CN118226808B - Dryer group scheduling optimization algorithm based on weight matrix - Google Patents
Dryer group scheduling optimization algorithm based on weight matrix Download PDFInfo
- Publication number
- CN118226808B CN118226808B CN202410155140.XA CN202410155140A CN118226808B CN 118226808 B CN118226808 B CN 118226808B CN 202410155140 A CN202410155140 A CN 202410155140A CN 118226808 B CN118226808 B CN 118226808B
- Authority
- CN
- China
- Prior art keywords
- dryer
- weight
- weight matrix
- time
- dryer group
- 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
- 239000011159 matrix material Substances 0.000 title claims abstract description 49
- 238000005457 optimization Methods 0.000 title claims abstract description 13
- 238000000034 method Methods 0.000 claims abstract description 17
- 239000000178 monomer Substances 0.000 claims abstract description 12
- 238000010981 drying operation Methods 0.000 claims abstract description 9
- 230000005611 electricity Effects 0.000 claims abstract description 9
- 230000008929 regeneration Effects 0.000 claims description 9
- 238000011069 regeneration method Methods 0.000 claims description 9
- 238000001035 drying Methods 0.000 claims description 5
- 238000004519 manufacturing process Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000004064 recycling Methods 0.000 description 2
- 238000007605 air drying Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F26—DRYING
- F26B—DRYING SOLID MATERIALS OR OBJECTS BY REMOVING LIQUID THEREFROM
- F26B21/00—Arrangements or duct systems, e.g. in combination with pallet boxes, for supplying and controlling air or gases for drying solid materials or objects
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
- G05B19/41865—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by job scheduling, process planning, material flow
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/32—Operator till task planning
- G05B2219/32252—Scheduling production, machining, job shop
Landscapes
- Engineering & Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Mechanical Engineering (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Quality & Reliability (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Automation & Control Theory (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Manufacturing & Machinery (AREA)
- Drying Of Solid Materials (AREA)
Abstract
The invention relates to the technical field of dryer group scheduling, in particular to a dryer group scheduling optimization algorithm based on a weight matrix. Which comprises the following steps: the working time of the dryer group to be scheduled is discretized into time nodes; constructing an internal state table of each dryer; a monomer weight value updating formula is formulated according to a pre-acquired electricity fee interval table of the place where the dryer group is located, and an integral weight value updating formula is formulated according to the completion condition of the integral gas drying operation task of the dryer group; constructing a weight matrix; outputting new weight values to update to the weight matrix according to the change of the dryer state corresponding to each time node in the internal state table, and repeating the process until all the time nodes are traversed; and the corresponding state table and weight matrix table until the consumed electricity fee in the output cycle is the lowest. The invention adopts a mode of updating the weight matrix table, and can achieve the aim of reducing the expenditure of electric charge to the maximum extent on the premise of not influencing normal gas drying operation.
Description
Technical Field
The invention relates to the technical field of dryer group scheduling, in particular to a dryer group scheduling optimization algorithm based on a weight matrix.
Background
As modern plants continue to increase in production efficiency, resource utilization, and reduced production costs, efficient plant job scheduling becomes increasingly important. Factory job scheduling refers to the process of optimizing job sequence, reasonably distributing resources, controlling production progress and the like in a factory comprising a plurality of available machines, so that enterprises can realize more efficient production operation. Factory job scheduling involves a number of factors such as job time, priority, machine job capacity and constraints. In order to find the optimal scheduling scheme, proper algorithms, techniques and strategies need to be applied from the global perspective.
However, conventional factory job scheduling is typically done by a person or simple procedure, which often does not meet these requirements well. In solving the problem of factory job scheduling, the current common methods are classical heuristic algorithms such as genetic algorithm, particle swarm algorithm and the like, and variants thereof. While heuristic algorithms can generally perform well on large-scale, complex problems, a near optimal solution is found in a short time. But such algorithms are typically based on local search strategies and do not guarantee that a globally optimal solution is found. Moreover, heuristic algorithms involve a large number of parameter settings, such as population size, number of iterations, neighborhood size, etc. The selection of different parameter values may lead to different results, requiring setting of specific parameters for specific situations. With the continued development and increasing complexity of the manufacturing industry, such approaches tend to be difficult to meet with ever-increasing scheduling demands.
Therefore, a simple and highly automated method for scheduling plant operations is highly demanded in modern plants.
Disclosure of Invention
The invention provides a dryer group scheduling optimization algorithm based on a weight matrix, which can overcome certain or certain defects in the prior art.
The invention relates to a dryer group scheduling optimization algorithm based on a weight matrix, which comprises the following steps of:
The working time of the dryer group to be scheduled is discretized into time nodes; first, 24 hours continued in one day are divided into 24 time intervals by whole points, and then the 24 node values of 0 to 23 are used to index the time intervals. Then, the node value S on the t-th day is calculated as s= (t-1) ×24+s, S being the node value obtained in the previous step.
Taking the time node as a row and the state of the dryer as a column, and constructing an internal state table of each dryer;
A monomer weight value updating formula is formulated according to a pre-acquired electricity fee interval table of the place where the dryer group is located, and an integral weight value updating formula is formulated according to the completion condition of the integral gas drying operation task of the dryer group;
Constructing a weight matrix for interacting with the internal state table; the weight matrix is used for storing estimated weight values of each time node action pair. The number of rows of the matrix table is the same as the number of the time nodes and is named according to the sequence numbers of the time nodes; the weight matrix table is 3 columns in total, and is respectively: tank a starts to regenerate, tank B starts to regenerate and no action is performed.
According to the state of the dryer corresponding to each time node in the internal state table, outputting a new weight value according to the change of the state by matching with the monomer and the whole weight updating formula to update the new weight value to a weight matrix, reaching the next time node, and repeating the process until all the time nodes are traversed;
And when the consumed electricity cost is lowest in the output cycle, a corresponding state table and a weight matrix table. The internal state table corresponds to the optimal scheduling result of the dryer group, and the weight matrix records the optimal action output of each time node in the optimal scheduling.
Preferably, the working time is the total running time length of the dryer group to be worked;
The discretization into time nodes includes: the total continuous time is divided into a plurality of time segments by taking the whole hour as a boundary, and the time segments are sequentially marked as time nodes in turn.
Preferably, the rows corresponding to the states of the dryer have two rows, and the two rows correspond to the states of the tank a and the tank B in the dryer, respectively.
Preferably, the state of the tank a and the state of the tank B have 3 state values including drying, regeneration and no action. The corresponding tank is in gas drying operation, tank regeneration operation and standby state at the corresponding time nodes.
Preferably, according to the state of the dryer corresponding to each time node in the internal state table, a new weight value is output according to the change of the state by matching with the monomer and the integral weight updating formula to update to a weight matrix, and the method specifically comprises the following steps:
Firstly, searching a current time node in the internal state table, and judging whether at least one tank body of the tank body A and the tank body B is in regeneration operation;
if not, not updating the monomer weight value;
If yes, corresponding weight updating values w 1、w2 and w 3 are determined according to the peak period, the flat period and the valley period divided in the electric charge interval table;
Let w out equal to the corresponding weight update value w 1、w2 or w 3, and according to the formula Outputting new monomer weight values to the weight matrix;
in the above formula, α is the learning rate; gamma is a prize discount;
The overall weight update is to ensure that the dryer group is internally collaborative. Firstly, determining an overall weight update value according to the insufficient times of the dryer in one complete cycle. Determining an overall weight update value according to the number of times of shortage of the dryer in one complete cycle; then updating the corresponding overall weight value according to the formula W (s, a) ≡W (s, a) +alpha [ all_w-W (s, a) ]; alpha is learning rate and takes a value of 0.02.
Preferably, the present invention refines the time node. The current continuous time division mode is full-point division, so that the dryer group can only switch states when full points are formed, and the scheduling flexibility can be improved by refining time nodes.
Preferably, the invention refines the two stages of gas drying and tank regeneration inside the dryer. The tank regeneration operation process comprises the following front-end and rear-end small processes, and the power consumption of the parts is relatively low.
Preferably, the present invention may use a neural network instead of a weight matrix table. When the number of days of a single cycle is too long, the dimension of the weight matrix table is increased, and the operation amount of the algorithm is also increased. While neural networks can handle problems with a large number of features or high dimensional inputs.
Compared with the prior art, the invention has the following remarkable improvements:
(1) The invention adopts a mode of updating the weight matrix table to solve the scheduling method of the dryer group, thereby achieving the aim of reducing the expenditure of electricity charge to the maximum extent on the premise of not influencing normal gas drying operation.
(2) The invention can formulate a monomer weight updating formula and an integral weight updating formula according to the electric charge interval table of the place where the dryer group is located and the gas drying task completion condition of the whole dryer group; thereby designing an independent weight matrix for all dryers within the dryer group.
(3) In the invention, the change of the internal value of the internal state table is matched with the updating formula according to the weight, so that a new weight value can be output to the weight matrix; and then the internal value of the weight matrix corresponding to each drier in the drier group can be optimized through multiple times of circulation, and then the optimized scheduling method of the drier group is output according to the weight matrix.
Drawings
FIG. 1 is a system block diagram of a dryer group according to the present invention;
FIG. 2 is a table showing the internal states of the dryer according to the present invention;
FIG. 3 is a weight matrix of dryer action output in the present invention;
FIG. 4 is an overall flow chart of an optimization algorithm in the present invention;
FIG. 5 is a diagram showing the comparison of the electric charges after training in the present invention;
FIG. 6 is a cross-sectional view of the present invention for the internal scheduling of a dryer group.
Detailed Description
The invention relates to a dryer group scheduling optimization algorithm based on a weight matrix, which comprises the following steps:
Dividing the continuous time of the total required work of the dryer group into a plurality of time nodes by taking 1 hour as a scale; drawing a state table for all dryers in the dryer group by combining the time nodes; a monomer weight updating formula and an integral weight updating formula are formulated according to an electric charge interval table of the place where the dryer group is located and the air drying task completion condition of the whole dryer group; designing action weight matrixes for all dryers in the dryer group, and making action output rules of the weight matrixes according to the internal operation flow of the dryers. Each dryer outputs actions to a state table according to the respective weight matrix and the output rule of the actions, the internal value of the state table is changed, a new weight value is output to the weight matrix according to a weight updating formula, the internal value of the weight matrix corresponding to each dryer in the dryer group is optimal through repeated circulation, and then the optimal scheduling method of the dryer group is output according to the weight matrix, so that the aim of reducing electricity charge under the premise of not influencing normal gas drying operation is achieved. The steps of the present invention will be described in detail by way of an example.
S01, 7 dryers are arranged on site, the continuous operation is carried out for 10 days, the regeneration operation power of each dryer is 150kw, the gas drying time is 8 hours, the regeneration heating time is 3 hours, and the electric charge planning table is shown in the table 1.4 dryers and 3 standby machines are needed in working conditions, and the circulation starting time is 0 point of the first day.
TABLE 1 electric charge planning sheet
S02, since the continuous loop operation is performed for 10 days, the total number of nodes is 10×24=240, and the number of rows of the dryer state table and the weight matrix is 240. The state table and the weight matrix table of each dryer are shown in fig. 2 and 3, and the internal values of both tables are all 0 in the initial state.
S03, determining three weight levels according to the peak period, the flat period and the valley period of the electric charge division table, formulating corresponding weight updating values w 1＝0、w2 =1 and w 3 =2, and according to a formulaNew weight values are output to the weight matrix. After traversing all time nodes, obtaining an overall weight updating value W all according to the insufficient times of the dryer in the current circulation, and then updating the corresponding weight value for the last time according to a formula W (s, a) ≡W (s, a) +alpha [ W all -W (s, a) ].
In this embodiment, α is a learning rate, and the value is preferably 0.02; gamma is a discount of rewards, and the value is preferably 0.9;
S04, starting circulation, inputting the state of the current time node into a weight matrix table by an internal state table, outputting an action value to the state table by the weight matrix table according to the state of the current time node, outputting a new weight value to update the weight matrix by a single body and an integral weight updating formula according to the action value, outputting a new weight value to update the weight matrix according to the change of the state, reaching the next time node, and repeating the process until all time nodes are traversed. The overall flow is shown in fig. 4.
S05, after the circulation is finished, outputting all results without insufficient dryer in the training process, wherein the horizontal axis represents successful serial numbers and the vertical axis represents electric charge as shown in fig. 5. In the cycle, the optimal scheduling result shows that the electricity charge can be saved by more than 40 ten thousand yuan each year.
Meanwhile, the scheduling condition in the dryer is output, as shown in fig. 6, the horizontal axis is a time node, the vertical axis is the number of the dryer, and after each number, two horizontal lines are arranged in the upper and lower directions, and the two horizontal lines correspond to two tanks in the dryer respectively. The red line segment represents that the corresponding tank body is subjected to gas drying operation, and the length is 8, which means that the gas drying operation is performed for 8 hours; the green line represents that the corresponding can is being subjected to a can recycling operation, with a length of 3, meaning that 3 hours of can recycling operation are being performed.
It is to be understood that, based on one or several embodiments provided in the present application, those skilled in the art may combine, split, reorganize, etc. the embodiments of the present application to obtain other embodiments, which do not exceed the protection scope of the present application.
The invention and its embodiments have been described above by way of illustration and not limitation, and the examples are merely illustrative of embodiments of the invention and the actual construction is not limited thereto. Therefore, if one of ordinary skill in the art is informed by this disclosure, the structural mode and the embodiments similar to the technical scheme are not creatively designed without departing from the gist of the present invention.
Claims (5)
1. The dryer group scheduling optimization algorithm based on the weight matrix is characterized by comprising the following steps of:
the working time of the dryer group to be scheduled is discretized into time nodes;
taking the time node as a row and the state of the dryer as a column, and constructing an internal state table of each dryer;
A monomer weight value updating formula is formulated according to a pre-acquired electricity fee interval table of the place where the dryer group is located, and an integral weight value updating formula is formulated according to the completion condition of the integral gas drying operation task of the dryer group;
constructing a weight matrix for interacting with the internal state table;
According to the state of the dryer corresponding to each time node in the internal state table, outputting a new weight value to update to a weight matrix according to the change of the state by matching with the single weight value updating formula and the whole weight value updating formula, reaching the next time node, and repeating the process until all the time nodes are traversed;
and when the consumed electricity cost is lowest in the output cycle, a corresponding internal state table and a weight matrix.
2. The dryer group scheduling optimization algorithm based on the weight matrix of claim 1, wherein: the working time is the total running time length of the dryer group to be worked;
The discretization into time nodes includes: the total continuous time is divided into a plurality of time segments by taking the whole hour as a boundary, and the time segments are sequentially marked as time nodes in turn.
3. The dryer group scheduling optimization algorithm based on the weight matrix of claim 1, wherein: the rows corresponding to the states of the dryer are two rows, and the two rows respectively correspond to the states of the tank A and the tank B in the dryer.
4. A dryer group scheduling optimization algorithm based on a weight matrix according to claim 3, wherein: the state of the tank a and the state of the tank B have 3 state values including drying, regeneration and no action.
5. The dryer group scheduling optimization algorithm based on the weight matrix of claim 4, wherein: according to the state of the dryer corresponding to each time node in the internal state table, a new weight value is output according to the change of the state by matching with the monomer and the integral weight updating formula to update to a weight matrix, and the method specifically comprises the following steps:
firstly, searching a current time node in the internal state table, and judging whether at least one tank body of the tank body A and the tank body B is in regeneration operation;
if not, not updating the monomer weight value;
If so, determining corresponding weight updating values according to the peak time periods, the flat sections and the valley time periods divided in the electric charge interval table 、And；
Make the following stepsEqual to the corresponding weight update value、Or (b)And according to the formulaOutputting new monomer weight values to the weight matrix;
determining an overall weight update value according to the number of times of shortage of the dryer in one complete cycle; then according to the formula And updating the corresponding overall weight value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410155140.XA CN118226808B (en) | 2024-02-04 | 2024-02-04 | Dryer group scheduling optimization algorithm based on weight matrix |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410155140.XA CN118226808B (en) | 2024-02-04 | 2024-02-04 | Dryer group scheduling optimization algorithm based on weight matrix |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118226808A CN118226808A (en) | 2024-06-21 |
| CN118226808B true CN118226808B (en) | 2024-08-30 |
Family
ID=91496871
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410155140.XA Active CN118226808B (en) | 2024-02-04 | 2024-02-04 | Dryer group scheduling optimization algorithm based on weight matrix |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118226808B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109949604A (en) * | 2019-04-01 | 2019-06-28 | 南京邮电大学 | A kind of large parking lot scheduling air navigation aid, system and application method |
| CN110968425A (en) * | 2019-11-22 | 2020-04-07 | 中盈优创资讯科技有限公司 | Dynamic allocation method and system for task resources |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110984062B (en) * | 2019-12-20 | 2020-12-29 | 华中科技大学 | Simulation scheduling method for large-scale reservoir group of watershed main and branch flows |
| CN115270627A (en) * | 2022-07-28 | 2022-11-01 | 湖南大学 | Ship section manufacturing iterative optimization scheduling method and system based on artificial neural network |
-
2024
- 2024-02-04 CN CN202410155140.XA patent/CN118226808B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109949604A (en) * | 2019-04-01 | 2019-06-28 | 南京邮电大学 | A kind of large parking lot scheduling air navigation aid, system and application method |
| CN110968425A (en) * | 2019-11-22 | 2020-04-07 | 中盈优创资讯科技有限公司 | Dynamic allocation method and system for task resources |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118226808A (en) | 2024-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107450498B (en) | Based on the production scheduling method and system for improving artificial bee colony algorithm | |
| CN104268722B (en) | Dynamic flexible job-shop scheduling method based on multi-objective Evolutionary Algorithm | |
| CN109946965A (en) | A kind of Discrete Production Workshop scheduled production method based on improved multiple target Jaya algorithm | |
| CN108710970B (en) | Multi-target scheduling parallel dimension reduction method for giant cascade hydroelectric system | |
| CN113222446A (en) | Distributed system scheduling optimization method based on unreliable machine and limited buffer | |
| CN117314055A (en) | Intelligent manufacturing workshop production-transportation joint scheduling method based on reinforcement learning | |
| CN116360355A (en) | Method for solving workshop scheduling problem based on NSGA-III algorithm | |
| CN116933939A (en) | Flexible workshop collaborative production method and system based on improved raccoon optimization algorithm | |
| CN118192472A (en) | Improved sparrow optimization method for scheduling problem of flexible job shop | |
| CN117707083A (en) | Scheduling method, terminal equipment and storage medium for distributed assembly line shop | |
| CN118226808B (en) | Dryer group scheduling optimization algorithm based on weight matrix | |
| CN114386843A (en) | Flexible workshop scheduling method based on improved deep reinforcement learning algorithm | |
| CN117891220A (en) | Distributed mixed flow shop scheduling method based on multi-agent deep reinforcement learning | |
| CN111290360B (en) | Multi-objective optimization method for casting production line | |
| CN115081755B (en) | Production and maintenance cooperative scheduling method and system based on variable neighborhood search algorithm | |
| CN112148446A (en) | Evolutionary strategy algorithm for multi-skill resource limited project scheduling | |
| CN112488543A (en) | Intelligent work site shift arrangement method and system based on machine learning | |
| CN116562584A (en) | Dynamic workshop scheduling method based on Conv-lasting and generalization characterization | |
| CN113723695B (en) | Remanufacturing scheduling optimization method based on scene | |
| CN116011723A (en) | Intelligent dispatching method and application of coking and coking mixed flow shop based on Harris eagle algorithm | |
| CN115496322A (en) | Distributed flow shop scheduling method and device | |
| CN115344011A (en) | Energy-saving distributed two-stage assembly zero-waiting flow shop scheduling problem optimization system | |
| CN114237166A (en) | Method for solving multi-rotating-speed energy-saving scheduling problem based on improved SPEA2 algorithm | |
| CN113792924B (en) | Single-piece job shop scheduling method based on Deep Q-network Deep reinforcement learning | |
| CN117829552B (en) | Robust optimization method, device and equipment based on Ru production scheduling |
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 |