CN109002914A - A kind of production scheduling method and device merging random algorithm and heuristic programming - Google Patents
A kind of production scheduling method and device merging random algorithm and heuristic programming Download PDFInfo
- Publication number
- CN109002914A CN109002914A CN201810758523.0A CN201810758523A CN109002914A CN 109002914 A CN109002914 A CN 109002914A CN 201810758523 A CN201810758523 A CN 201810758523A CN 109002914 A CN109002914 A CN 109002914A
- Authority
- CN
- China
- Prior art keywords
- algorithm
- cause
- effect diagram
- cost
- heuristic
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 70
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 67
- 238000010586 diagram Methods 0.000 claims abstract description 69
- 230000009471 action Effects 0.000 claims abstract description 39
- 238000012546 transfer Methods 0.000 claims description 30
- 230000004927 fusion Effects 0.000 claims description 15
- 238000005457 optimization Methods 0.000 abstract description 14
- 230000008901 benefit Effects 0.000 abstract description 11
- 238000013178 mathematical model Methods 0.000 abstract description 5
- 230000000694 effects Effects 0.000 description 8
- 230000008569 process Effects 0.000 description 5
- 238000010276 construction Methods 0.000 description 4
- 230000000875 corresponding effect Effects 0.000 description 4
- 238000009826 distribution Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000004040 coloring Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000010924 continuous production Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003860 storage 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
- G06Q10/04—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 discloses a kind of production scheduling method and devices for merging random algorithm and heuristic programming, comprising: the action sequence for passing through random algorithm random selection production to any one uncertainty movement；The inspiration value of each variable in target is solved by heuristic programming mode and cause-and-effect diagram maximum cost method for action sequence, it is intuitive, simple and practical to solve existing heuristic, spend less time the feasible solution that scheduling problem can be obtained, but it is relatively low that it optimizes performance, the powerful optimization advantage of mathematical model is not given full play to, it is difficult to carry out optimizing with global viewpoint, is difficult the deviation of quantitative assessment obtained result and globally optimal solution, degree of optimization is not easy the technical issues of holding.
Description
Technical field
The present invention relates to planning field more particularly to a kind of production scheduling methods for merging random algorithm and heuristic programming
And device.
Background technique
Production scheduling be exactly tissue execute scheduling of production work, production scheduling using scheduling of production as foundation,
Scheduling of production will realize that the necessity of production scheduling is to produce movable property by industrial enterprise to determine by production scheduling
Fixed, modern industrial enterprises, production link is more, and cooperation relation is complicated, and continuous production is strong, and situation variation is fast, a certain part hair
Raw failure or a certain measure are not realized on schedule, often involve the operation of entire production system.
Therefore, reinforce production scheduling work, for understanding in time, grasping manufacturing schedule, researching and analysing influences each of production
Kind factor, takes Corresponding Countermeasures according to different situations, and gap is made to reduce or restore normally to be very important.Process industry production
Scheduling can bring apparent economic benefit for actual production, be widely concerned by researchers at home and abroad.Currently, production scheduling
Method can be divided into it is rule-based scheduling and the scheduling two major classes based on model.Rule-based dispatching method is according to certain
Rule and the tactful next step operation to determine production process, also known as heuristic.Heuristic is intuitive, simple real
With, spend less time the feasible solution that scheduling problem can be obtained, but its to optimize performance relatively low, do not give full play to mathematics
The powerful optimization advantage of model, it is difficult to optimizing be carried out with global viewpoint, be difficult the obtained result of quantitative assessment and the overall situation most
The deviation of excellent solution, degree of optimization are not easy to hold.
Summary of the invention
A kind of production scheduling method merging random algorithm and heuristic programming provided by the invention, solves existing inspiration
Formula method is intuitive, simple and practical, spends less time the feasible solution that scheduling problem can be obtained, but it optimizes performance relatively
It is low, do not give full play to the powerful optimization advantage of mathematical model, it is difficult to optimizing be carried out with global viewpoint, be difficult quantitative assessment institute
The deviation of obtained result and globally optimal solution, degree of optimization are not easy the technical issues of holding.
A kind of production scheduling method merging random algorithm and heuristic programming provided by the invention, comprising:
The action sequence that random algorithm random selection production is passed through to any one uncertainty movement；
Pass through for the action sequence each in heuristic programming mode and cause-and-effect diagram maximum cost method solution target
The inspiration value of variable.
Optionally, described to be solved for the action sequence by heuristic programming mode and cause-and-effect diagram maximum cost method
The inspiration value of each variable includes: in target
It calculates an action state s and reaches target sgDistanceAnd the cause-and-effect diagram that the distance is denoted as state s is opened
Hair value, it is describedSpecifically:
Wherein, cos tv(s(v),sg(v)) indicate that variable is transferred to target s from the value in state sgIn value needed for generation
Valence.
Optionally, the uncertain movement is available at state s.
Optionally, the method also includes:
By dijkstra's algorithm, and binding domain transfer figure and cause-and-effect diagram calculate the cos tv(s(v),sg(v))。
Optionally, described by dijkstra's algorithm, and binding domain transfer figure and cause-and-effect diagram calculate the cos tv(s
(v),sg(v)) it specifically includes:
If variable v does not have the direct older generation in cause-and-effect diagram, then cos tv(s(v),sg(v)) for v domain shift figure in from
S (v) arrives sg(v) shortest path length, if it does not exist from s (v) to sg(v) path, then cos tv(s(v),sg(v)) it is positive
It is infinite；
Enable vvFor the set of the direct older generation of v in cause-and-effect diagram, if v is from s (v) to sg(v) transfer is with vvMiddle variable vpTax
Premised on value, then in vpDomain transfer figure in find and meet the shortest path length of the assignment, and the length is added to cos tv
(s(v),sg(v)) in.
A kind of production scheduling device merging random algorithm and heuristic programming provided by the invention, comprising:
Random selection module, the movement for passing through random algorithm random selection production to any one uncertainty movement
Sequence；
It inspires and solves module, for passing through heuristic programming mode and cause-and-effect diagram maximum cost for the action sequence
Method solves the inspiration value of each variable in target.
Optionally, the inspiration solves module, is used for:
It calculates an action state s and reaches target sgDistanceAnd the cause-and-effect diagram that the distance is denoted as state s is opened
Hair value, it is describedSpecifically:
Wherein, cos tv(s(v),sg(v)) indicate that variable is transferred to target s from the value in state sgIn value needed for generation
Valence.
Optionally, the uncertain movement is available at state s.
Optionally, further includes:
Computing module, for passing through dijkstra's algorithm, and binding domain transfer figure and cause-and-effect diagram calculate the cos tv(s
(v),sg(v))。
Optionally, the computing module, is used for:
If variable v does not have the direct older generation in cause-and-effect diagram, then cos tv(s(v),sg(v)) for v domain shift figure in from
S (v) arrives sg(v) shortest path length, if it does not exist from s (v) to sg(v) path, then cos tv(s(v),sg(v)) it is positive
It is infinite；
Enable vvFor the set of the direct older generation of v in cause-and-effect diagram, if v is from s (v) to sg(v) transfer is with vvMiddle variable vpTax
Premised on value, then in vpDomain transfer figure in find and meet the shortest path length of the assignment, and the length is added to cos tv
(s(v),sg(v)) in.
As can be seen from the above technical solutions, the invention has the following advantages that
A kind of production scheduling method merging random algorithm and heuristic programming provided by the invention, comprising: to any one
The action sequence that a uncertain movement passes through random algorithm random selection production；Pass through heuristic rule for the action sequence
The mode of drawing and cause-and-effect diagram maximum cost method solve the inspiration value of each variable in target, on the basis of production scheduling model, knot
The model based on random algorithm and heuristic programming has been closed, i.e., selection corresponding actions have been carried out to each link by random algorithm
Sequence, and problem is solved with heuristic, so that obtained solution quality is secure, there is linear relationship,
It solves that existing heuristic is intuitive, simple and practical, spends less time the feasible solution that scheduling problem can be obtained, but its is excellent
It is relatively low to change performance, does not give full play to the powerful optimization advantage of mathematical model, it is difficult to optimizing be carried out with global viewpoint, very
The deviation of difficult quantitative assessment obtained result and globally optimal solution, degree of optimization are not easy the technical issues of holding.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention without any creative labor, may be used also for those of ordinary skill in the art
To obtain other attached drawings according to these attached drawings.
Fig. 1 is the one of the production scheduling method of a kind of fusion random algorithm provided in an embodiment of the present invention and heuristic programming
The flow diagram of a embodiment；
Fig. 2 is the another of the production scheduling method of a kind of fusion random algorithm provided in an embodiment of the present invention and heuristic programming
The flow diagram of one embodiment；
Fig. 3 is the one of the production scheduling device of a kind of fusion random algorithm provided in an embodiment of the present invention and heuristic programming
The structural schematic diagram of a embodiment.
Specific embodiment
In order to make the invention's purpose, features and advantages of the invention more obvious and easy to understand, below in conjunction with the present invention
Attached drawing in embodiment, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that disclosed below
Embodiment be only a part of the embodiment of the present invention, and not all embodiment.Based on the embodiments of the present invention, this field
Those of ordinary skill's all other embodiment obtained without making creative work, belongs to protection of the present invention
Range.
Please refer to Fig. 1, a kind of production scheduling side for merging random algorithm and heuristic programming provided in an embodiment of the present invention
Method may include:
Step S100: the action sequence that random algorithm random selection production is passed through to any one uncertainty movement；
In the embodiment of the present invention, when carrying out the production scheduling of fusion random algorithm and heuristic programming, need to any
The action sequence that one uncertain movement passes through random algorithm random selection production；
It should be noted that random algorithm develops to today in the algorithm of many types from the tool of an inversion
It is all widely applied, it is shown that the powerful vitality of random algorithm itself.So-called random algorithm is exactly in the process of implementation
Make randomly selected algorithm.There are two advantages for random algorithm: simple and quick.Random algorithm is studied, in addition to its research
It brings outside new method and new thought.Another major reason is the relationship of it and algorithm complexity.The time of algorithm is discussed
Complexity estimation.To answer P class problem algorithm analysis, several compellent results are had been achieved with.But for NP double linear problems of difficulty for solving,
Still it does not make substantial progress.In this background, research answer NP is difficult to resolve causes people very big with the approximate algorithm inscribed
Interest.But the estimation of pairing approximation algorithm approximation performance ratio requires all to set up all examples answered a question, and just seems
It is too harsh.Causing in the case where P ≠ NP assumes, several NP double linear problems of difficulty for solving can not find approximate algorithm of its approximate performance than bounded, such as
Itinerant pedlar's optimization problem.Even if some NP double linear problems of difficulty for solving have algorithm of the approximate performance than bounded, but this upper bound is too big, such as schemes
Coloring is the same as topic.In response to this, Karp attempts to presuppose problem-instance obeys certain probability distribution in instance space, if
It counts out and answers the probabilistic algorithm that several NP are difficult to resolve same topic.
There are two types of thinkings for probabilistic algorithm, one is algorithm is deterministic type algorithm.The example answered a question obeys certain probability
Distribution, can parser expected time complexity or probability be 1 (almost everywhere) under conditions of.Provide the essence answered a question
Really solution or approximate solution.Another thinking of estimate by design algorithm is that the instance space of Solve problems is determining, and by random language
Sentence introduces algorithm, and here it is the random algorithms usually said.
Their real difficulty of many np problems is perhaps that we cannot find out the thing of consistency in its construction.Change sentence
It talks about, their construction is too bad so that we cannot portray them, according to classical based on clear clear and accurately
The theory of algorithm of discrete topology can not find the good algorithm solved the problems, such as naturally.At this point, random algorithm is had to take the second best, do not consider
Time consumption under worst case, and consider its average time and expend, finding one is in most cases a good algorithm
Solution route, although it may not be absolutely good.But compare problem be unable to estimate it is far better, in numerous applications-
Random algorithm is to find simplest perhaps most fast algorithm or want it both ways.
Problem solving difficulty has various reasons, wherein it is important that: the condition of problem changes over time, or even can
Some undesirable factors can occur to need to carry out special operation feasible to generate in addition, practical problem is usually present constraint
Solution.Heuristic is a kind of strategy for seeking to solve the problems, such as these difficulties, it establishes the base in people's experience and judgement
On plinth, the subjective initiative and creativity of people are embodied, its main feature is that utilizing past experience, selection when solving the problems, such as
The effective method of maximum possible, rather than systematically, with determining step go to seek answer.
Common algorithm often all attempts various possibilities, can finally find the solution of required problem, but it may be needed
Will be in very big problem space, answer can just be acquired by devoting a tremendous amount of time, the time complexity and space complexity of problem
It is quite big.And heuristic is to take current most satisfied action in limited search space, reduces the number of trial to the greatest extent
Amount, can quickly solve the problems, such as.But due to this method meeting trial and error, it is possible that will fail.
Since stringent optimal solution is not present in many problems, or it is not necessarily to obtain the solution with height-precision, Huo Zheyao
It obtains that the cost that optimal solution is spent is too big, at this moment often more can accurately describe the selection of people than optimality to satisfaction property of target
Behavior.So heuristic is not emphasized to obtain optimal solution, but obtain the satisfactory solution of problem.
Heuristic search also exactly handles the main stream approach of large space search problem, its highly effective effect in recent years
Constantly it is proved.In certainty planning, heuristic has been widely used, and asks to handle high-dimensional space search
Topic, and good effect is obtained, but for practical problem, the optimality of solution is required to be frequently not so strictly, in satisfaction
Suboptimal solution in time more corresponds to actual needs.
Therefore, the present invention introduces random algorithm before heuristic programming: passing through to any one uncertainty movement random
The action sequence of algorithm random selection production improves efficiency under the premise of guaranteeing certain accuracy.
Step S101, it is solved in target for action sequence by heuristic programming mode and cause-and-effect diagram maximum cost method
The inspiration value of each variable；
After any one action sequence of uncertainty movement by random algorithm random selection production, need to be directed to
Action sequence solves the inspiration value of each variable in target by heuristic programming mode and cause-and-effect diagram maximum cost method；
The basic thought of cause-and-effect diagram heuristic is: calculating an action state s and reaches target sgDistanceBy this
A distance is denoted as the cause-and-effect diagram inspiration value of state s, hereIs defined as:
Wherein cos tv(s(v),sg(v)) indicate that variable is transferred to target s from the value in state sgIn value needed for generation
Valence.As can be seen from the above equation, the cause-and-effect diagram inspiration value of a state is the sum of the inspiration value of each variable in target, therefore this is
A kind of and cost method.Another value s is transferred to from a value s (v) in domain calculating variable vg(v) cost cos tv(s
(v),sg(v)) when, binding domain transfer figure and cause-and-effect diagram are needed, what is mainly utilized is similar dijkstra's algorithm:
If variable v does not have the direct older generation in cause-and-effect diagram, then cos tv(s(v),sg(v))cos tv(s(v),sg
(v)) it shifts in figure for the domain of v from s (v) to sg(v) shortest path length, if it does not exist from s (v) to sg(v) path, then
cos tv(s(v),sg(v)) it is positive infinite；
Enable vvFor the set of the direct older generation of v in cause-and-effect diagram, if v is from s (v) to sg(v) transfer is with vvMiddle variable vpTax
Premised on value, then in vpDomain transfer figure in find and meet the shortest path length of the assignment, and the length is added to cos tv
(s(v),sg(v)) in.
Wherein, the basic cost of all high-rise variable transfers is all 1.
It is not usually independent between each variable especially in production scheduling problems but in practical problem, mesh
When a variable in mark is satisfied, positive interaction may be generated to other variables.Effect is shown as when maximum so that other variables
Also target is reached, therefore, introduces maximum cost method, the calculation method of cause-and-effect diagram maximum cost method more excellent to the appraisal of state
Are as follows:
Cause-and-effect diagram maximum cost method is also that can not receive, because calculating cos tv(s(v),sg(v)) when, which is embedded
With cost method.But the admissibility of cause-and-effect diagram maximum cost method is better than cause-and-effect diagram and cost method, actively makees when generating between variable
Used time, if excessively high using the inspiration value that cause-and-effect diagram and cost method obtain, it is possible to so that search finding process slows down, or
To planning solution length be not optimal, and optimal solution can be obtained with cause-and-effect diagram maximum cost rule.And if only if the domain of variable
When jump condition in transfer figure is not more than one, i.e., when the effect generated between no jump condition, cause-and-effect diagram maximum cost method
It is admissible.
A kind of production scheduling method merging random algorithm and heuristic programming provided by the invention, comprising: to any one
The action sequence that a uncertain movement passes through random algorithm random selection production；Pass through heuristic programming side for action sequence
The inspiration value that formula and cause-and-effect diagram maximum cost method solve each variable in target combines on the basis of production scheduling model
Model based on random algorithm and heuristic programming carries out selection corresponding actions sequence to each link by random algorithm
Column, and problem is solved with heuristic, so that obtained solution quality is secure, there is linear relationship, solution
Existing heuristic of having determined is intuitive, simple and practical, spends less time the feasible solution that scheduling problem can be obtained, but it optimizes
Performance is relatively low, does not give full play to the powerful optimization advantage of mathematical model, it is difficult to carry out optimizing with global viewpoint, be difficult
The deviation of quantitative assessment obtained result and globally optimal solution, degree of optimization are not easy the technical issues of holding.
The above is carried out to one embodiment of a kind of production scheduling method of fusion random algorithm and heuristic programming
Detailed description below will carry out another embodiment of a kind of fusion random algorithm and the production scheduling method of heuristic programming
Detailed description.
Referring to Fig. 2, a kind of production scheduling side for merging random algorithm and heuristic programming provided in an embodiment of the present invention
Method may include:
Step S200: the action sequence that random algorithm random selection production is passed through to any one uncertainty movement；
In the embodiment of the present invention, when carrying out the production scheduling of fusion random algorithm and heuristic programming, need to any
The action sequence that one uncertain movement passes through random algorithm random selection production；
It should be noted that random algorithm develops to today in the algorithm of many types from the tool of an inversion
It is all widely applied, it is shown that the powerful vitality of random algorithm itself.So-called random algorithm is exactly in the process of implementation
Make randomly selected algorithm.There are two advantages for random algorithm: simple and quick.Random algorithm is studied, in addition to its research
It brings outside new method and new thought.Another major reason is the relationship of it and algorithm complexity.The time of algorithm is discussed
Complexity estimation.To answer P class problem algorithm analysis, several compellent results are had been achieved with.But for NP double linear problems of difficulty for solving,
Still it does not make substantial progress.In this background, research answer NP is difficult to resolve causes people very big with the approximate algorithm inscribed
Interest.But the estimation of pairing approximation algorithm approximation performance ratio requires all to set up all examples answered a question, and just seems
It is too harsh.Causing in the case where P ≠ NP assumes, several NP double linear problems of difficulty for solving can not find approximate algorithm of its approximate performance than bounded, such as
Itinerant pedlar's optimization is the same as topic.Even if some NP double linear problems of difficulty for solving have algorithm of the approximate performance than bounded, but this upper bound is too big, such as schemes
Coloring is the same as topic.In response to this, Karp attempts to presuppose problem-instance obeys certain probability distribution in instance space, if
It counts out and answers the probabilistic algorithm that several NP are difficult to resolve same topic.
There are two types of thinkings for probabilistic algorithm, one is algorithm is deterministic type algorithm.The example answered a question obeys certain probability
Distribution, can parser expected time complexity or probability be 1 (almost everywhere) under conditions of.Provide the essence answered a question
Really solution or approximate solution.Another thinking of estimate by design algorithm is that the instance space of Solve problems is determining, and by random language
Sentence introduces algorithm, and here it is the random algorithms usually said.
Their real difficulty of many np problems is perhaps that we cannot find out the thing of consistency in its construction.Change sentence
It talks about, their construction is too bad so that we cannot portray them, according to classical based on clear clear and accurately
The theory of algorithm of discrete topology can not find the good algorithm solved the problems, such as naturally.At this point, random algorithm is had to take the second best, do not consider
Time consumption under worst case, and consider its average time and expend, finding one is in most cases a good algorithm
Solution route, although it may not be absolutely good.But compare problem be unable to estimate it is far better, in numerous applications-
Random algorithm is to find simplest perhaps most fast algorithm or want it both ways.
Problem solving difficulty has various reasons, wherein it is important that: the condition of problem changes over time, or even can
Some undesirable factors can occur to need to carry out special operation feasible to generate in addition, practical problem is usually present constraint
Solution.Heuristic is a kind of strategy for seeking to solve the problems, such as these difficulties, it establishes the base in people's experience and judgement
On plinth, the subjective initiative and creativity of people are embodied, its main feature is that utilizing past experience, selection when solving the problems, such as
The effective method of maximum possible, rather than systematically, with determining step go to seek answer.
Common algorithm often all attempts various possibilities, can finally find the solution of required problem, but it may be needed
Will be in very big problem space, answer can just be acquired by devoting a tremendous amount of time, the time complexity and space complexity of problem
It is quite big.And heuristic is to take current most satisfied action in limited search space, reduces the number of trial to the greatest extent
Amount, can quickly solve the problems, such as.But due to this method meeting trial and error, it is possible that will fail.
Since stringent optimal solution is not present in many problems, or it is not necessarily to obtain the solution with height-precision, Huo Zheyao
It obtains that the cost that optimal solution is spent is too big, at this moment often more can accurately describe the selection of people than optimality to satisfaction property of target
Behavior.So heuristic is not emphasized to obtain optimal solution, but obtain the satisfactory solution of problem.
Heuristic search also exactly handles the main stream approach of large space search problem, its highly effective effect in recent years
Constantly it is proved.In certainty planning, heuristic has been widely used, and asks to handle high-dimensional space search
Topic, and good effect is obtained, but for practical problem, the optimality of solution is required to be frequently not so strictly, in satisfaction
Suboptimal solution in time more corresponds to actual needs.
Therefore, the present invention introduces random algorithm before heuristic programming: passing through to any one uncertainty movement random
The action sequence of algorithm random selection production improves efficiency under the premise of guaranteeing certain accuracy.
Step S201, it calculates an action state s and reaches target sgDistanceAnd by distance be denoted as state s because
Fruit figure inspiration value,Specifically:
Wherein, cos tv(s(v),sg(v)) indicate that variable is transferred to target s from the value in state sgIn value needed for generation
Valence, by dijkstra's algorithm, and binding domain transfer figure and cause-and-effect diagram calculate the cos tv(s(v),sg(v))；
Wherein, if in cause-and-effect diagram, variable v does not have the direct older generation, then cos tv(s(v),sg(v))cos tv(s(v),
sg(v)) it shifts in figure for the domain of v from s (v) to sg(v) shortest path length, if it does not exist from s (v) to sg(v) path,
Then cos tv(s(v),sg(v)) it is positive infinite；
Enable vvFor the set of the direct older generation of v in cause-and-effect diagram, if v is from s (v) to sg(v) transfer is with vvMiddle variable vpTax
Premised on value, then in vpDomain transfer figure in find and meet the shortest path length of the assignment, and the length is added to cos tv
(s(v),sg(v)) in
After any one action sequence of uncertainty movement by random algorithm random selection production, need to calculate
One action state s reaches target sgDistanceAnd distance is denoted as to the cause-and-effect diagram inspiration value of state s,Specifically
Are as follows:
Wherein, cos tv(s(v),sg(v)) indicate that variable is transferred to target s from the value in state sgIn value needed for generation
Valence；
It should be noted that uncertain movement is available at state s.
The basic thought of cause-and-effect diagram heuristic is: calculating an action state s and reaches target sgDistanceBy this
A distance is denoted as the cause-and-effect diagram inspiration value of state s, hereIs defined as:
Wherein cos tv(s(v),sg(v)) indicate that variable is transferred to target s from the value in state sgIn value needed for generation
Valence.As can be seen from the above equation, the cause-and-effect diagram inspiration value of a state is the sum of the inspiration value of each variable in target, therefore this is
A kind of and cost method.Another value s is transferred to from a value s (v) in domain calculating variable vg(v) cost cos tv(s
(v),sg(v)) when, binding domain transfer figure and cause-and-effect diagram are needed, what is mainly utilized is similar dijkstra's algorithm:
If variable v does not have the direct older generation in cause-and-effect diagram, then cos tv(s(v),sg(v))cos tv(s(v),sg
(v)) it shifts in figure for the domain of v from s (v) to sg(v) shortest path length, if it does not exist from s (v) to sg(v) path, then
cos tv(s(v),sg(v)) it is positive infinite；
Enable vvFor the set of the direct older generation of v in cause-and-effect diagram, if v is from s (v) to sg(v) transfer is with vvMiddle variable vpTax
Premised on value, then in vpDomain transfer figure in find and meet the shortest path length of the assignment, and the length is added to cos tv
(s(v),sg(v)) in.
Wherein, the basic cost of all high-rise variable transfers is all 1.
It is not usually independent between each variable especially in production scheduling problems but in practical problem, mesh
When a variable in mark is satisfied, positive interaction may be generated to other variables.Effect is shown as when maximum so that other variables
Also target is reached, therefore, introduces maximum cost method, the calculation method of cause-and-effect diagram maximum cost method more excellent to the appraisal of state
Are as follows:
Cause-and-effect diagram maximum cost method is also that can not receive, because calculating cos tv(s(v),sg(v)) when, which is embedded
With cost method.But the admissibility of cause-and-effect diagram maximum cost method is better than cause-and-effect diagram and cost method, actively makees when generating between variable
Used time, if excessively high using the inspiration value that cause-and-effect diagram and cost method obtain, it is possible to so that search finding process slows down, or
To planning solution length be not optimal, and optimal solution can be obtained with cause-and-effect diagram maximum cost rule.And if only if the domain of variable
When jump condition in transfer figure is not more than one, i.e., when the effect generated between no jump condition, cause-and-effect diagram maximum cost method
It is admissible.
A kind of production scheduling method merging random algorithm and heuristic programming provided by the invention, comprising: to any one
The action sequence that a uncertain movement passes through random algorithm random selection production；Pass through heuristic programming side for action sequence
The inspiration value that formula and cause-and-effect diagram maximum cost method solve each variable in target combines on the basis of production scheduling model
Model based on random algorithm and heuristic programming carries out selection corresponding actions sequence to each link by random algorithm
Column, and problem is solved with heuristic, so that obtained solution quality is secure, there is linear relationship, solution
Existing heuristic of having determined is intuitive, simple and practical, spends less time the feasible solution that scheduling problem can be obtained, but it optimizes
Performance is relatively low, does not give full play to the powerful optimization advantage of mathematical model, it is difficult to carry out optimizing with global viewpoint, be difficult
The deviation of quantitative assessment obtained result and globally optimal solution, degree of optimization are not easy the technical issues of holding.
Referring to Fig. 3, Fig. 3 shows a kind of life for merging random algorithm and heuristic programming provided in an embodiment of the present invention
Produce the structural schematic diagram of dispatching device, comprising:
Random selection module 301, for passing through random algorithm random selection production to any one uncertainty movement
Action sequence；
It inspires and solves module 302, for passing through heuristic programming mode and cause-and-effect diagram maximum cost for action sequence
Method solves the inspiration value of each variable in target.
Optionally, it inspires and solves module 302, be used for:
It calculates an action state s and reaches target sgDistanceAnd the cause-and-effect diagram that distance is denoted as state s is inspired
Value,Specifically:
Wherein, cos tv(s(v),sg(v)) indicate that variable is transferred to target s from the value in state sgIn value needed for generation
Valence.
Optionally, uncertain movement is available at state s.
Optionally, further includes:
Computing module 303, for passing through dijkstra's algorithm, and binding domain transfer figure and cause-and-effect diagram calculate cos tv(s
(v),sg(v))。
Optionally, computing module 303 are used for:
If variable v does not have the direct older generation in cause-and-effect diagram, then cos tv(s(v),sg(v)) for v domain shift figure in from
S (v) arrives sg(v) shortest path length, if it does not exist from s (v) to sg(v) path, then cos tv(s(v),sg(v)) it is positive
It is infinite；
Enable vvFor the set of the direct older generation of v in cause-and-effect diagram, if v is from s (v) to sg(v) transfer is with vvMiddle variable vpTax
Premised on value, then in vpDomain transfer figure in find and meet the shortest path length of the assignment, and the length is added to cos tv
(s(v)),sg(v)) in.
Professional further appreciates that, unit described in conjunction with the examples disclosed in the embodiments of the present disclosure
And algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate hardware and
The interchangeability of software generally describes each exemplary composition and step according to function in the above description.These
Function is implemented in hardware or software actually, the specific application and design constraint depending on technical solution.Profession
Technical staff can use different methods to achieve the described function each specific application, but this realization is not answered
Think beyond the scope of this invention.
The step of method described in conjunction with the examples disclosed in this document or algorithm, can directly be held with hardware, processor
The combination of capable software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only deposit
Reservoir (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technology
In any other form of storage medium well known in field.
The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present invention.
Various modifications to these embodiments will be readily apparent to those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore, of the invention
It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one
The widest scope of cause.
Claims (10)
1. a kind of production scheduling method for merging random algorithm and heuristic programming characterized by comprising
The action sequence that random algorithm random selection production is passed through to any one uncertainty movement；
Each variable in target is solved by heuristic programming mode and cause-and-effect diagram maximum cost method for the action sequence
Inspiration value.
2. the production scheduling method of fusion random algorithm and heuristic programming according to claim 1, which is characterized in that institute
It states and each variable in target is solved by heuristic programming mode and cause-and-effect diagram maximum cost method for the action sequence
Inspiration value includes:
It calculates an action state s and reaches target sgDistanceAnd the cause-and-effect diagram that the distance is denoted as state s is inspired
Value, it is describedSpecifically:
Wherein, costv(s(v),sg(v)) indicate that variable is transferred to target s from the value in state sgIn value needed for cost.
3. the production scheduling method of fusion random algorithm and heuristic programming according to claim 2, which is characterized in that institute
Uncertain movement is stated to be available at state s.
4. the production scheduling method of fusion random algorithm and heuristic programming according to claim 3, which is characterized in that institute
State method further include:
By dijkstra's algorithm, and binding domain transfer figure and cause-and-effect diagram calculate the costv(s(v),sg(v))。
5. the production scheduling method of fusion random algorithm and heuristic programming according to claim 4, which is characterized in that institute
It states through dijkstra's algorithm, and binding domain transfer figure and cause-and-effect diagram calculate the costv(s(v),sg(v)) it specifically includes:
If variable v does not have the direct older generation in cause-and-effect diagram, then costv(s(v),sg(v)) it shifts in figure for the domain of v from s (v)
To sg(v) shortest path length, if it does not exist from s (v) to sg(v) path, then costv(s(v),sg(v)) it is positive infinite；
Enable vvFor the set of the direct older generation of v in cause-and-effect diagram, if v is from s (v) to sg(v) transfer is with vvMiddle variable vpBe assigned a value of
Premise, then in vpDomain transfer figure in find and meet the shortest path length of the assignment, and the length is added to costv(s
(v),sg(v)) in.
6. a kind of production scheduling device for merging random algorithm and heuristic programming characterized by comprising
Random selection module, the movement sequence for passing through random algorithm random selection production to any one uncertainty movement
Column；
It inspires and solves module, for being asked for the action sequence by heuristic programming mode and cause-and-effect diagram maximum cost method
Solve the inspiration value of each variable in target.
7. the production scheduling device of fusion random algorithm and heuristic programming according to claim 6, which is characterized in that institute
It states inspiration and solves module, be used for:
It calculates an action state s and reaches target sgDistanceAnd the cause-and-effect diagram that the distance is denoted as state s is inspired
Value, it is describedSpecifically:
Wherein, costv(s(v),sg(v)) indicate that variable is transferred to target s from the value in state sgIn value needed for cost.
8. the production scheduling device of fusion random algorithm and heuristic programming according to claim 7, which is characterized in that institute
Uncertain movement is stated to be available at state s.
9. the production scheduling device of fusion random algorithm and heuristic programming according to claim 8, which is characterized in that also
Include:
Computing module, for passing through dijkstra's algorithm, and binding domain transfer figure and cause-and-effect diagram calculate the costv(s(v),sg
(v))。
10. the production scheduling device of fusion random algorithm and heuristic programming according to claim 9, which is characterized in that
The computing module, is used for:
If variable v does not have the direct older generation in cause-and-effect diagram, then costv(s (v), sg(v)) it shifts in figure for the domain of v from s (v)
To sg(v) shortest path length, if it does not exist from s (v) to sg(v) path, then costv(s(v),sg(v)) it is positive infinite；
Enable vvFor the set of the direct older generation of v in cause-and-effect diagram, if v is from s (v) to sg(v) transfer is with vvMiddle variable vpBe assigned a value of
Premise, then in vpDomain transfer figure in find and meet the shortest path length of the assignment, and the length is added to costv(s
(v),sg(v)) in.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810758523.0A CN109002914A (en) | 2018-07-11 | 2018-07-11 | A kind of production scheduling method and device merging random algorithm and heuristic programming |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810758523.0A CN109002914A (en) | 2018-07-11 | 2018-07-11 | A kind of production scheduling method and device merging random algorithm and heuristic programming |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109002914A true CN109002914A (en) | 2018-12-14 |
Family
ID=64598954
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810758523.0A Pending CN109002914A (en) | 2018-07-11 | 2018-07-11 | A kind of production scheduling method and device merging random algorithm and heuristic programming |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109002914A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110850893A (en) * | 2019-11-29 | 2020-02-28 | 北京理工大学 | Spacecraft task planning method based on maximum cost evaluation |
| CN112811273A (en) * | 2021-01-27 | 2021-05-18 | 广东工业大学 | Probability parallel planning based elevator scheduling method based on real-time additivity heuristic method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102521203A (en) * | 2011-11-18 | 2012-06-27 | 东南大学 | Hierarchical reinforcement learning task graph evolution method based on cause-and-effect diagram |
| CN104166903A (en) * | 2014-08-18 | 2014-11-26 | 四川航天系统工程研究所 | Task planning method and system based on working procedure division |
| CN105844364A (en) * | 2016-04-08 | 2016-08-10 | 上海派毅智能科技有限公司 | Service robot optimal path program method based on heuristic function |
| CN106203634A (en) * | 2016-07-20 | 2016-12-07 | 广东工业大学 | A kind of based on the didactic parallel probability plan method of cause-and-effect diagram |
-
2018
- 2018-07-11 CN CN201810758523.0A patent/CN109002914A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102521203A (en) * | 2011-11-18 | 2012-06-27 | 东南大学 | Hierarchical reinforcement learning task graph evolution method based on cause-and-effect diagram |
| CN104166903A (en) * | 2014-08-18 | 2014-11-26 | 四川航天系统工程研究所 | Task planning method and system based on working procedure division |
| CN105844364A (en) * | 2016-04-08 | 2016-08-10 | 上海派毅智能科技有限公司 | Service robot optimal path program method based on heuristic function |
| CN106203634A (en) * | 2016-07-20 | 2016-12-07 | 广东工业大学 | A kind of based on the didactic parallel probability plan method of cause-and-effect diagram |
Non-Patent Citations (2)
| Title |
|---|
| 李晋丽: "基于因果图的启发式规划研究与实现", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 饶东宁等: "基于因果图启发式的并行概率规划求解", 《计算机应用研究》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110850893A (en) * | 2019-11-29 | 2020-02-28 | 北京理工大学 | Spacecraft task planning method based on maximum cost evaluation |
| CN112811273A (en) * | 2021-01-27 | 2021-05-18 | 广东工业大学 | Probability parallel planning based elevator scheduling method based on real-time additivity heuristic method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gandomani et al. | An empirically-developed framework for Agile transition and adoption: A Grounded Theory approach | |
| Neumann et al. | Resource levelling for projects with schedule-dependent time windows | |
| Kaiser et al. | Differences in managerial jobs at the bottom, middle, and top: A review of empirical research | |
| Williams | The contribution of mathematical modelling to the practice of project management | |
| Nassar et al. | Developing a complexity measure for project schedules | |
| Mazzucato et al. | The Green New Deal: A bold mission-oriented approach | |
| Obit et al. | Designing a multi-agent approach system for distributed course timetabling | |
| CN109002914A (en) | A kind of production scheduling method and device merging random algorithm and heuristic programming | |
| Sulaiman et al. | A new hybrid conjugate gradient algorithm for optimization models and its application to regression analysis | |
| Mastio et al. | Distributed agent-based traffic simulations | |
| Zilberstein | Resource-bounded sensing and planning in autonomous systems | |
| Zahid et al. | Investigation of surrogate measures of robustness for project scheduling problems | |
| Sengupta et al. | Coping with staffing delays in software project management: an experimental investigation | |
| Mailler et al. | A Cooperative mediation-based protocol for dynamic distributed resource allocation | |
| Liu et al. | A Data-Driven Surrogate Modeling for Train Rescheduling in High-Speed Railway Networks Under Wind-Caused Speed Restrictions | |
| EP3588326A1 (en) | Solving a deterministic global optimisation problem | |
| Decker et al. | Analyzing a quantitative coordination relationship | |
| Chakrabortty et al. | Robust optimization based heuristic approach for solving stochastic multi-mode resource constrained project scheduling problem | |
| Parker et al. | Max-sum for allocation of changing cost tasks | |
| Wensveen et al. | Real-time optimistic planning with action sequences | |
| Magdaleno et al. | A practical experience in designing business processes to improve collaboration | |
| Chen et al. | Community Life Project Platform (CLPP) for Collective Impact–Based on Community Design 3.0 Framework | |
| Li et al. | Self-Organization, Dynamic Meta-governance, and Value Creation in Megaprojects | |
| Hindi | Efficient solution of the single-item, capacitated lot-sizing problem with start-up and reservation costs | |
| Faria | Project Management Under Uncertainty: a mixed approach using flexible resource management to exploit schedule flexibility |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181214 |
|
| RJ01 | Rejection of invention patent application after publication |