CN107808221A - Blast furnace material distribution Parameter Decision Making method based on case matching - Google Patents
Blast furnace material distribution Parameter Decision Making method based on case matching Download PDFInfo
- Publication number
- CN107808221A CN107808221A CN201711046346.5A CN201711046346A CN107808221A CN 107808221 A CN107808221 A CN 107808221A CN 201711046346 A CN201711046346 A CN 201711046346A CN 107808221 A CN107808221 A CN 107808221A
- Authority
- CN
- China
- Prior art keywords
- carbon monoxide
- utilization rate
- variable
- model
- model parameter
- 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
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"
-
- C—CHEMISTRY; METALLURGY
- C21—METALLURGY OF IRON
- C21B—MANUFACTURE OF IRON OR STEEL
- C21B5/00—Making pig-iron in the blast furnace
- C21B5/008—Composition or distribution of the charge
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Chemical & Material Sciences (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Organic Chemistry (AREA)
- Marketing (AREA)
- Metallurgy (AREA)
- Game Theory and Decision Science (AREA)
- Materials Engineering (AREA)
- Manufacturing & Machinery (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Blast Furnaces (AREA)
Abstract
The invention provides the blast furnace material distribution Parameter Decision Making method matched based on case, methods described includes step:Choose potential model parameter variable;Model parameter variable is pre-processed using box traction substation;Correlation analysis is carried out to pretreated model parameter variable and utilization rate of carbon monoxide;Count utilization rate of carbon monoxide variation tendency corresponding to the model parameter variable of utilization rate of carbon monoxide trend prediction;Setting sampling amount threshold value will simultaneously be contrasted by the utilization rate of carbon monoxide trend result of threshold extraction with legitimate reading;Then utilization rate of carbon monoxide trend precision is obtained;Setting accuracy threshold value;If precision of prediction meets precision threshold, model is valid model, is otherwise invalid model.The present invention is to instructing blast furnace material distribution parameter to have great importance.
Description
Technical field
The present invention relates to blast furnace ironmaking field, and in particular to the blast furnace material distribution Parameter Decision Making method based on case matching.
Background technology
Steel industry is the important component of Chinese national economy, relates generally to construction, the energy, automobile, basis set
The core industry of national development such as apply, be that China realizes national off, builds the powerful support of economic power.Given birth in blast furnace
During production, cloth is that it decides the distribution of furnace charge in stove most actively with crucial operational means, affects Gas Flow development
And blast furnace other states.But because blast furnace is a closed black box, detection information is limited, causes the pass of blast furnace material distribution and state
System is difficult to clearly, and scene relies primarily on artificial experience and cloth is adjusted.The scene of blast furnace possesses substantial amounts of history production number
According to containing blast furnace inner working rule.Therefore, depth excavate blast furnace historical data useful information, establish blast furnace material distribution with
The relational model of blast furnace state, so as to realize the blast furnace material distribution operation adjustment of blast fumance index optimization, just turn into industry and pay close attention to
Subject matter.
The content of the invention
The invention provides the blast furnace material distribution Parameter Decision Making method matched based on case, can effectively solve the above problems.
Technical scheme provided by the invention is:Based on the blast furnace material distribution Parameter Decision Making method of case matching, methods described bag
Include step:Choose potential model parameter variable;The model parameter variable is pre-processed using box traction substation
Model parameter variable afterwards;Correlation analysis is carried out to the pretreated model parameter variable with utilization rate of carbon monoxide to obtain
To the model parameter variable of utilization rate of carbon monoxide trend prediction;Count the model ginseng of the utilization rate of carbon monoxide trend prediction
Utilization rate of carbon monoxide variation tendency corresponding to number variable obtains utilization rate of carbon monoxide trend result;Set sample
Extraction amount threshold value and the utilization rate of carbon monoxide trend result as described in the sampling amount threshold extraction;It will extract
Utilization rate of carbon monoxide trend result contrasted to obtain with utilization rate of carbon monoxide variation tendency legitimate reading
Comparing result;Utilization rate of carbon monoxide trend precision is obtained with the comparing result;Setting accuracy threshold value;If institute
State utilization rate of carbon monoxide trend precision and meet precision threshold, then the model is valid model, can be used for height
The Parameter Decision Making of stove cloth;If the utilization rate of carbon monoxide trend precision is unsatisfactory for precision threshold, the mould
Type is invalid model, then finds new case again and be used to establish valid model.
The beneficial effects of the invention are as follows:The invention provides the blast furnace material distribution Parameter Decision Making method matched based on case, leads to
The foundation using utilization rate of carbon monoxide as decision-making cloth parameter is crossed, analysis utilization rate of carbon monoxide is (high with model parameter variable
Oven-like state variable) between relevance, it is determined that the input variable for case matching library.It is then based on cloth parameter and obtains model
The predicted value of parametric variable, it is finally based on the prediction mould that case matching process establishes the pre- variation tendency of utilization rate of carbon monoxide respectively
Type, effective prediction to utilization rate of carbon monoxide variation tendency is realized, to instructing blast furnace material distribution parameter to have great importance.
Brief description of the drawings
Fig. 1 is the overall flow figure of the blast furnace material distribution Parameter Decision Making method based on case matching in the embodiment of the present invention;
Fig. 2 is that box traction substation rejecting abnormalities model parameter variable method flow chart of steps is utilized in the embodiment of the present invention.
Embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with accompanying drawing to embodiment party of the present invention
Formula is further described, the particular technique details hereinafter mentioned, such as:Method, equipment etc., are only better understood from reader
Technical scheme, does not represent that present invention is limited only by following ins and outs.
The embodiment provides the blast furnace material distribution Parameter Decision Making method matched based on case.Refer to Fig. 1, Fig. 1
It is the overall flow figure of the blast furnace material distribution Parameter Decision Making method based on case matching in the embodiment of the present invention, methods described is by hardware
Equipment realizes that specific steps include:
S101:Potential model parameter variable is chosen, model parameter variable specifically includes:Stove cold flow, furnace wall thermic load,
Injecting coal quantity, material speed, silicone content, permeability index, ore coke ratio, Rich Oxygen Amount and furnace roof rise tube top temperature;The potential model parameter
Variable acquisition interval is 6 hours, takes 4 data daily.
S102:The model parameter variable is pre-processed using box traction substation to obtain pretreated model parameter change
Amount, is specifically included:Reject obvious abnormal model parameter variable.
S103:Correlation analysis is carried out to the pretreated model parameter variable and utilization rate of carbon monoxide and obtains an oxygen
Change the model parameter variable of carbon utilisation rate trend prediction.The utilization rate of carbon monoxide calculation formula is
Wherein, η represents utilization rate of carbon monoxide, Volume (CO2) it is CO in blast furnace roof increase in pipeline2Percent by volume, Volume (CO) for height
CO percent by volume in stove furnace roof increase in pipeline.The specific formula of correlation analysis is:
Wherein, I (X, Y) represents the association relationship between variable X and variable Y, is determined and carbon monoxide profit according to the size of mutual information
With the big model parameter variable of rate dependence, p (Xj,Ym) represent the joint probability distribution of variable X and variable Y, p (Xj) represent to become
Measure X marginal probability distribution, p (Ym) represent variable Y marginal probability distribution.Table 1 is different model parameter variables and an oxidation
Carbon utilisation rate association relationship.
Table 1
As can be seen from Table 1, have that five model parameter variables and utilization rate of carbon monoxide correlation are larger, therefore select heat
Load, top temperature, permeability index, oxygen-enriched, coal powder injection this mode input of five variables as utilization rate of carbon monoxide trend prediction.
Can be by artificially determining in view of oxygen-enriched and coal powder injection, therefore the model based on support vector regression only needs pre- calorimetric to bear
Lotus, top temperature, permeability index these three state variables.
S104:Count utilization rate of carbon monoxide corresponding to the model parameter variable of the utilization rate of carbon monoxide trend prediction
Variation tendency obtains utilization rate of carbon monoxide trend result.Table 2 is utilization rate of carbon monoxide variation tendency statistics rule
Rule.
Table 2
From table 2, model parameter variable and carbon monoxide change utilization rate are divided into increase and reduce by two major classes, according to it
Preceding selected model parameter variable, the Sample Storehouse have 32 groups of situations altogether.Wherein, it is increased to represent that model parameter variable is presented for "+"
Trend, "-" represent that the trend reduced is presented in model parameter variable.
S105:Set sampling amount threshold value and utilization rate of carbon monoxide becomes as described in the sampling amount threshold extraction
Change trend prediction result.
S106:The utilization rate of carbon monoxide trend result of extraction and utilization rate of carbon monoxide variation tendency is true
Real result is contrasted to obtain comparing result.Table 3 is the utilization rate of carbon monoxide trend result extracted and an oxidation
The comparing result of carbon utilisation rate variation tendency legitimate reading.
Table 3
S107:Utilization rate of carbon monoxide trend precision is obtained with the comparing result, specific formula is:Wherein, n (Correct_Set) represents correct comparing result, and N (Test_Set) represents to extract
Total sample number.Acc=13/18 ≈ 0.72 can be obtained according to the result of table 3.
S108:Setting accuracy threshold value.
S109:If the utilization rate of carbon monoxide trend precision meets precision threshold, the model is to have
Model is imitated, can be used for the Parameter Decision Making of blast furnace material distribution.For cloth Parameter Decision Making, when the utilization rate of carbon monoxide of prediction
During to increase, illustrate that now given cloth parameter is rational.When the utilization rate of carbon monoxide of prediction is reduces, then illustrate
Given cloth parameter is unreasonable, energy consumption can be caused to increase, and now cloth parameter should be updated, and model after renewal is joined
Number variable re-enters model, then predicts the variation tendency of utilization rate of carbon monoxide, untill utilization rate of carbon monoxide increase.
S110:If the utilization rate of carbon monoxide trend precision is unsatisfactory for precision threshold, the model is
Invalid model, then new case is found again and is used to establish valid model.
Referring to Fig. 2, Fig. 2 is that box traction substation rejecting abnormalities model parameter variable method steps flow chart is utilized in the embodiment of the present invention
Figure, including:
S201:Model parameter Variables Sequence is subjected to sequence from small to large and obtains new model parametric variable sequence Xj。
S202:Solve XjUpper quartile Q3, median Q2With lower quarter back's number Q1。
S203:Determine Exception Model parametric variable control constant k.
S204:The minimum estimate and maximum estimated value, calculation formula of computation model parametric variable be:x1=Q1-k(Q3-
Q1), x2=Q3+k(Q3-Q1), wherein, x1Represent the minimum estimate, x2Represent the maximum estimated value.
S205:Definition is less than x1Or more than x2Model parameter variable be Exception Model parametric variable.
S206:Reject the Exception Model parametric variable.
By performing embodiments of the invention, all technical characteristics in the claims in the present invention are obtained for detailed explain
State.
Prior art is different from, the embodiment provides the blast furnace material distribution Parameter Decision Making side matched based on case
Method, by the way that utilization rate of carbon monoxide to be used as to the foundation of decision-making cloth parameter, analysis utilization rate of carbon monoxide becomes with model parameter
The relevance between (blast furnace state variable) is measured, it is determined that the input variable for case matching library.Cloth parameter is then based on to obtain
To the predicted value of model parameter variable, it is finally based on case matching process and establishes the pre- variation tendency of utilization rate of carbon monoxide respectively
Forecast model, effective prediction to utilization rate of carbon monoxide variation tendency is realized, to instructing blast furnace material distribution parameter to have important guiding
Meaning.
The foregoing is only presently preferred embodiments of the present invention, be not intended to limit the invention, it is all the present invention spirit and
Within principle, any modification, equivalent substitution and improvements made etc., it should be included in the scope of the protection.
Claims (5)
1. based on the blast furnace material distribution Parameter Decision Making method of case matching, methods described is realized by hardware device, it is characterised in that:Bag
Include following steps:Choose potential model parameter variable;The model parameter variable is pre-processed to obtain using box traction substation pre-
Model parameter variable after processing;Correlation point is carried out to the pretreated model parameter variable and utilization rate of carbon monoxide
Analysis obtains the model parameter variable of utilization rate of carbon monoxide trend prediction;Count the mould of the utilization rate of carbon monoxide trend prediction
Utilization rate of carbon monoxide variation tendency corresponding to shape parameter variable obtains utilization rate of carbon monoxide trend result;Setting
Sampling amount threshold value and the utilization rate of carbon monoxide trend result as described in the sampling amount threshold extraction;Will
The utilization rate of carbon monoxide trend result of extraction is contrasted with utilization rate of carbon monoxide variation tendency legitimate reading
Obtain comparing result;Utilization rate of carbon monoxide trend precision is obtained with the comparing result;Setting accuracy threshold value;
If the utilization rate of carbon monoxide trend precision meets precision threshold, the model is valid model, Ke Yiyong
In the Parameter Decision Making of blast furnace material distribution;If the utilization rate of carbon monoxide trend precision is unsatisfactory for precision threshold, institute
It is invalid model to state model, then finds new case again and be used to establish valid model.
2. the blast furnace material distribution Parameter Decision Making method as claimed in claim 1 based on case matching, it is characterised in that:The selection
Potential model parameter variable specifically includes:Stove cold flow, furnace wall thermic load, injecting coal quantity, material speed, silicone content, permeability index,
Ore coke ratio, Rich Oxygen Amount and furnace roof rise tube top temperature;The potential model parameter variable acquisition interval is 6 hours, takes 4 numbers daily
According to.
3. the blast furnace material distribution Parameter Decision Making method as claimed in claim 2 based on case matching, it is characterised in that:The utilization
Box traction substation is pre-processed to obtain pretreated model parameter variable and specifically included to the model parameter variable:Reject obvious
Abnormal model parameter variable;The model parameter variable specific steps for rejecting obvious exception include:By model parameter variable
Sequence carries out sequence from small to large and obtains new model parametric variable sequence Xj;Solve XjUpper quartile Q3, median Q2With under
Quarter back's number Q1;Determine Exception Model parametric variable control constant k;The minimum estimate and maximum of computation model parametric variable are estimated
Evaluation, calculation formula are:x1=Q1-k(Q3-Q1), x2=Q3+k(Q3-Q1), wherein, x1Represent the minimum estimate, x2Represent
The maximum estimated value;Definition is less than x1Or more than x2Model parameter variable be Exception Model parametric variable;Reject the exception
Model parameter variable.
4. the blast furnace material distribution Parameter Decision Making method as claimed in claim 3 based on case matching, it is characterised in that:The correlation
Property the specific formula of analysis be:Wherein, I (X, Y) represents variable X and change
The association relationship between Y is measured, determines to become with the big model parameter of utilization rate of carbon monoxide correlation according to the size of mutual information
Amount, p (Xj,Ym) represent the joint probability distribution of variable X and variable Y, p (Xj) represent variable X marginal probability distribution, p (Ym) table
Show the marginal probability distribution of variable Y.
5. the blast furnace material distribution Parameter Decision Making method as claimed in claim 1 based on case matching, it is characterised in that:The utilization
The comparing result obtains the specific formula of utilization rate of carbon monoxide trend precision:
Wherein, n (Correct_Set) represents correct comparing result, and N (Test_Set) represents the total sample number extracted.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711046346.5A CN107808221A (en) | 2017-10-31 | 2017-10-31 | Blast furnace material distribution Parameter Decision Making method based on case matching |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711046346.5A CN107808221A (en) | 2017-10-31 | 2017-10-31 | Blast furnace material distribution Parameter Decision Making method based on case matching |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107808221A true CN107808221A (en) | 2018-03-16 |
Family
ID=61591063
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711046346.5A Pending CN107808221A (en) | 2017-10-31 | 2017-10-31 | Blast furnace material distribution Parameter Decision Making method based on case matching |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107808221A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109284942A (en) * | 2018-10-31 | 2019-01-29 | 山西太钢不锈钢股份有限公司 | Method for determining blast furnace crude fuel Rational Parameters range |
CN109766331A (en) * | 2018-12-06 | 2019-05-17 | 中科恒运股份有限公司 | Method for processing abnormal data and device |
CN111471819A (en) * | 2020-04-29 | 2020-07-31 | 江苏省沙钢钢铁研究院有限公司 | Method and system for regulating and controlling material distribution system of blast furnace |
CN112906196A (en) * | 2021-01-21 | 2021-06-04 | 山西太钢不锈钢股份有限公司 | Method for determining reasonable range of fuel ratio of blast furnace |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1598534A (en) * | 2004-07-28 | 2005-03-23 | 东北大学 | Soft investigating method for overflow grain index of ore grinding system based on case inference |
CN103246801A (en) * | 2013-03-04 | 2013-08-14 | 北京工业大学 | Shaft furnace fault condition forecasting method based on improved case-based reasoning |
CN103729571A (en) * | 2014-01-23 | 2014-04-16 | 中南大学 | Modeling method for utilization rate of carbon monoxide in iron-making process of blast furnace |
CN105671227A (en) * | 2016-03-21 | 2016-06-15 | 中国地质大学(武汉) | Method and device for clustering states of blast furnace |
-
2017
- 2017-10-31 CN CN201711046346.5A patent/CN107808221A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1598534A (en) * | 2004-07-28 | 2005-03-23 | 东北大学 | Soft investigating method for overflow grain index of ore grinding system based on case inference |
CN103246801A (en) * | 2013-03-04 | 2013-08-14 | 北京工业大学 | Shaft furnace fault condition forecasting method based on improved case-based reasoning |
CN103729571A (en) * | 2014-01-23 | 2014-04-16 | 中南大学 | Modeling method for utilization rate of carbon monoxide in iron-making process of blast furnace |
CN105671227A (en) * | 2016-03-21 | 2016-06-15 | 中国地质大学(武汉) | Method and device for clustering states of blast furnace |
Non-Patent Citations (5)
Title |
---|
DENGFENG XIAO ET AL.: "The chaotic characteristic of the carbon-monoxide utilization ratio in the blast furnace", 《ISA TRANSACTIONS》 * |
JIANQI AN ET AL.: "Analysis of Influencing Factors on Carbon Monoxide Utilization Rate of Blast Furnace Based on Multi-Timescale Characteristics", 《PROCEEDINGS OF THE 36TH CHINESE CONTROL CONFERENCE》 * |
XIAO DENGFENG ET AL.: "Chaotic Prediction of Carbon-Monoxide Utilization Ratio in the Blast Furnace", 《PROCEEDINGS OF THE 35TH CHINESE CONTROL CONFERENCE》 * |
安剑奇等: "基于改进支持向量机的高炉一氧化碳利用率预测方法", 《化工学报》 * |
许永华等: "基于案例与规则推理的热风炉燃烧控制方法与应用", 《计算机测量与控制》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109284942A (en) * | 2018-10-31 | 2019-01-29 | 山西太钢不锈钢股份有限公司 | Method for determining blast furnace crude fuel Rational Parameters range |
CN109284942B (en) * | 2018-10-31 | 2021-08-03 | 山西太钢不锈钢股份有限公司 | Method for determining reasonable range of blast furnace raw fuel parameters |
CN109766331A (en) * | 2018-12-06 | 2019-05-17 | 中科恒运股份有限公司 | Method for processing abnormal data and device |
CN111471819A (en) * | 2020-04-29 | 2020-07-31 | 江苏省沙钢钢铁研究院有限公司 | Method and system for regulating and controlling material distribution system of blast furnace |
CN111471819B (en) * | 2020-04-29 | 2022-03-29 | 江苏省沙钢钢铁研究院有限公司 | Method and system for regulating and controlling material distribution system of blast furnace |
CN112906196A (en) * | 2021-01-21 | 2021-06-04 | 山西太钢不锈钢股份有限公司 | Method for determining reasonable range of fuel ratio of blast furnace |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107808221A (en) | Blast furnace material distribution Parameter Decision Making method based on case matching | |
CN104630410A (en) | Real-time dynamic converter steelmaking quality prediction method based on data analysis | |
Li et al. | Dynamic time features expanding and extracting method for prediction model of sintering process quality index | |
CN102163261A (en) | Case-reasoning-based molten steel temperature prediction method | |
CN103729571B (en) | Modeling method for utilization rate of carbon monoxide in iron-making process of blast furnace | |
CN102031319A (en) | Method for forecasting silicon content in blast-furnace hot metal | |
CN104267610B (en) | High-precision blast furnace smelting process abnormal data detection and repair method | |
CN104881707B (en) | A kind of sintering energy consumption Forecasting Methodology based on integrated model | |
CN110264041A (en) | A kind of non-intrusion type load decomposition and monitoring method based on difference | |
CN103258130A (en) | Method for predicting sinter tumbler strength | |
CN107463093A (en) | A kind of blast-melted quality monitoring method based on KPLS robust reconstructed errors | |
Han et al. | Hybrid intelligent control of BOF oxygen volume and coolant addition | |
CN111915080A (en) | Raw fuel cost optimal proportioning method based on molten iron quality constraint | |
CN102864766A (en) | Liquefaction judgment method based on standard penetration and static cone penetration test correlation | |
CN105574297A (en) | Self-adaption blast-furnace melt silicon content tendency forecasting method | |
Yin et al. | Dynamic real–time abnormal energy consumption detection and energy efficiency optimization analysis considering uncertainty | |
CN103489007B (en) | SVMs Selection of kernel function method based on sample prior information and application | |
CN110097141A (en) | A kind of acquisition operational system intelligent trouble detection method | |
CN111798023B (en) | Comprehensive coke ratio prediction method in steelmaking sintering production | |
CN102968644A (en) | Method for predicting smelting finishing point of argon-oxygen refined iron alloy | |
CN103160629B (en) | Method for prediction of blast furnace heat tendency | |
CN107729676B (en) | Correlation analysis method for blast furnace operation parameters, state variables and CO utilization rate | |
CN114525372B (en) | Multi-mode fusion based blast furnace state monitoring method and device | |
Jiang et al. | A novel intelligent monitoring method for the closing time of the taphole of blast furnace based on two-stage classification | |
CN107832880B (en) | Blast furnace state variable prediction method based on material distribution parameters |
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: 20180316 |
|
RJ01 | Rejection of invention patent application after publication |