CN114395649A - Blast furnace raw fuel evaluation method based on data self-learning - Google Patents

Blast furnace raw fuel evaluation method based on data self-learning Download PDF

Info

Publication number
CN114395649A
CN114395649A CN202210131794.XA CN202210131794A CN114395649A CN 114395649 A CN114395649 A CN 114395649A CN 202210131794 A CN202210131794 A CN 202210131794A CN 114395649 A CN114395649 A CN 114395649A
Authority
CN
China
Prior art keywords
parameter
raw fuel
data
raw
blast furnace
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
Application number
CN202210131794.XA
Other languages
Chinese (zh)
Inventor
李壮年
赵新民
李宝峰
陈树文
***
武树永
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanxi Taigang Stainless Steel Co Ltd
Original Assignee
Shanxi Taigang Stainless Steel Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shanxi Taigang Stainless Steel Co Ltd filed Critical Shanxi Taigang Stainless Steel Co Ltd
Priority to CN202210131794.XA priority Critical patent/CN114395649A/en
Publication of CN114395649A publication Critical patent/CN114395649A/en
Pending legal-status Critical Current

Links

Classifications

    • C—CHEMISTRY; METALLURGY
    • C21—METALLURGY OF IRON
    • C21B—MANUFACTURE OF IRON OR STEEL
    • C21B5/00—Making pig-iron in the blast furnace
    • C21B5/001—Injecting additional fuel or reducing agents
    • C21B5/003—Injection of pulverulent coal
    • C—CHEMISTRY; METALLURGY
    • C21—METALLURGY OF IRON
    • C21B—MANUFACTURE OF IRON OR STEEL
    • C21B2300/00—Process aspects
    • C21B2300/04—Modeling of the process, e.g. for control purposes; CII

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Manufacturing & Machinery (AREA)
  • Materials Engineering (AREA)
  • Metallurgy (AREA)
  • Organic Chemistry (AREA)
  • Manufacture Of Iron (AREA)

Abstract

The invention relates to the technical field of blast furnace production, in particular to a blast furnace raw fuel evaluation method based on data self-learning, which comprises the steps of obtaining the types of raw fuels used by a blast furnace and the types of raw materials included by each raw fuel, wherein the types of the raw fuels are four types including coke, pulverized coal, sintering and pellets, the types of the raw materials are numbers for distinguishing the same raw fuel provided by different manufacturers or production procedures, setting the types and weights of parameters to be investigated by each raw fuel and a grading mode to which each parameter belongs, setting the actual value of each parameter of each raw fuel in each raw fuel type as the product R & phi of a basic value and a parameter grading coefficient, and setting the actual value R & phi of each raw fuel in each raw fuelSynthesis ofIs the cumulative sum of the products of the base scores of each parameter and the parameter scoring coefficients for all the feedstock models of the raw fuel.

Description

Blast furnace raw fuel evaluation method based on data self-learning
Technical Field
The invention relates to the technical field of blast furnace production, in particular to the field of blast furnace raw fuel quality management.
Background
In the production process of the blast furnace, the quality and the stability of the raw fuel directly influence the stable and smooth operation of the blast furnace and the technical indexes of the blast furnace production. In the blast furnace ironmaking industry, the statement of 'seven-component raw material and three-component operation' is widely accepted. Due to the limitations of production conditions and resource environments and the aim of reducing cost, the quality of the raw fuel of the blast furnace often changes greatly, however, different blast furnaces have different corresponding modes for a blast furnace raw fuel quality evaluation system, most blast furnaces mainly conduct extensive and qualitative evaluation, and a scientific and systematic evaluation system is not formed, so that the condition of the blast furnace cannot be greatly optimized.
Disclosure of Invention
The purpose of the invention is: strictly relating to the quality of the raw fuel, and strengthening the management of the raw fuel by establishing a data self-learning raw fuel quantitative evaluation model taking statistics as a mechanism, thereby providing guarantee for stable and smooth running of furnace conditions and providing support for improving the technical indexes of blast furnace production.
The technical scheme adopted by the invention is as follows: a blast furnace raw fuel evaluation method based on data self-learning is characterized by obtaining the types of raw fuels used by a blast furnace and the types of raw materials contained by each type of raw fuel, wherein the types of the raw fuels are four types including coke, coal powder, sintering and pellets, the types of the raw materials are numbers for distinguishing the same type of raw fuel provided by different manufacturers or production processes, the types and the weights of parameters to be investigated by each type of raw fuel and a scoring mode to which each parameter belongs are set, the lower the index value X of the parameter is, the better the first scoring mode is, and the evaluation scoring formula is as follows: base score
Figure BDA0003502945120000011
L1The index value X of the parameter is less than or equal to L1When R is 100, U1The second scoring mode is that the index value X of the parameter is higher and better, and the evaluation scoring formula is as follows: base score
Figure BDA0003502945120000012
L2The index value X of the parameter is less than or equal to L as the second lower limit value2When R is 0, U2The index value X of the parameter is greater than or equal to U2When R is 100, the index value X of the parameter in the third scoring mode is inThe evaluation scoring formula is the best between the upper limit and the lower limit: base score
Figure BDA0003502945120000013
Or base score
Figure BDA0003502945120000014
L3Is a third lower limit value, L4Is a fourth lower limit value, U3Is a third upper limit value, U4The index value X of the parameter is greater than or equal to L4While being less than or equal to U4When R is 100, the index value X of the parameter is equal to or greater than U3Or the index value X of the parameter is less than or equal to L3When R is 0; each parameter scoring coefficient phi of each raw material model of each raw fuel is lambda omega eta, wherein the weight of the raw fuel corresponding to the lambda raw material model, the sum of the weights of all raw fuel types is 100, omega is the percentage of the weight of the raw material model to the total weight of the corresponding raw fuel, eta is the parameter weight, namely the weight of the parameter in all the parameters of each raw fuel, each parameter actual score of each raw fuel of each raw material model is the product R phi of the basic score and the parameter scoring coefficient, and the actual score R of each raw fuelSynthesis ofThe cumulative sum of the products of the base score and the parameter scoring coefficient for each parameter of all feedstock types of the raw fuel, i.e.
Figure BDA0003502945120000015
n is the sum of all parameters of all raw material models of the raw fuel, i is the corresponding serial number of each parameter, i is more than or equal to 1 and less than or equal to n, RiFor the base score, phi, of each parameteriThe total actual score of all raw fuels is the cumulative sum of the actual scores of each raw fuel.
First lower limit value L in first scoring mode1And a first upper limit value U1The acquisition method is that1＝P5％+σ，U1＝P95％+ sigma; second lower limit value L of second scoring mode2And a second upper limit value U2The acquisition method is that2＝P5％-σ，U2＝P95％- σ; third lower limit value L of third scoring mode3And a fourth lower limit value L4And the third upper limit value U3And the fourth upper limit value U4The obtaining method is L3＝P5％-σ，U3＝P95％+σ，L4＝P25％，U4＝P75％(ii) a Where σ is the standard deviation of the parameter, P5％The minimum 5% of data is deleted after 5% quantile of the historical data of the parameter, namely the historical data of the parameter is sorted from small to large, the minimum value of the rest data is taken, and P is95％The minimum 95% of the data is deleted after the 95% quantile of the historical data of the parameter, namely the historical data of the parameter is sorted from small to large, and the minimum value P of the rest data is taken25％The minimum 25% of the data is deleted after the 25% quantile of the historical data of the parameter, namely the historical data of the parameter is sorted from small to large, and the minimum value of the rest data is taken, P75％The minimum 75% of the data is deleted after 75% quantiles of the historical data of the parameters, namely the historical data of the parameters are sorted from small to large, and the minimum value of the rest data is taken.
The historical data of the parameters refers to the historical data of the parameters in the last 1-2 years, and is data changing along with time. When the limit parameters (the upper limit value and the lower limit value) deviate from the actual production condition (experience judgment), the limit parameters are multiplied by a correction coefficient alpha to correct the limit parameters, wherein the alpha is generally 0.7-1.3, and mainly aims at the application of different blast furnace historical data, for example, when one blast furnace starts to operate, the data of other blast furnaces can be adopted and then multiplied by the correction coefficient alpha to correct the data.
The invention has the beneficial effects that: the method can obtain the single score of the raw fuel index by adopting a data self-learning method, and then obtain the comprehensive score of the raw fuel by a weighting method, namely the index for judging the total quality level of the raw fuel, thereby providing a reliable quantitative basis for the quality analysis of the raw fuel of the blast furnace. With a scientific raw fuel evaluation system, a blast furnace operator can adjust and deal with the blast furnace according to an evaluation result and the condition of the blast furnace so as to realize early warning and prior adjustment and provide support for stable and smooth running and reinforced smelting of the blast furnace.
Detailed Description
The method is combined with the production practice of the blast furnace, a statistical analysis principle is technically applied, a raw fuel evaluation model with data self-learning is established, the quality level of the raw fuel is accurately reflected by a quantitative analysis mode, the scientific evaluation on the quality of the raw fuel of the blast furnace is realized, and the method has obvious guiding significance and practical value on the operation of the blast furnace. The evaluation method is successfully applied to the Tai steel 6# blast furnace, achieves good effect, and has the following main benefits:
the daily output of molten iron of the Tai steel 6# blast furnace is improved from 9577t/d to 10089t/d, and according to the normal production and operation of steel enterprises, the economic benefit brought to the enterprises by each ton of molten iron is at least 200 yuan/tFe; the fuel ratio is reduced from 513.8kg/tFe to 506.0kg/tFe (wherein, the coal ratio is reduced from 166.5 to 151.5kg/tFe, the coke ratio is increased from 347.4 to 354.5kg/tFe), the coke price is 1900 yuan/t, the coal dust price is 1070 yuan/t, and the contribution rate of the patent is 30%.
The economic benefits generated each year after the yield is improved: (10089-9577) × 360 × 200/10000 × 30% ═ 1105.9 ten thousand yuan.
The cost of the iron fuel per ton is reduced: (166.5-151.5) × 1070+ (347.4-354.5) × 1900 ═ 2.56 m/tFe.
The reduction of fuel cost generates economic benefits each year: 10089 × 2.56 × 360/10000 × 30% ═ 278.9 ten thousand yuan. In conclusion, the economic benefit generated by the invention patent is 1384.8 ten thousand yuan each year.
With a scientific raw fuel evaluation system, a blast furnace operator can adjust and deal with the blast furnace according to an evaluation result and the condition of the blast furnace so as to realize early warning and prior adjustment and provide support for stable and smooth running of the blast furnace.
Description of the drawings: the time range of the data after the patent is adopted is 1 month to 7 months 2021 in 2020, and the time range of the comparative data is 6# blast furnace blow-in to 6 months 2020.
Firstly, determining lambda, omega and eta of each raw fuel according to the production condition of a blast furnace, thereby obtaining the weight coefficient phi of each raw fuel; counting percentile parameters P of each index in the last 2 years5％、P25％、P75％、P95％And a standard deviation σ. Raw fuelThe weighting parameters and statistical parameters are shown in table 1.
TABLE 1 raw Fuel weight parameters and statistical parameters
Figure BDA0003502945120000041
Secondly, obtaining a limit parameter L of each index according to the raw fuel parameter mode and the statistical result1、L2、U1、U2And substituting the index value into a formula to obtain the single-term score R of each raw fuel index, wherein the result is shown in Table 2.
TABLE 2 raw Fuel index Limit parameters and evaluation results
Figure BDA0003502945120000051
And accumulating the single scores of the 31 raw fuel indexes according to the given weight to obtain the comprehensive score of the raw fuel, wherein the final score result is 72.84.
By the method, the systematic and scientific evaluation of the quality of the raw fuel can be realized.

Claims (3)

1. A blast furnace raw fuel evaluation method based on data self-learning is characterized by comprising the following steps: the method comprises the steps of obtaining the type of raw fuel used by a blast furnace and the type of raw material included by each type of raw fuel, wherein the type of raw fuel is four types of coke, coal powder, sintering and pellets, the type of raw material is a number for distinguishing the same type of raw fuel provided by different manufacturers or production processes, setting the type and weight of parameters to be investigated of each type of raw fuel and a scoring mode to which each parameter belongs, and setting a first scoring mode to ensure that the lower the index value X of the parameter is, the better the first scoring mode is, and the evaluation scoring formula is as follows: base score
Figure FDA0003502945110000011
L1The index value X of the parameter is less than or equal to L1When R is 100, U1Is a first upper limit valueThe second scoring mode is that the higher the index value X of the parameter is, the better the evaluation scoring formula is: base score
Figure FDA0003502945110000012
L2The index value X of the parameter is less than or equal to L as the second lower limit value2When R is 0, U2The index value X of the parameter is greater than or equal to U2When R is 100, the index value X of the parameter in the third scoring mode is in the best upper and lower limit interval, and the evaluation scoring formula is as follows: base score
Figure FDA0003502945110000013
Or base score
Figure FDA0003502945110000014
L3Is a third lower limit value, L4Is a fourth lower limit value, U3Is a third upper limit value, U4The index value X of the parameter is greater than or equal to L4While being less than or equal to U4When R is 100, the index value X of the parameter is equal to or greater than U3Or the index value X of the parameter is less than or equal to L3When R is 0; each parameter scoring coefficient phi of each raw material model of each raw fuel is lambda omega eta, wherein the weight of the raw fuel corresponding to the lambda raw material model, the sum of the weights of all raw fuel types is 100, omega is the percentage of the weight of the raw material model to the total weight of the corresponding raw fuel, eta is the parameter weight, namely the weight of the parameter in all the parameters of each raw fuel, each parameter actual score of each raw fuel of each raw material model is the product R phi of the basic score and the parameter scoring coefficient, and the actual score R of each raw fuelSynthesis ofThe cumulative sum of the products of the base score and the parameter scoring coefficient for each parameter of all feedstock types of the raw fuel, i.e.
Figure FDA0003502945110000015
n is the sum of all parameters of all raw material models of the raw fuel, i is the corresponding serial number of each parameter, i is more than or equal to 1 and less than or equal ton，RiFor the base score, phi, of each parameteriThe total actual score of all raw fuels is the cumulative sum of the actual scores of each raw fuel.
2. The method for evaluating the raw fuel of the blast furnace based on the data self-learning as claimed in claim 1, wherein: first lower limit value L in first scoring mode1And a first upper limit value U1The acquisition method is that1＝P5％+σ，U1＝P95％+ sigma; second lower limit value L of second scoring mode2And a second upper limit value U2The acquisition method is that2＝P5％-σ，U2＝P95％- σ; third lower limit value L of third scoring mode3And a fourth lower limit value L4And the third upper limit value U3And the fourth upper limit value U4The obtaining method is L3＝P5％-σ，U3＝P95％+σ，L4＝P25％，U4＝P75％(ii) a Where σ is the standard deviation of the parameter, P5％The minimum 5% of data is deleted after 5% quantile of the historical data of the parameter, namely the historical data of the parameter is sorted from small to large, the minimum value of the rest data is taken, and P is95％The minimum 95% of the data is deleted after the 95% quantile of the historical data of the parameter, namely the historical data of the parameter is sorted from small to large, and the minimum value P of the rest data is taken25％The minimum 25% of the data is deleted after the 25% quantile of the historical data of the parameter, namely the historical data of the parameter is sorted from small to large, and the minimum value of the rest data is taken, P75％The minimum 75% of the data is deleted after 75% quantiles of the historical data of the parameters, namely the historical data of the parameters are sorted from small to large, and the minimum value of the rest data is taken.
3. The method for evaluating the raw fuel of the blast furnace based on the data self-learning as claimed in claim 2, wherein: the historical data of the parameters refers to the historical data of the parameters in the last 1-2 years, and is data changing along with time.
CN202210131794.XA 2022-02-14 2022-02-14 Blast furnace raw fuel evaluation method based on data self-learning Pending CN114395649A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210131794.XA CN114395649A (en) 2022-02-14 2022-02-14 Blast furnace raw fuel evaluation method based on data self-learning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210131794.XA CN114395649A (en) 2022-02-14 2022-02-14 Blast furnace raw fuel evaluation method based on data self-learning

Publications (1)

Publication Number Publication Date
CN114395649A true CN114395649A (en) 2022-04-26

Family

ID=81233873

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210131794.XA Pending CN114395649A (en) 2022-02-14 2022-02-14 Blast furnace raw fuel evaluation method based on data self-learning

Country Status (1)

Country Link
CN (1) CN114395649A (en)

Citations (5)

* Cited by examiner, † Cited by third party
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
CN111445142A (en) * 2020-03-26 2020-07-24 华润电力技术研究院有限公司 Fuel blending combustion evaluation method, system and device
CN112884294A (en) * 2021-01-27 2021-06-01 广东韶钢松山股份有限公司 Fine powder resource evaluation method, device, equipment and storage medium
CN113077132A (en) * 2021-03-22 2021-07-06 山西太钢不锈钢股份有限公司 Method for evaluating cost performance of pulverized coal injection
CN113420426A (en) * 2021-06-03 2021-09-21 北京首钢股份有限公司 Method, device, medium and computer equipment for determining forward running condition of blast furnace

Patent Citations (5)

* Cited by examiner, † Cited by third party
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
CN111445142A (en) * 2020-03-26 2020-07-24 华润电力技术研究院有限公司 Fuel blending combustion evaluation method, system and device
CN112884294A (en) * 2021-01-27 2021-06-01 广东韶钢松山股份有限公司 Fine powder resource evaluation method, device, equipment and storage medium
CN113077132A (en) * 2021-03-22 2021-07-06 山西太钢不锈钢股份有限公司 Method for evaluating cost performance of pulverized coal injection
CN113420426A (en) * 2021-06-03 2021-09-21 北京首钢股份有限公司 Method, device, medium and computer equipment for determining forward running condition of blast furnace

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
郭泽琴: "太钢5号高炉低燃料比冶炼实践", 《山西冶金》 *

Similar Documents

Publication Publication Date Title
CN109583118B (en) Sintering ratio calculation and sinter cost optimization method
CN104164558B (en) A kind of sintering circuit sinter quality control method
CN110484714B (en) Method for improving quality of sinter by improving proportion of sinter materials
CN114091871A (en) Blast furnace iron-making ore blending method and system
WO2015045766A1 (en) Control device and control method for converter furnace blowing equipment
CN113919206A (en) Narrow composition control method for steelmaking alloying
CN104593540A (en) Method for evaluating energy efficiency in converter steelmaking process
CN105734276A (en) Comprehensive assessment method and comprehensive assessment system for economical efficiency of iron ores for blast furnace process
CN106228011B (en) The evaluation method of coke quality steady
CN110669963A (en) Aluminum alloy casting batching system
CN112695153A (en) Method for optimizing steelmaking alloy feeding amount and reducing cost
CN109252005A (en) The method for establishing Blending Optimized model
CN107818408A (en) A kind of silicon steel dynamic mass design method based on mark
CN106521059A (en) Method for controlling blast furnace gas flow distribution by using phased array radar to measure ore coke ratio of blast furnace material surface
CN116469481B (en) LF refined molten steel composition forecasting method based on XGBoost algorithm
CN110246547B (en) Ore blending optimization method in sintering process
CN114395649A (en) Blast furnace raw fuel evaluation method based on data self-learning
CN112680566B (en) Refining furnace decarburization end point detection method and system
CN113626993A (en) Method for evaluating smelting value of iron ore
CN116306232A (en) Blast furnace energy consumption and carbon emission analysis method and system based on industrial big data
CN103901179B (en) A kind of method evaluating coke Blending Efficiency of Blending and coke quality steady
CN112036628B (en) Method for establishing model for representing coal blending cost variation
CN215066336U (en) Sintering ore FeO content on-line detection device based on pipe belt machine
CN116401491A (en) Method for quantifying maximum blast furnace air quantity according to fluctuation level of furnace condition
CN110766305A (en) Raw fuel quality tracking and early warning method

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
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20220426