Step（2）：FromILOne product of middle taking-up, find itpPosition.It willpIn the product be inserted into other n-1
Desired positions in a position are new to obtainp；
Step（3）：IfILIt is not sky, then goes to step（2）；Otherwise, terminate.
Step 4.2：Follower intersects corresponding individual and finder's individual, to obtain and finder's individual phase
As solve.Current solution and population preferably solution are intersected, present embodiment uses partially matched crossover, by what is obtained after intersection
Preferably individual and current follower's individual compare, if more excellent, current individual is replaced, and is otherwise remained unchanged.
Step 4.3：Patrolman primarily serves the effect gone around at random, it is therefore an objective to which the solution randomly generated using one is replaced
Poor individual in population.Randomly generating the scheduling solution for meeting and producing and constraining can be arranged by arbitrarily generating a random product
It obtains；It can also randomly choose some products on the basis of currently preferably solution and be inserted into other random sites, it is larger to carry out
Disturbance.Both modes are carried out according to respective 0.5 probability, and then patrolman's operation replaces the individual of generation in population
Worst person.
Step 4.4：Judge whether end condition meets, uses maximum iteration for end condition.Due to algorithm performance
Preferably, present embodiment problem scale is 20, using 50 on behalf of maximum iteration.If cycle-index less than 50, returns to step
Rapid 4.Otherwise, end loop and optimum results are exported, the optimal scheduling scheme for being scheduling to algorithm and finding obtained at this time.
For example, the optimum individual product that group searching algorithm is found after 50 iteration is arranged as（3, 1, 8, 4, 5, 6,
9, 10, 7, 2）, solution desired value is 500.11.If according to the mean value computation of processing time, operation plan Gantt chart such as Fig. 3
Show.The scheduling scheme is all shown as more excellent scheduling in many experiments test, and the run time of entire dispatching algorithm is in 0.5 milli
In second, the requirement of scheduling system real time can be met.
The above content is the explanations to a preferred embodiment of the present invention, can help those skilled in the art more fully
Understand technical scheme of the present invention.But these embodiments are merely illustrative the specific embodiment party, and it cannot be said that the present invention
Formula is only limitted to the explanation of these embodiments.
Claims (3)
1. a kind of multi-products ordering production scheduling method under processing time condition of uncertainty, it is characterised in that：This method
Uncertain processing time is described using the mode of probability distribution, establishes the random expectation using Maximal Makespan as target
It is worth model, sliding-model control has been carried out to the group searching algorithm of Filled function, institute is solved using the group searching algorithm of discrete codes
The stochastic model of foundation obtains preferably production scheduling scheme, this method and specifically includes：
Preparation stage：It is loaded into the processing time information of each process of each product and the storage class of multi-products ordering in real time
Type；
The stochastic simulation stage：According to the temporal information of each process of loading, using inverse transformation method generate NS obey correspondence with
The processing time of machine distribution, NS are stochastic simulation frequency in sampling；
The algorithm initialization stage：To multi-products ordering, handled according to the case where displacement, i.e. product in every equipment
Process sequence is the same, and by the arrangement that each individual UVR exposure is product serial number, to each equipment, the product for being arranged in front is first
Processing arranges posterior product post-processing, is then initialized group searching algorithm population as follows：One solution is by NEH rules
It generates, other popsize-1 solution is random to be generated, and using each solution of stochastic simulation data assessment, popsize is that population is big
Small, Three role finder, follower, the patrolman of group searching algorithm have carried out discretization design, and the finder is most
Well the local search based on insertion is carried out on the basis of scheduling solution；Current solution and preferably scheduling solution are carried out intersection behaviour by follower
Make, to obtain more preferable individual；Patrolman disturbs global preferably solution, and replaces the worst scheduling scheme in group；
The phylogenetic scale of group searching algorithm：The operational design for carrying out finder, follower and patrolman obtains preferably production and adjusts
Degree scheme.
2. the multi-products ordering production scheduling method under processing time condition of uncertainty according to claim 1,
It is characterized in that：The uncertain processing time, probability distribution are known, and Monte Carlo mould is carried out according to the probabilistic model
It is quasi-, carry out approximate calculation expectation target value.
3. the multi-products ordering production scheduling method under processing time condition of uncertainty according to claim 1,
It is characterized in that：The stochastic model is the multi-products ordering production scheduling problems model under processing time condition of uncertainty,
Including unlimited intermediate storage tank and without intermediate storage tank two types.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510279760.5A CN105045223B (en) | 2015-05-28 | 2015-05-28 | A kind of multi-products ordering production scheduling method under condition of uncertainty |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510279760.5A CN105045223B (en) | 2015-05-28 | 2015-05-28 | A kind of multi-products ordering production scheduling method under condition of uncertainty |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105045223A CN105045223A (en) | 2015-11-11 |
| CN105045223B true CN105045223B (en) | 2018-08-03 |
Family
ID=54451828
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510279760.5A Expired - Fee Related CN105045223B (en) | 2015-05-28 | 2015-05-28 | A kind of multi-products ordering production scheduling method under condition of uncertainty |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105045223B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1484704A1 (en) * | 2003-06-02 | 2004-12-08 | Institut Francais Du Petrole | Method for helping oilfield management decision making in the presence of technically and economically uncertain parameters |
| CN1912861A (en) * | 2005-08-08 | 2007-02-14 | 上海市计量测试技术研究院 | Method of analog computing synthesis indeterminacy using Monte carlo acounting |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8234295B2 (en) * | 2009-06-03 | 2012-07-31 | International Business Machines Corporation | Managing uncertain data using Monte Carlo techniques |
-
2015
- 2015-05-28 CN CN201510279760.5A patent/CN105045223B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1484704A1 (en) * | 2003-06-02 | 2004-12-08 | Institut Francais Du Petrole | Method for helping oilfield management decision making in the presence of technically and economically uncertain parameters |
| CN1912861A (en) * | 2005-08-08 | 2007-02-14 | 上海市计量测试技术研究院 | Method of analog computing synthesis indeterminacy using Monte carlo acounting |
Non-Patent Citations (4)
| Title |
|---|
| 不确定条件下中间存储时间有限多产品间歇生产过程调度;耿佳灿 等;《化工学报》;20150131;第66卷(第1期);第357-365页 * |
| 不确定条件下柔性作业车间的鲁棒优化调度方法;何明;《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》;20150415(第4期);第C029-98页 * |
| 操作时间不确定的间歇过程最优化设计;曹德铭 等;《化工学报》;20020131;第53卷(第1期);第72-77页 * |
| 组搜索算法研究及其应用;阎兴頔;《中国优秀博士学位论文全文数据库信息科技辑》;20130615(第6期);第I140-6页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105045223A (en) | 2015-11-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zheng et al. | Energy-efficient scheduling for multi-objective two-stage flow shop using a hybrid ant colony optimisation algorithm | |
| Lee et al. | Adaptive transmission scheduling in wireless networks for asynchronous federated learning | |
| CN106527381B (en) | A kind of fast evaluation method towards parallel batch processing machine dynamic dispatching | |
| CN106855957A (en) | Factory's bus load prediction based on similar day and least square method supporting vector machine | |
| Salcedo-Sanz et al. | New coral reefs-based approaches for the model type selection problem: a novel method to predict a nation's future energy demand | |
| CN111367657A (en) | Computing resource collaborative cooperation method based on deep reinforcement learning | |
| CN114565237B (en) | Comprehensive energy low-carbon scheduling method and system for multi-energy combined power station | |
| CN105590143B (en) | Multi-machine assembly line chip mounter load balancing optimization method in PCB assembly process | |
| CN102521706A (en) | KPI data analysis method and device for the same | |
| CN113239639A (en) | Policy information generation method, policy information generation device, electronic device, and storage medium | |
| CN108710970A (en) | A kind of parallel dimension reduction method of Multiobjective Scheduling of huge Hydro Power Systems with Cascaded Reservoirs | |
| CN110290542A (en) | Marine UAV Communication coverage optimization method and system | |
| CN111126707B (en) | Energy consumption equation construction and energy consumption prediction method and device | |
| CN109784570A (en) | Wisdom workshop flexible production dispatching method based on information physical emerging system | |
| CN113688542A (en) | Intelligent optimization water resource configuration method and device, computer equipment and storage medium | |
| CN105045223B (en) | A kind of multi-products ordering production scheduling method under condition of uncertainty | |
| Zheng et al. | A co-evolutionary teaching-learning-based optimization algorithm for stochastic RCPSP | |
| Anita et al. | Solution of emission constrained unit commitment problem using shuffled frog leaping algorithm | |
| CN109635328A (en) | Integrated circuit layout method and distributed design approach | |
| CN112488343B (en) | Smart city intelligent perception terminal address selection method based on hyper-heuristic algorithm | |
| Li et al. | Extended hk evolving network model | |
| Vijay et al. | A detailed investigation on conventional and meta-heuristic optimization algorithms for economic power scheduling problems | |
| CN106970840A (en) | A kind of Method for HW/SW partitioning of combination task scheduling | |
| CN105005842A (en) | Parallel equipment scheduling strategy for just-in-time production | |
| CN111324444A (en) | Cloud computing task scheduling method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180803 Termination date: 20190528 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |