CN105787509B - A kind of pre- distribution of iron ore blending process - Google Patents
A kind of pre- distribution of iron ore blending process Download PDFInfo
- Publication number
- CN105787509B CN105787509B CN201610105488.3A CN201610105488A CN105787509B CN 105787509 B CN105787509 B CN 105787509B CN 201610105488 A CN201610105488 A CN 201610105488A CN 105787509 B CN105787509 B CN 105787509B
- Authority
- CN
- China
- Prior art keywords
- slot
- total
- iron ore
- class
- raw material
- 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
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 title claims abstract description 116
- 238000000034 method Methods 0.000 title claims abstract description 65
- 229910052742 iron Inorganic materials 0.000 title claims abstract description 58
- 238000002156 mixing Methods 0.000 title claims abstract description 41
- 230000008569 process Effects 0.000 title claims abstract description 33
- 238000009826 distribution Methods 0.000 title claims abstract description 21
- 239000000463 material Substances 0.000 claims abstract description 104
- 238000005457 optimization Methods 0.000 claims abstract description 27
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 22
- 239000000203 mixture Substances 0.000 claims abstract description 20
- 229910052729 chemical element Inorganic materials 0.000 claims abstract description 12
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 claims abstract description 7
- 229910052710 silicon Inorganic materials 0.000 claims abstract description 7
- 239000010703 silicon Substances 0.000 claims abstract description 7
- 239000004615 ingredient Substances 0.000 claims description 64
- 239000002994 raw material Substances 0.000 claims description 61
- 239000000470 constituent Substances 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 2
- 230000007717 exclusion Effects 0.000 claims description 2
- 238000005194 fractionation Methods 0.000 claims description 2
- 230000004044 response Effects 0.000 claims description 2
- 238000012163 sequencing technique Methods 0.000 claims description 2
- OYPRJOBELJOOCE-UHFFFAOYSA-N Calcium Chemical compound [Ca] OYPRJOBELJOOCE-UHFFFAOYSA-N 0.000 claims 1
- 229910052791 calcium Inorganic materials 0.000 claims 1
- 239000011575 calcium Substances 0.000 claims 1
- 238000000354 decomposition reaction Methods 0.000 claims 1
- 235000013399 edible fruits Nutrition 0.000 claims 1
- 229920001296 polysiloxane Polymers 0.000 claims 1
- 239000007787 solid Substances 0.000 claims 1
- 239000000126 substance Substances 0.000 abstract description 16
- 238000005520 cutting process Methods 0.000 abstract description 3
- 238000007781 pre-processing Methods 0.000 abstract description 2
- 238000004519 manufacturing process Methods 0.000 description 23
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 17
- 229910000831 Steel Inorganic materials 0.000 description 6
- 239000010959 steel Substances 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 229910052500 inorganic mineral Inorganic materials 0.000 description 3
- 239000011707 mineral Substances 0.000 description 3
- 241001269238 Data Species 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 241001417490 Sillaginidae Species 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000003556 assay Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- FFBHFFJDDLITSX-UHFFFAOYSA-N benzyl N-[2-hydroxy-4-(3-oxomorpholin-4-yl)phenyl]carbamate Chemical compound OC1=C(NC(=O)OCC2=CC=CC=C2)C=CC(=C1)N1CCOCC1=O FFBHFFJDDLITSX-UHFFFAOYSA-N 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000005245 sintering Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2431—Multiple classes
-
- 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
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Business, Economics & Management (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Health & Medical Sciences (AREA)
- Manufacturing & Machinery (AREA)
- Probability & Statistics with Applications (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Manufacture And Refinement Of Metals (AREA)
Abstract
The present invention relates to the technical fields such as data processing, clustering algorithm, Combinatorial Optimization more particularly to a kind of pre- distributions of iron ore blending process.The invention proposes a kind of pre- distribution of the iron ore blending process based on clustering algorithm and Combinatorial Optimization, step includes: material composition data preprocessing operation；A cluster operation is completed according to silicon content difference in material composition, i.e., is divided into N major class using Canopy-Kmeans clustering method；On this basis, secondary cluster is carried out using K-means algorithm according to iron content difference；Consider the constraint conditions such as chemical element index, classification results are handled using the method for merging and decomposing；Then, Optimum panel cutting order is determined with Combinatorial Optimization thought, obtain final proportion scheme.The pre- distribution of iron ore blending process provided by the invention saves and calculates time-consuming, it is as small as possible to guarantee that per tour mixes material chemical component error, and mix between class and class and expect that the variance fluctuation range of chemical component is small, precision is higher.
Description
Technical field
The present invention relates to the technical fields such as data processing, clustering algorithm, Combinatorial Optimization more particularly to a kind of iron ore to mix
The pre- distribution of journey.
Background technique
Currently, steel production in China industry is all constantly being promoted in the demand of steel production quantity and quality, arrive
2014 stop, and the iron and steel output share in China has occupied the 45% of the whole world.But due to starting late, China is in terms of steel production
Technology there are larger gaps still between developed country.And stock yard blending process, first of work as steel production
Sequence is different by a variety of sources and component difference is biggish by limited mixing facility according to windrow mission requirements monthly
Iron-bearing mineral is uniformly mixed by different ratio, so that being able to the entire moon be finished and guarantee that the whole month appoints to the various iron ore raw materials of factory
The chemical component what class mixes material is uniform and stable, i.e. per tour mixes material in the fluctuation range that enterprise and state quota require, and
The chemical component variance fluctuation that material is mixed between class and class is as small as possible, meets follow-up sintering and smelts iron the production technology used and wants
It asks, and then obtains final proportion scheme.The key of stock yard blending process is to determine in blending process the object of slot ingredient altogether
Material combination and their blanking order.And monthly iron-bearing mineral to be mixed is many kinds of, mineralogical composition differs greatly, and
Dosage bunker limited amount, it is ensured that the chemical component of per tour mixing material is uniform and stable extremely difficult in the whole month, still lacks so far full
The mixing dosage calculation method of sufficient requirement of real-time.Can blending process proportion calculates promptly and accurately be directly related in mixing material
Whether the chemical components such as silicon, iron meet subsequent production requirement, and then influence the production efficiency and product matter of entire iron and steel enterprise
Amount, therefore a kind of real-time calculation method for iron ore blending process is needed, the calculation method in the field is broadly divided at present
Based on artificial experience calculate mixing dosage method and based on the mixing dosage method of fuzzy theory Modeling Calculation.
Summary of the invention
(1) technical problems to be solved
The technical problem to be solved by the present invention is to solve since many kinds of of iron ore raw material iron-bearing mineral, mineralogical composition are poor
Different larger, ingredient fluctuates big and dosage bunker limited amount, it cannot be guaranteed that the chemical component that per tour mixes material in the whole month is uniformly asked
Topic.
(2) technical solution
In order to solve the above-mentioned technical problems, the present invention provides a kind of pre- distributions of iron ore blending process comprising
Following step:
Step 1: according to the of that month various iron ore raw material compositional datas for arriving factory, carrying out pretreatment operation, excludes silicon, iron member
Plain component difference is larger or its physicochemical characteristics can not be total to the special material of slot, obtains raw material collection to be clustered；
Step 2: the raw material collection to be clustered that step 1 obtains once is gathered using Canopy-Kmeans clustering algorithm
Generic operation.The division that major class is carried out according to primary constituent content makes it be divided into reasonable N number of major class automatically, reaches primary element
Balance；
Step 3: secondary cluster is carried out using K-Means algorithm to the primary cluster raw material set obtained by step 2
Operation.Cluster again is carried out according to minor element content in contained material composition in each major class and realizes that group divides, full
Under the premise of the primary element of foot is in a basic balance, the ingredient for reaching minor element in each group is substantially approximate；
Step 4: the thought that each small class a operation merges and decomposes obtained in step 3 is handled.Consider practical
Ingredient slot number, mixing material each element index, the constraint conditions such as several and total siccative amount of arranging an order according to class and grade in production, to gained cluster result and spy
Different raw material is merged and is automatically brought into operation with decomposing etc., is determined final clustering schemes, is guaranteed that all kinds of middle raw material types, quantity meet in fact
Border production requirement ensures finally to cluster number and is equal to available ingredient slot number, every class raw material type is no more than the raw material that can be total to slot
The total upper limit, while meeting the requirement of other production constraints.Every class raw material participates in ingredient in a dosage bunker as a result, to total slot
Plurality of raw materials determine blanking precedence in slot at random, and then generate initial proportion scheme；
Step 5: total slot raw material blanking is carried out to initial ingredients scheme obtained in step 4 with combined optimization method
The optimization of sequence determines final proportion scheme.Consider that raw material type is various, ingredient fluctuates the shadows of the factors such as larger and constraint condition is more
It rings, each class, which is total to the successive arrangement order of slot ingredient, to be determined with the method for exhaustion in Combinatorial Optimization for each raw material in total slot,
So that the mixing material ingredient of any shift of the entire moon fluctuates within the scope of the minimal ripple of plan component target, final match is obtained
Material scheme.
The pretreated process of raw material compositional data in step 1 are as follows:
It differs greatly for silicon, ferro element component content or plans the biggish raw material of proportion, because that can not be total to unclassified stores
Slot is excluded, it is prevented to be added in cluster sample；For raw material physicochemical characteristics such as toughness, aqueous itself
Amount etc. leads to not total slot, this kind of material is distinguished and excluded according to practical production experience and feedstock assay result, hinders
Only it is added in cluster sample.
The process for using Canopy-Kmeans clustering algorithm once to be clustered raw material set to be clustered in step 2
Are as follows:
According to constituent content difference primary in material composition, using Canopy-Kmeans clustering technique by raw material to be clustered
Collection is automatically divided into reasonable N number of major class, reaches the balance of primary constituent content, avoids artificially specified number of clusters and drops
The robustness of low method；Such as feedstock property used in different iron-smelters and blanking require to be different, it need to be according to its raw material
The feature of compositional data carries out intelligent selection to the quantity and initial point of cluster, is overcome using Canopy-Kmeans clustering technique
The shortcomings that traditional K-Means method falls into local optimum when number of clusters and initial point are chosen.Specific step is as follows:
Respectively enter pretreated in the primary constituent content deposit data set of factory's iron ore material, selects two threshold value T1And T2,
Wherein T1>T2.One point P of arbitrary extracting again calculates P point at a distance from Canopy, if P point is in the distance T of some Canopy1With
It is interior, then this Canopy is added in P point；If the P point is once in the distance T of some Canopy2Within, then P from data set
It deletes, repeats above-mentioned two step until data set is sky, the final number N for counting Canopy class passes through Canopy clustering method
Legacy data is divided into N number of major class.
Carry out the process of secondary cluster operation in step 3 using K-Means algorithm to primary cluster raw material collection are as follows:
By once clustering in the N major class material obtained, the K-Means in class is carried out to minor element in each major class and is gathered
Material in major class is divided into two groups by class, so that minor element ingredient gap in the approximate material of primary element is become smaller, by institute
Two group data being divided to temporarily are assigned in total slot memory space, to reach the substantially approximate mesh of all kinds of middle minor elements
's；
Step 4 handles the thought that each small class a operation merges and decomposes obtained in step 3.Consider practical
Ingredient slot number, mixing material each element index, the constraint conditions such as several and total siccative amount of arranging an order according to class and grade in production, to gained cluster result and spy
Different raw material is merged and is automatically brought into operation with decomposing etc., is determined final clustering schemes, is guaranteed that all kinds of middle raw material types, quantity meet in fact
Border production requirement ensures finally to cluster number and is equal to available ingredient slot number, every class raw material type is no more than the raw material that can be total to slot
The total upper limit, while meeting the requirement of other production constraints.Every class raw material participates in ingredient in a dosage bunker as a result, to total slot
Plurality of raw materials determine blanking precedence in slot at random, and then generate initial proportion scheme.The upper limit of the number for remembering slot altogether is k
It is a:
1. illustrate that its primary element and minor element and unclassified stores difference are big when material number is 1 in group, it can not
Cluster, then put it into single slot, but if the discharge quantity of the material is lower than Mmin, then ingredient cannot individually be carried out, it is necessary to find
Total slot processing is carried out with its suitable material, or when the material purchase volume is smaller, is added using the characteristic for itself matching small
It is added to the very end of the train for matching biggish material in certain single slot, minimizes it to the influence power of bulk chemical element index；
2. being divided into two groups after clustering according to minor element has much when the material number of major class is between k+1~2k
Kind of combinatory possibility, such as: if k is 3, possible combination have 23,22,13,33, these may be combined, due to
Every class material number is few, it may be considered that is directly total to slot；Other the case where, such as 14,4 have been more than the number upper limit k=of total slot
3, then it needs to carry out again for that 4 materials once to cluster and assign in 2 total slots；
3. then such preceding k kind material is moved in other total slots, then sentences when the material number of major class is greater than 2k kind
Disconnected major class surplus material number, if being still greater than 2k kind so continues to execute above step；If it is less than 2k kind, then jump to
1 and 2 steps are stated to be handled.
By the preliminary optimization of dosage bunker material number, slot material number can be total to by making in each dosage bunker meets actual production
It is required that further, it is contemplated that can be used in actual production ingredient slot number constraint, be respectively processed:
1. if the usage quantity of dosage bunker and enterprise practical ingredient slot number are just identical after above-mentioned preliminary optimization,
Without current procedures processing；
2. being total to if the usage quantity of dosage bunker is less than enterprise practical ingredient slot number after above-mentioned preliminary optimization
The fractionation of slot；Treating method in response to this are as follows: choose k and be total to slot, maximum material will be matched in slot in k kind material and moved
Out, it puts into single slot, so that original k is total to slot and is changed into a k-1 and be total to the single slot of slot and one, dosage bunker usage quantity is made to increase by 1
It is a, and so on, it corrects with ensure that the minimum cost of original ingredient scheme while increasing dosage bunker；
3. if the usage quantity of dosage bunker is greater than enterprise practical ingredient slot number after above-mentioned preliminary optimization, to ingredient
It is combined between slot, there are mainly two types of exchanges combined method:
Slot single in the proportion scheme preliminarily formed is analyzed, does not consider the material that must carry out single slot processing, and unite
Meter can redistribute the number of material, and the material that can be redistributed is ranked up according to the ratio, select two kinds of the smallest objects of proportion
Expect slot altogether, dosage bunker usage quantity is made to reduce 1；
As described above, statistics can redistribute the number of material, if can redistribute in material only have 1 can be into
The total slot of row then combines it with other 2 total slots, makees 3 slot processing altogether, make dosage bunker usage quantity reduction 1；It can also similarly incite somebody to action
Material in single slot is put into 3 total slots, is constituted 4 total slots, need to only be met and be no more than total slot upper limit k.
Step 5 carries out total slot raw material blanking to initial ingredients scheme obtained in step 4 with combined optimization method
The optimization of sequence:
The arrangement order being randomly generated according to step 4 counts the index situation of current allocation result, judges whether to meet
The production requirement of current enterprise, and the fluctuation situation of index is analyzed, material information is constant in guaranteeing original slot altogether
Under the premise of, fully intermeshing analysis is carried out by its possible blanking sequencing to current all slot materials altogether, after arranging every time
Material combinations carry out index verification, and assessment made to the index, the optimum combination mode that last index for selection is assessed makes to work as
Preceding index reaches global optimum, obtains the optimal solution of material blanking order in dosage bunker, so that the entire moon is former to the various iron ores of factory
Material is able to be finished and guarantee that any class of the whole month chemical component for mixing material is uniform and stable, i.e. per tour mixes material and refers in enterprise and country
It marks in desired fluctuation range, and the chemical component variance fluctuation that material is mixed between class and class is as small as possible, and then obtains final
Proportion scheme.
This patent generates mixing dosage scheme by combining clustering algorithm and combined optimization method quick, intelligently.It is counting
Time-consuming aspect is calculated, final proportion scheme can be obtained within this patent whole process 1~2 second, and most of factories of China manually calculate time-consuming 2
It differs within~4 hours or even 5 hours can not all acquire satisfactory solution；In precision aspect, the present invention can make per tour index raw in enterprise
It produces in desired fluctuation range, while guaranteeing that the variance fluctuation for mixing material chemical component between class and class is as small as possible, and it is artificial
Calculating is difficult to meet all indexs simultaneously, or even the larger situation of chemical component fluctuation occurs.This patent is big using clustering algorithm
Width reduces solution room, and under the premise of considering constraints a variety of in actual production, makes to gather in conjunction with the thought of Combinatorial Optimization
Result after class reaches global optimum, and then effectively instructs actual production process.It is now most of that the present invention solves China
The pre- ingredient link of factory automates generation on a large scale to industrial production and pushes and make still using the artificial status for calculating proportion
With effectively improving the speed and precision of pre- ingredient, lay good basis for enterprise's subsequent production link.
(3) beneficial effect
Above-mentioned technical proposal of the invention has the advantages that the pre- ingredient side of iron ore blending process provided by the invention
Method saves and calculates time-consuming, it is as small as possible to guarantee that per tour mixes material chemical component error, and mix between class and class and expect chemical component
Variance fluctuation range it is small, precision is higher.
Detailed description of the invention
Fig. 1 is the step schematic diagram of the pre- distribution of iron ore blending process of the embodiment of the present invention；
Fig. 2 is the pre- distribution intelligent clustering and combinatorial optimization algorithm process of iron ore blending process of the embodiment of the present invention
Figure；
Fig. 3 is the pre- distribution Canopy algorithm principle figure of iron ore blending process of the embodiment of the present invention；
Fig. 4 is the pre- distribution Canopy-Kmeans clustering algorithm flow chart of iron ore blending process of the embodiment of the present invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiments of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, ordinary skill people
Member's every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
Embodiment 1
As shown in Figure 1 to 4, a kind of pre- distribution of the iron ore blending process proposed in the present invention,
Step 1: according to the of that month various iron ore raw material compositional datas for arriving factory, carrying out pretreatment operation, excludes silicon, iron member
Plain component difference is larger or its physicochemical characteristics can not be total to the special material of slot, obtains raw material collection to be clustered；
Step 2: the raw material collection to be clustered that step 1 obtains once is gathered using Canopy-Kmeans clustering algorithm
Generic operation.The division that major class is carried out according to primary constituent content makes it be divided into reasonable N number of major class automatically, reaches primary element
Balance；
Step 3: secondary cluster is carried out using K-Means algorithm to the primary cluster raw material set obtained by step 2
Operation.Cluster again is carried out according to minor element content in contained material composition in each major class and realizes that group divides, full
Under the premise of the primary element of foot is in a basic balance, the ingredient for reaching minor element in each group is substantially approximate；
Step 4: the thought that each small class a operation merges and decomposes obtained in step 3 is handled.Consider practical
Ingredient slot number, mixing material each element index, the constraint conditions such as several and total siccative amount of arranging an order according to class and grade in production, to gained cluster result and spy
Different raw material is merged and is automatically brought into operation with decomposing etc., is determined final clustering schemes, is guaranteed that all kinds of middle raw material types, quantity meet in fact
Border production requirement ensures finally to cluster number and is equal to available ingredient slot number, every class raw material type is no more than the raw material that can be total to slot
The total upper limit, while meeting the requirement of other production constraints.Every class raw material participates in ingredient in a dosage bunker as a result, to total slot
Plurality of raw materials determine blanking precedence in slot at random, and then generate initial proportion scheme；
Step 5: total slot raw material blanking is carried out to initial ingredients scheme obtained in step 4 with combined optimization method
The optimization of sequence determines final proportion scheme.Consider that raw material type is various, ingredient fluctuates the shadows of the factors such as larger and constraint condition is more
It rings, each class, which is total to the successive arrangement order of slot ingredient, to be determined with the method for exhaustion in Combinatorial Optimization for each raw material in total slot,
So that the mixing material ingredient of any shift of the entire moon fluctuates within the scope of the minimal ripple of plan component target, final match is obtained
Material scheme.
Known parameters are set and require index fluctuation range；
It is expected that 92130 tons of total amount, average 4500 tons of per tour blanking, set and can be used for the slot number of ingredient as 11, Yi Jiyuan
Disk blanking velocity upper and lower bound；
Every content ingredient fluctuation range: primary element within ± 0.2%, minor element within ± 0.62% and
Other compositions requirement, chemical element fluctuation is as small as possible between class and class.
Table 1 arrives the various iron ore raw material data 1 of factory for certain month
Exclusion can not be total to after the special material of slot with unclassified stores, obtain raw material set to be clustered, then carry out to it
Canopy-Kmeans clustering, by material rough segmentation in current planning list at major class, then it is poly- using K-Means to each major class
After class algorithm is finely divided, obtain by secondary cluster as a result, as shown in table 2.
1.1 material cluster result of table
Result after secondary cluster is merged and resolution process, i.e. the 3rd class and the 4th class only have a kind of material, wherein
No. 11 material proportion amounts are greater than MminCan carry out single slot processing, however No. 20 materials then cannot single slot processing, can only select and its
The material of slot altogether, therefore original proportion is singly analyzed, guarantee as close as possible and do not violate physical constraint in ingredient
Under principle, selection is suitably total to slot material, i.e. material 4；
For statistical analysis to the usage quantity of dosage bunker after preliminary optimization, statistics actual use dosage bunker number is 9,
Lower than setting value 11, it is therefore desirable to 2 slots or 3 material carries out resolution process in slots altogether altogether, select No. 1 total slot and No. 3 total slots
It rejects wherein proportion the maximum to be put into single slot, generates initial proportion scheme.
With Combinatorial Optimization thought Optimization Cutting order, material chemical component fluctuation minimum and satisfaction are mixed meeting as far as possible
Under the premise of every constraint condition, the proportion blanking table finally optimized is obtained, (wherein dotted line is point of slot and slot as shown in table 3
Boundary line).
Certain moon of table 1.2 is to the final proportion scheme of factory's iron ore
Embodiment 2
Known parameters are set and require index fluctuation range；
It is expected that 92130 tons of total amount, average 4500 tons of per tour blanking, set and can be used for the slot number of ingredient as 11, Yi Jiyuan
Disk blanking velocity upper and lower bound；
Every content ingredient fluctuation range: primary element within ± 0.2%, minor element within ± 0.62% and
Other compositions requirement, chemical element fluctuation is as small as possible between class and class.
Certain moon of table 2 is to factory's iron ore raw material inventory 2
Certain moon of table 2.1 is to the final proportion scheme of factory's iron ore
Embodiment 3
Known parameters are set and require index fluctuation range；
It is expected that 98000 tons of total amount, average 5300 tons of per tour blanking, set and can be used for the slot number of ingredient as 11, Yi Jiyuan
Disk blanking velocity upper and lower bound；
Every content ingredient fluctuation range: primary element within ± 0.2%, minor element within ± 0.62% and
Other compositions requirement, chemical element fluctuation is as small as possible between class and class.
Certain moon of table 3 is to factory's iron ore raw material inventory 3
Certain moon of table 3.1 is to the final proportion scheme of factory's iron ore
Embodiment 4
Known parameters are set and require index fluctuation range；
It is expected that 92130 tons of total amount, average 4500 tons of per tour blanking, set and can be used for the slot number of ingredient as 11, Yi Jiyuan
Disk blanking velocity upper and lower bound；
Every content ingredient fluctuation range: primary element within ± 0.2%, minor element within ± 0.62% and
Other compositions requirement, chemical element fluctuation is as small as possible between class and class.
Certain moon of table 4 is to factory's iron ore raw material inventory 4
Certain moon of table 4.1 is to the final proportion scheme of factory's iron ore
Embodiment 5
Known parameters are set and require index fluctuation range；
It is expected that 98000 tons of total amount, average 4635 tons of per tour blanking, set and can be used for the slot number of ingredient as 11, Yi Jiyuan
Disk blanking velocity upper and lower bound；
Every content ingredient fluctuation range: primary element within ± 0.2%, minor element within ± 0.62% and
Other compositions requirement, chemical element fluctuation is as small as possible between class and class.
Certain moon of table 5 is to factory's iron ore raw material inventory 5
Certain moon of table 5.1 is to the final proportion scheme of factory's iron ore
Embodiment 6
Known parameters are set and require index fluctuation range；
It is expected that 98000 tons of total amount, average 5600 tons of per tour blanking, set and can be used for the slot number of ingredient as 11, Yi Jiyuan
Disk blanking velocity upper and lower bound；
Every content ingredient fluctuation range: primary element within ± 0.2%, minor element within ± 0.62% and
Other compositions requirement, chemical element fluctuation is as small as possible between class and class.
Certain moon of table 6 is to factory's iron ore raw material inventory 6
Certain moon of table 6.1 is to the final proportion scheme of factory's iron ore
Embodiment 7
Known parameters are set and require index fluctuation range；
It is expected that 92130 tons of total amount, average 4500 tons of per tour blanking, set and can be used for the slot number of ingredient as 11, Yi Jiyuan
Disk blanking velocity upper and lower bound；
Every content ingredient fluctuation range: primary element within ± 0.2%, minor element within ± 0.62% and
Other compositions requirement, chemical element fluctuation is as small as possible between class and class.
Certain moon of table 7 is to factory's iron ore raw material inventory 7
Certain moon of table 7.1 is to the final proportion scheme of factory's iron ore
By verifying, the index of the primary element of per tour and minor element is within enterprise practical claimed range, other yuan
There are individual over-standard phenomenons for element, but have complied fully with the demand of practical blanking.The result not only has compared with artificial calculated result
There is rapidity, also there is accuracy and operator's generality.
In conclusion the invention proposes a kind of pre- ingredient sides of the iron ore blending process based on clustering algorithm and Combinatorial Optimization
Method, step include: material composition data preprocessing operation；Primary cluster is completed according to silicon content difference in material composition
Operation, i.e., be divided into N major class using Canopy-Kmeans clustering method；On this basis, according to iron content difference benefit
Secondary cluster is carried out with K-means algorithm；Consider the constraint conditions such as chemical element index, using the method for merging and decomposing to point
Class result is handled；Then, Optimum panel cutting order is determined with Combinatorial Optimization thought, obtain final proportion scheme, this hair
The pre- distribution of the iron ore blending process of bright offer saves and calculates time-consuming, guarantees that per tour mixes material chemical component error and to the greatest extent may be used
Can be small, and the variance fluctuation range that material chemical component is mixed between class and class is small, precision is higher.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；
And these are modified or replaceed, technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (8)
1. a kind of pre- distribution of iron ore blending process, which is characterized in that comprise the steps of:
Step 1: pretreatment operation is carried out to iron ore raw material, exclusion elemental composition differs greatly or its physicochemical characteristics can not
The special material of slot altogether, obtains raw material collection to be clustered；
Step 2: the raw material set to be clustered is subjected to a cluster operation using Canopy-Kmeans clustering algorithm, is obtained
Primary cluster raw material set, the primary cluster raw material collection are combined into multiple major class of most important chemical element index balance；
Step 3: the primary cluster raw material set is subjected to secondary cluster operation using K-Means algorithm, obtains secondary cluster
Raw material set, the secondary cluster raw material collection are combined into the close multiple groups of minor element content in iron ore raw material ingredient；
Step 4: by the secondary cluster raw material set and the special material for ingredient slot number, mix material each element index,
Several and total siccative amount condition of arranging an order according to class and grade merges or operation splitting, obtains initial ingredients scheme, described to merge or decompose specific
Method is as follows, remembers that the upper limit of the number of slot altogether is k；
(1) when material number is 1 in group, illustrate that its primary element and minor element and unclassified stores difference are big, Wu Faju
Class is then put it into single slot, but if the discharge quantity of the material is lower than Mmin, then cannot individually carry out ingredient, it is necessary to find with
Its suitable material carries out total slot processing, or when the material purchase volume is smaller, is added using small characteristic itself is matched
The very end of the train that biggish material is matched into certain single slot, minimizes it to the influence power of bulk chemical element index；
(2) when the material number of major class is between k+1~2k, two groups are divided into after clustering according to minor element, if two groups small
Material number in class is less than k value, can directly be total to slot；If wherein the material number in one group of group is more than k value, it is directed to
This group of group, which carries out once clustering again, to be assigned in two total slots；
(3) when the material number of major class is greater than 2k kind, then such preceding k kind material is moved in other total slots, then judge
Major class surplus material number, if being still greater than 2k kind so continues to execute above step；If it is less than 2k kind, then jump to above-mentioned
Step (1) and step (2) are handled；
Step 5: total slot raw material blanking order is carried out to initial ingredients scheme obtained in step 4 with combined optimization method
Optimization, determines final proportion scheme.
2. the pre- distribution of iron ore blending process according to claim 1, it is characterised in that: the ingredient in the step 1
The element to differ greatly includes silicon, iron, calcium constituent.
3. the pre- distribution of iron ore blending process according to claim 1, it is characterised in that: the Canopy-Kmeans
Clustering algorithm carries out the division of major class according to primary element silicone content.
4. the pre- distribution of iron ore blending process according to claim 3, it is characterised in that: the Canopy-Kmeans
Clustering algorithm is two threshold value T of selection1And T2, wherein T1>T2, solid circles T1, virtual coil T2；One point P of arbitrary extracting again, meter
P point is calculated at a distance from Canopy, if P point is in the distance T of some Canopy1Within, then this Canopy is added in P point；Such as
The fruit P point is once in the distance T of some Canopy2Within, then P is deleted from data set, repeats above-mentioned two step until data set
For sky, the final number N for counting Canopy class.
5. the pre- distribution of iron ore blending process according to claim 1, it is characterised in that: to secondary member in each major class
Plain sheet content carries out K-Means cluster, and the material in major class is divided into two groups.
6. the pre- distribution of iron ore blending process according to claim 1, it is characterised in that: the side of the merging or decomposition
Method further includes following steps；
(1) it if usage quantity and practical ingredient slot number through above-mentioned initial ingredients scheme dosage bunker are just identical, does not repair
The just described initial ingredients scheme；
(2) it if the usage quantity through above-mentioned initial ingredients scheme dosage bunker is less than enterprise practical ingredient slot number, is total to
The fractionation of slot；Treating method in response to this are as follows: choose k and be total to slot, maximum material will be matched in slot in k kind material and moved
Out, it puts into single slot, so that original k is total to slot and is changed into a k-1 and be total to the single slot of slot and one, dosage bunker usage quantity is made to increase by 1
It is a, and so on, it corrects with ensure that the minimum cost of original ingredient scheme while increasing dosage bunker；
(3) if the usage quantity through above-mentioned initial ingredients scheme dosage bunker is greater than enterprise practical ingredient slot number, to ingredient
It is combined between slot by exchanging combined method, obtains final proportion scheme.
7. the pre- distribution of iron ore blending process according to claim 6, it is characterised in that: the exchange combined method has
Two kinds；
Method 1: slot single in final proportion scheme is analyzed, does not consider the material that must carry out single slot processing, and counting can
The material that can be redistributed is ranked up by the number for redistributing material according to the ratio, selects two kinds of the smallest materials of proportion total
Slot makes dosage bunker usage quantity reduce 1；
Method 2: the number of material can be redistributed by counting final proportion scheme, there was only 1 object in material if can redistribute
Material can carry out total slot, then it is total to slot with other 2 materials and combined, and make 3 materials and be total to slot processing, reduce dosage bunker usage quantity
1；The material in single slot can also be similarly put into 3 total slots, constitute 4 total slots, need to only meet and be no more than total slot upper limit k i.e.
It can.
8. the pre- distribution of iron ore blending process according to claim 1, which is characterized in that the combined optimization method
Are as follows: the index situation for counting current allocation result, under the premise of material information is constant in guaranteeing original slot altogether, to current all
Slot material carries out fully intermeshing analysis by its possible blanking sequencing altogether, and the material combinations after each arrangement are carried out index and are tested
Card, and assessment is made to the index, the optimum combination mode of last index for selection assessment makes current criteria reach global optimum,
The optimal solution of material blanking order in dosage bunker is obtained, and then obtains final proportion scheme.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610105488.3A CN105787509B (en) | 2016-02-25 | 2016-02-25 | A kind of pre- distribution of iron ore blending process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610105488.3A CN105787509B (en) | 2016-02-25 | 2016-02-25 | A kind of pre- distribution of iron ore blending process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105787509A CN105787509A (en) | 2016-07-20 |
| CN105787509B true CN105787509B (en) | 2018-12-14 |
Family
ID=56403033
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610105488.3A Active CN105787509B (en) | 2016-02-25 | 2016-02-25 | A kind of pre- distribution of iron ore blending process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105787509B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106484838A (en) * | 2016-09-30 | 2017-03-08 | 中国南方电网有限责任公司 | Safety inspection java standard library dynamic updating method based on data mining |
| CN108596375B (en) * | 2018-04-11 | 2021-07-27 | 武汉理工大学 | Equipment manufacturing material demand scheme design method and system |
| CN110681855A (en) * | 2018-07-06 | 2020-01-14 | 米亚索乐装备集成（福建）有限公司 | Preparation method of alloy powder |
| CN114118898A (en) * | 2021-10-22 | 2022-03-01 | 中冶南方工程技术有限公司 | Improved blending ingredient Block plan generation method and device and storage medium |
| CN114913932B (en) * | 2022-07-15 | 2022-10-14 | 北矿智云科技(北京)有限公司 | Batching method, batching device, electronic equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102193502A (en) * | 2011-06-14 | 2011-09-21 | 中南大学 | Comprehensive optimization method for steel sintering secondary burdening |
| CN104561411A (en) * | 2015-01-23 | 2015-04-29 | 武钢集团昆明钢铁股份有限公司 | Blending method capable of effectively improving quality of uniformly-mixed ores |
| CN104649036A (en) * | 2014-12-22 | 2015-05-27 | 天津钢铁集团有限公司 | Piling method for improving stability of uniformly blended material |
| CN105095266A (en) * | 2014-05-08 | 2015-11-25 | 中国科学院声学研究所 | Method and system for clustering optimization based on Canopy algorithm |
-
2016
- 2016-02-25 CN CN201610105488.3A patent/CN105787509B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102193502A (en) * | 2011-06-14 | 2011-09-21 | 中南大学 | Comprehensive optimization method for steel sintering secondary burdening |
| CN105095266A (en) * | 2014-05-08 | 2015-11-25 | 中国科学院声学研究所 | Method and system for clustering optimization based on Canopy algorithm |
| CN104649036A (en) * | 2014-12-22 | 2015-05-27 | 天津钢铁集团有限公司 | Piling method for improving stability of uniformly blended material |
| CN104561411A (en) * | 2015-01-23 | 2015-04-29 | 武钢集团昆明钢铁股份有限公司 | Blending method capable of effectively improving quality of uniformly-mixed ores |
Non-Patent Citations (1)
| Title |
|---|
| Canopy-Kmeans聚类和组合优化的铁矿预配料智能调度;曹跃 等;《控制理论与应用》;20170731;第34卷(第7期);947-955页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105787509A (en) | 2016-07-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105787509B (en) | A kind of pre- distribution of iron ore blending process | |
| Spence et al. | Setup reduction and increased effective capacity | |
| Wang et al. | Information processing structures and decision making delays in MRP and JIT | |
| CN102622667A (en) | Balance production scheduling method based on multi-product and multi-production-line production mode | |
| CN110728073A (en) | Multi-objective optimization method for coal washing and blending | |
| CN106406099B (en) | Cement batching system and method based on fuzzy matching sum value feedback | |
| CN104680012A (en) | Calculating model for sintering and burdening | |
| Monkiewicz | International technology flows and the technology gap: the experience of Eastern European socialist countries in international perspective | |
| CN104576441B (en) | Semiconductor package wiring LOT more rules dispatching methods based on operating area | |
| CN106094751B (en) | A kind of dispatching method and device of raw material | |
| Thenarasu et al. | Development and selection of hybrid dispatching rule for dynamic job shop scheduling using multi-criteria decision making analysis (MCDMA) | |
| CN106707994B (en) | Steel rolls dynamic debugging system and its method | |
| CN106395404B (en) | A kind of beating and double roasting sack is unloaded the method for regulation and control of homogenizing | |
| Arzi et al. | Real-time production control of an FMS in a produce-to-order environment | |
| Wang et al. | Multi-objective optimization model for seru production system formation under uncertain condition | |
| CN107918820A (en) | A kind of supply chain order collocation method for considering income | |
| Fernandes et al. | POLCA: centralised vs. decentralised job release | |
| CN106920004A (en) | A kind of one-dimensional stock-cutting method based on cost dynamic equilibrium | |
| Günther et al. | Workload planning in small lot printed circuit board assembly | |
| CN112349357B (en) | Long-term batching method and system for production of non-ferrous copper | |
| Yue et al. | Regional allocation of energy intensity reduction target: the case of Henan province in China | |
| Urnauer et al. | Pareto Heuristic for Product Family-Oriented Product-Workstation Allocation Planning with Restricted Capacities | |
| Adeyeye et al. | Goal programming model for production planning in a toothpaste factory | |
| CN109754186A (en) | Method for calculating probability and device | |
| CHUNG et al. | Using the virtual cell concept for short-term production planning for flexible job shop manufacturing systems |
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 | ||
| GR01 | Patent grant |