CN109359723A - Based on the converter terminal manganese content prediction technique for improving regularization extreme learning machine - Google Patents
Based on the converter terminal manganese content prediction technique for improving regularization extreme learning machine Download PDFInfo
- Publication number
- CN109359723A CN109359723A CN201811382479.4A CN201811382479A CN109359723A CN 109359723 A CN109359723 A CN 109359723A CN 201811382479 A CN201811382479 A CN 201811382479A CN 109359723 A CN109359723 A CN 109359723A
- Authority
- CN
- China
- Prior art keywords
- manganese content
- data
- prediction
- learning machine
- extreme learning
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- C—CHEMISTRY; METALLURGY
- C21—METALLURGY OF IRON
- C21C—PROCESSING OF PIG-IRON, e.g. REFINING, MANUFACTURE OF WROUGHT-IRON OR STEEL; TREATMENT IN MOLTEN STATE OF FERROUS ALLOYS
- C21C5/00—Manufacture of carbon-steel, e.g. plain mild steel, medium carbon steel or cast steel or stainless steel
- C21C5/28—Manufacture of steel in the converter
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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"
Abstract
The invention discloses a kind of based on the converter terminal manganese content prediction technique for improving regularization extreme learning machine, first, acquire data involved in pneumatic steelmaking actual production, find the influence factor for influencing BOF Steelmaking Endpoint manganese content, determine the input variable of model, establish the BOF Steelmaking Endpoint manganese content prediction model based on regularization extreme learning machine (RELM), during the prediction of terminal manganese content, it is optimized using input layer weight and hidden layer deviation of the Modified particle swarm optimization algorithm (IPSO) to RELM model, it establishes based on the BOF Steelmaking Endpoint manganese content prediction model for improving particle swarm algorithm Optimal Regularization extreme learning machine (IPSO-RELM).It is tested using the practical data of smelting in converter scene to the manganese content prediction technique, the results showed that the precision of prediction and arithmetic speed of this method, which have, to be obviously improved, and then prediction promptly and accurately can be carried out to BOF Steelmaking Endpoint manganese content.
Description
Technical field
The invention belongs to field of steel metallurgy, and in particular to a kind of based on the converter terminal for improving regularization extreme learning machine
Manganese content prediction technique.
Background technique
Pneumatic steelmaking is current steel-making side most important in the world as link highly important in Steel Production Flow Chart
Method.The purpose of pneumatic steelmaking is to provide chemical component and temperature qualified first steel-making water for next procedure, thus it is guaranteed that blowing
Steady progress and Accurate Prediction and the ingredient and temperature that control smelting endpoint molten steel, be pneumatic steelmaking vital task it
One.In converter steelmaking process, if being able to achieve the accurate and quick prediction to endpoint molten steel manganese content, operation can be improved
The accuracy that personnel judge tapping, and the efficiency of steel alloying operation out, so that production cost is reduced, raising molten steel matter
Amount.In order to realize that steel smelting procedure is precisely controlled, how timely and accurately to forecast that the content of BOF Steelmaking Endpoint molten steel manganese is
Major issue urgently to be resolved.
In order to realize the target of the high hit rate of BOF Steelmaking Endpoint Control, bessemerize from pervious artificial experience operate to
Automatic Steelmaking direction develop.Currently, the prediction technique of BOF Steelmaking Endpoint manganese content mainly includes statistical method and non-
Statistical method, statistical method include: linear regression, nonlinear regression etc., and non-statistical method includes: expert system, BP
Neural network etc..Since the converter steelmaking process molten bath chemical reaction mechanism of action is sufficiently complex, the factor of molten steel manganese content is influenced
Very much, and between these factors it influences each other, has stronger non-linear relation with terminal manganese content.Based on statistical method
The adaptability and generalization ability of the BOF Steelmaking Endpoint manganese content prediction model of foundation are weaker, and certain statistical methods are in base
On the basis of sublance sampling analysis molten steel manganese content, is established using thermodynamics and material balance from sublance and be sampled to blowing eventually
The manganese content prediction model of point, higher cost, and it is only applicable to the converter equipped with sublance；And turned based on non-statistical method
Furnace make steel terminal manganese content prediction model have stronger adaptability and generalization ability, due to influence steel-making terminal manganese content because
Element is numerous, and has stronger non-linear relation between each influence factor and terminal manganese content, and artificial neural network is stronger non-
Linear approximation ability can be very good to solve the problems, such as this, BP neural network be current widely used neural network it
One.Equally, this method also has predicts applied to BOF Steelmaking Endpoint manganese content.But the model based on this method foundation is in training
Process needs to consume a large amount of time, easily fall into local optimum and training process needs to be arranged a large amount of network training ginseng
Number, and precision of prediction is low, it is difficult to BOF Steelmaking Endpoint manganese content is accurately predicted rapidly, in time, is unfavorable for steel
The high-efficiency reform of enterprise's high-quality steel.Although RELM model is on the basis of ELM model by introducing structural risk minimization
Theory improves generalization ability, but when solving the problems, such as gradient decline, due to concealed nodes parameter (the input layer power of RELM model
Value and hidden layer deviation) it is randomly generated, do not ensure that the RELM model trained is optimal, it is therefore desirable to more implicit
The problems such as node layer can be only achieved ideal precision, slow so as to cause convergence rate.Further, since exporting weight matrix in RELM
Acquisition is calculated by input weight matrix and hidden layer deviation, there may be invalid hidden layer node, this will affect RELM
Precision of prediction, efficiency and the stability of model, are finally not achieved ideal Generalization Capability.Therefore, it is adaptable to develop one kind,
The BOF Steelmaking Endpoint manganese content prediction technique that arithmetic speed is fast and predictablity rate is high, for improving BOF Steelmaking Endpoint ingredient
Controlled level is of great significance.
Summary of the invention
To solve the above problems, improving the precision of prediction of BOF Steelmaking Endpoint manganese content, the present invention passes through analysis population
The principle of optimization algorithms SO proposes a kind of based on Modified particle swarm optimization algorithm optimization regularization extreme learning machine (IPSO-
RELM converter smelting endpoint manganese content prediction model), and carried out training with certain steel mill's converter actual production data and verified,
Significantly improve the precision of prediction and arithmetic speed of BOF Steelmaking Endpoint manganese content.
The present invention relates to a kind of based on the BOF Steelmaking Endpoint manganese for improving particle swarm algorithm Optimal Regularization extreme learning machine
Content prediction method, this method is mainly by during smelting medium and high carbon steel, passing through structure to domestic 80 tons of converters of certain Special Steel Works
Molten steel manganese content prediction model (at the end of master blows) is built the converter smelting later period, is Content of Medium-high Carbon Steels in Converter Steelmaking Process endpoint molten steel manganese content
Offer guidance is be provided.
PSO determines the global optimum of particle by iterating to calculate.However, particle swarm algorithm is in iteration searching process
There are problems that being easily trapped into local extremum.Therefore, in order to solve this problem, convergence speed of the algorithm is improved, the present invention adopts
With a kind of modified particle swarm optiziation, improvement part is as follows:
(1) improve the deficiency of standard particle group's algorithm using a kind of nonlinear weight weighing method herein, this algorithm can be retouched
It states are as follows:
When current iteration number k is smaller, ω is close to ωmaxIt ensure that the ability of searching optimum of algorithm；With iteration time
The increase of number k, ω ensure that the local search ability of algorithm with decreases in non-linear, so that algorithm be enable neatly to adjust the overall situation
Search and local search ability.
(2) " variation " factor is added to particle swarm algorithm in the variation thought for using for reference genetic algorithm, that is, allows particle to exist certain
Aberration rate, after particle updates every time, with certain probability initialize particle, make its energy in the case where falling into local extremum
It enough jumps out and continues searching, to expand its search range.
Firstly, being pre-processed to sample data collected, metallurgical basic principle is combined with correlation analysis, really
The required data of this fixed model；Secondly, the data sample being collected into is divided into two parts, wherein 4/5ths number is randomly selected
It is used to establish model according to sample, the data sample of residue 1/5th is used to verify the accuracy of model；Finally, according to model
Prediction result is compared with the numerical value of actual measurement and draws a conclusion, and the time needed for counting the model calculation.Acceptable
Training time in, generalization ability, stability and the accuracy of prediction of prediction model are obviously improved.
Specific step is as follows by the present invention:
Step 1, the pretreatment of data pre-processes the abnormal data in data sample, is finally determined as this mould
The sample data of type, since the factor for influencing BOF Steelmaking Endpoint manganese content is numerous, and the quantity between each influence factor data
Grade has biggish difference, will affect to the precision of prediction of model, therefore, needs before model training to sample
Data are normalized, and can eliminate the problem for causing precision of prediction not high due to data dimension difference and guarantor in this way
The raw information of each variable is held, data normalization range is [- 1,1].
Step 2, construction is based on regularization extreme learning machine BOF Steelmaking Endpoint manganese content prediction model, randomly selects step
(1) training data of 4/5ths data samples as model in, and model is established with this.
Step 3, remaining 1/5th data sample of step 2 is used to verify the accuracy of model, is entered into instruction
In the regularization extreme learning machine perfected, prediction result, and the precision of prediction by comprehensively considering model and prediction are finally obtained
Mean square error select suitable regularization coefficient, node in hidden layer and the activation primitive of hidden layer.
Step 4, the data of actual production process are acquired in real time, and the obtaining in real time to data by industrial computer
It takes and then Accurate Prediction can be carried out to BOF Steelmaking Endpoint manganese content.
BP neural network (Back Propagation, abbreviation BP)
Extreme learning machine (Extreme learning machine, abbreviation ELM)
Regularization extreme learning machine (Regularized Extreme Learning Machine, abbreviation RELM)
Particle swarm optimization algorithm (Particle Swarm Optimization, abbreviation PSO)
Modified particle swarm optimization algorithm (Improved Particle Swarm Optimization, abbreviation IPSO).
By the above content it is found that provided by the present application is a kind of based on the converter terminal for improving regularization extreme learning machine
Manganese content prediction technique, firstly, data involved in acquisition pneumatic steelmaking actual production, finding, which influences BOF Steelmaking Endpoint manganese, contains
Then the influence factor of amount pre-process to selected sample data and determines extreme learning machine according to these data
Training dataset is input in regularization extreme learning machine and completes to model later by input node number, output node number
Training complete the prediction to BOF Steelmaking Endpoint manganese content finally, input remaining sample data.The application passes through improvement
Particle swarm algorithm Optimal Regularization extreme learning machine model predicts BOF Steelmaking Endpoint manganese content, using improvement population
Algorithm carries out optimizing to the biasing of regularization extreme learning machine input layer weight and hidden layer, and the optimal value that search obtains is put into
Steel-making terminal manganese content is predicted in regularization extreme learning machine.The precision of prediction and arithmetic speed of this model have bright
Aobvious promotion, and then prediction promptly and accurately can be carried out to BOF Steelmaking Endpoint manganese content.
Detailed description of the invention
Fig. 1 is the structural diagram of the present invention；
Fig. 2 is model construction of the present invention and calculation flow chart.
Specific embodiment
In order to make those skilled in the art that the present invention may be better understood, with reference to the accompanying drawings and examples to this hair
Bright technical solution further illustrates.
- 2 pairs of embodiment of the present invention are described further referring to Fig.1.
In the specific implementation process, industrial control computer calculation process of the present invention including the following steps:
(1) the medium and high carbon steel historical data of certain domestic Special Steel Works converter smelting is acquired first, and determines that influence turns
Furnace makes steel terminal manganese content influence factor collection.The present invention is as example, by data application Pearson (Pierre collected
It is inferior) related coefficient progress correlation analysis, determine the input variable number of model.
(2) pretreatment that data are carried out to the factor for influencing BOF Steelmaking Endpoint manganese content, rejects abnormal data,
Determine final sample data set.
(3) data sample set derived above is normalized, data normalization range is [- 1,1].
(4) regularization extreme learning machine is constructed, the data according to handled by (3) randomly select wherein from data sample
4/5ths data sample carrys out training pattern, and the data sample of residue 1/5th verifies the accuracy of model.By comprehensive
The mean square error for considering the prediction of regularization extreme learning machine and precision of prediction are closed reasonable hidden layer node number is arranged, implies
Layer activation primitive and regularization coefficient.
(5) input layer weight and hidden layer deviation optimizing --- initialization population.Initial population is produced at random by RELM
Raw input layer weight matrix and hidden layer deviation matrix composition.Population scale P is set as n, if node in hidden layer is h, input
Layer neuron number is m, dimensionality of particle L=h (m+1).Choose Studying factors c1And c1, position and speed is limited in [- 1,1].
(6) fitness of each particle is calculated.The particle in initialization population is trained using RELM algorithm, is calculated
The training set mean square error (MSE) of each particle out, as the fitness value of particle swarm optimization algorithm.Wherein fitness is public
Formula:
In formula, fiFor converter smelting endpoint Mn content measured value, yiFor converter smelting endpoint Mn content prediction value.
(7) individual extreme value and group's extreme value are found.After each iteration, by of calculated fitness value MSE and particle
Body extreme value, group's extreme value are compared, if the fitness value is smaller, as individual extreme value and group's extreme value.It repeats
The above process terminates to iteration, can obtain one group of preferably input layer weight matrix and hidden layer deviation matrix.
(8) model training and test.Input layer weight matrix and hidden layer deviation square will be obtained by above-mentioned searching process
Battle array is brought into RELM model, is substituted into data and is trained and tests.
(9) after bessemerizing beginning, Database Systems collect and record the heat information in real time, and obtained data are carried out
Normalized, data processing range are [- 1,1], these numbers that industrial control computer is provided according to process database system
According to, be input to the established BOF Steelmaking Endpoint manganese content prediction model based on regularization extreme learning machine, to scene blow
Heat carry out manganese content prediction.
The present invention is to implement carrier with certain steel mill 80t converter producing medium and high carbon steel, according to metallurgical basic principle, in conjunction with correlation
Property analysis is determining is contained with molten iron charge weight, steel scrap charge weight, molten iron temperature, molten iron phosphorus content, molten iron manganese content, molten iron sulphur
Amount, oxygen consumption, lime adding amount, light dolomite additional amount and slagging agent additional amount etc. are input variable, terminal manganese content
As output variable, the converter smelting endpoint molten steel manganese content prediction model based on regularization extreme learning machine RELM is established.
Table 1 is the test result after the present invention is implemented, and predicts that hit rate of the error in ± 0.025% range is 94%, mean square error
It is 2.18 × 10-8.Method provided by the invention can be to can relatively accurately predict converter smelting endpoint manganese content, and then can be real
Border production, which is precisely controlled, provides a kind of important references.
Test result after 1 present invention implementation of table
The preferred embodiment of the present invention above described embodiment only expresses, the description thereof is more specific and detailed, but simultaneously
Limitations on the scope of the patent of the present invention therefore cannot be interpreted as.It should be pointed out that for those of ordinary skill in the art
For, without departing from the inventive concept of the premise, several deformations can also be made, improves and substitutes, these belong to this hair
Bright protection scope.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
Claims (3)
1. a kind of based on the converter terminal manganese content prediction technique for improving regularization extreme learning machine, characterized in that it includes such as
Lower step:
Step 1, the selection of the input variable of regularization extreme learning machine, according to the influence factor of BOF Steelmaking Endpoint manganese content and
Terminal manganese content carries out correlation analysis in conjunction with metallurgical basic principle and with Pearson came (Pearson) related coefficient, finds shadow
Ring the influence factor of BOF Steelmaking Endpoint manganese content；
Step 2, the influence factor data of BOF Steelmaking Endpoint manganese content are acquired, and these data are pre-processed,
Ensure the real effectiveness of data, the final sample data for determining this method and using；
Step 3, regularization extreme learning machine input data is normalized, select the range of data normalization for [- 1,
1]；
Step 4, initialization population, initial population is by the RELM input layer weight matrix being randomly generated and hidden layer deviation square
Battle array composition, population scale P is set as n, if node in hidden layer is h, input layer number is m, dimensionality of particle L=h (m+1),
Choose Studying factors c1And c1, position and speed is limited in [- 1,1]；
Step 5, the fitness for calculating each particle is trained the particle in initialization population using RELM algorithm, calculates
The training set mean square error (MSE) of each particle out, as the fitness value of particle swarm optimization algorithm, wherein fitness is public
Formula:
In formula, fiFor converter smelting endpoint Mn content measured value, yiFor converter smelting endpoint Mn content prediction value；
Step 6, individual extreme value and group's extreme value are found, after each iteration, by the individual of calculated fitness value MSE and particle
Extreme value, group's extreme value are compared, if the fitness value is smaller, as individual extreme value and group's extreme value, in repetition
State process terminates to iteration, obtains one group of preferably input layer weight matrix and hidden layer deviation matrix；
Step 7, model training and test, will obtain input layer weight matrix and hidden layer deviation matrix by above-mentioned searching process
It is brought into RELM model, to historical data collected, chooses 4/5ths data therein to train the regularization limit
Habit machine chooses remaining 1/5th data to verify the accuracy of this method, by comprehensively considering regularization extreme learning machine
Training degree of fitting and the precision of prediction of manganese content reasonable hidden layer node number and hidden layer activation primitive and just be set
Change coefficient, then to guarantee the optimization of network structure；
Step 8, after bessemerizing beginning, Database Systems collect and record the heat information in real time, and obtained data are carried out
Normalized, data processing range are [- 1,1], these numbers that industrial control computer is provided according to process database system
According to, be input to the established BOF Steelmaking Endpoint manganese content prediction model based on extreme learning machine, to scene blowing heat
Carry out the prediction of manganese content.
2. special as described in claim 1 based on the converter terminal manganese content prediction technique for improving regularization extreme learning machine
Sign is, in the step 2, the data preprocessing method is carried out in advance with data smoothing technique to abnormal data by rejecting
Processing.
3. special as described in claim 1 based on the converter terminal manganese content prediction technique for improving regularization extreme learning machine
Sign is that this method is realized pre- in real time to the progress of BOF Steelmaking Endpoint manganese content using industrial control computer and process database
Report, wherein industrial control computer for predicting BOF Steelmaking Endpoint manganese content in real time；Process database and Industry Control calculate
Machine is connected, and for acquiring in real time, recording convertor steelmaking process data, provides data branch for the operation of industrial control computer
Support.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811382479.4A CN109359723A (en) | 2018-11-20 | 2018-11-20 | Based on the converter terminal manganese content prediction technique for improving regularization extreme learning machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811382479.4A CN109359723A (en) | 2018-11-20 | 2018-11-20 | Based on the converter terminal manganese content prediction technique for improving regularization extreme learning machine |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109359723A true CN109359723A (en) | 2019-02-19 |
Family
ID=65332316
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811382479.4A Pending CN109359723A (en) | 2018-11-20 | 2018-11-20 | Based on the converter terminal manganese content prediction technique for improving regularization extreme learning machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109359723A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110942208A (en) * | 2019-12-10 | 2020-03-31 | 萍乡市恒升特种材料有限公司 | Method for determining optimal production conditions of silicon carbide foam ceramic |
| CN111933223A (en) * | 2020-07-03 | 2020-11-13 | 大冶特殊钢有限公司 | Automatic batching method in steel-making alloying process |
| CN112541625A (en) * | 2020-12-07 | 2021-03-23 | 东北大学 | Self-adaptive multi-working-condition steel secondary energy generation amount dynamic prediction method |
| CN112992285A (en) * | 2020-12-31 | 2021-06-18 | 无锡东研信科科技研发有限公司 | IPSO-HKELM-based blast furnace molten iron silicon content prediction method |
| CN112992129A (en) * | 2021-03-08 | 2021-06-18 | 中国科学技术大学 | Attention-keeping mechanism monotonicity keeping method in voice recognition task |
| CN113033705A (en) * | 2021-04-22 | 2021-06-25 | 江西理工大学 | Intelligent judgment and verification method for copper converter blowing slagging period end point based on pattern recognition |
| CN113112011A (en) * | 2020-01-13 | 2021-07-13 | 中移物联网有限公司 | Data prediction method and device |
| CN113192568A (en) * | 2021-03-15 | 2021-07-30 | 山东钢铁股份有限公司 | Refining furnace desulfurization terminal point forecasting method and system |
| CN114638555A (en) * | 2022-05-18 | 2022-06-17 | 国网江西综合能源服务有限公司 | Power consumption behavior detection method and system based on multilayer regularization extreme learning machine |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104200268A (en) * | 2014-09-03 | 2014-12-10 | 辽宁大学 | PSO (Particle Swarm Optimization) extremity learning machine based strip steel exit thickness predicting method |
| CN105354646B (en) * | 2015-12-04 | 2017-04-12 | 福州大学 | Power load forecasting method for hybrid particle swarm optimization and extreme learning machine |
| CN106650920A (en) * | 2017-02-19 | 2017-05-10 | 郑州大学 | Prediction model based on optimized extreme learning machine (ELM) |
| CN108256260A (en) * | 2018-02-05 | 2018-07-06 | 北京科技大学 | A kind of continuous casting billet quality Forecasting Methodology based on extreme learning machine |
-
2018
- 2018-11-20 CN CN201811382479.4A patent/CN109359723A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104200268A (en) * | 2014-09-03 | 2014-12-10 | 辽宁大学 | PSO (Particle Swarm Optimization) extremity learning machine based strip steel exit thickness predicting method |
| CN105354646B (en) * | 2015-12-04 | 2017-04-12 | 福州大学 | Power load forecasting method for hybrid particle swarm optimization and extreme learning machine |
| CN106650920A (en) * | 2017-02-19 | 2017-05-10 | 郑州大学 | Prediction model based on optimized extreme learning machine (ELM) |
| CN108256260A (en) * | 2018-02-05 | 2018-07-06 | 北京科技大学 | A kind of continuous casting billet quality Forecasting Methodology based on extreme learning machine |
Non-Patent Citations (2)
| Title |
|---|
| 张小晨: ""基于极限学习机的转炉炼钢终点预测模型研究"", 《中国优秀硕士学位论文全文数据库 工程科技Ⅰ辑》 * |
| 陶海龙: ""基于混合智能算法的铁路运量预测研究"", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110942208A (en) * | 2019-12-10 | 2020-03-31 | 萍乡市恒升特种材料有限公司 | Method for determining optimal production conditions of silicon carbide foam ceramic |
| CN110942208B (en) * | 2019-12-10 | 2023-07-07 | 萍乡市恒升特种材料有限公司 | Method for determining optimal production conditions of silicon carbide foam ceramic |
| CN113112011A (en) * | 2020-01-13 | 2021-07-13 | 中移物联网有限公司 | Data prediction method and device |
| CN113112011B (en) * | 2020-01-13 | 2024-02-27 | 中移物联网有限公司 | Data prediction method and device |
| CN111933223A (en) * | 2020-07-03 | 2020-11-13 | 大冶特殊钢有限公司 | Automatic batching method in steel-making alloying process |
| CN111933223B (en) * | 2020-07-03 | 2024-01-19 | 大冶特殊钢有限公司 | Automatic batching method in steelmaking alloying process |
| CN112541625A (en) * | 2020-12-07 | 2021-03-23 | 东北大学 | Self-adaptive multi-working-condition steel secondary energy generation amount dynamic prediction method |
| CN112541625B (en) * | 2020-12-07 | 2023-12-08 | 东北大学 | Self-adaptive multi-working-condition steel secondary energy generation amount dynamic prediction method |
| CN112992285A (en) * | 2020-12-31 | 2021-06-18 | 无锡东研信科科技研发有限公司 | IPSO-HKELM-based blast furnace molten iron silicon content prediction method |
| CN112992129A (en) * | 2021-03-08 | 2021-06-18 | 中国科学技术大学 | Attention-keeping mechanism monotonicity keeping method in voice recognition task |
| CN113192568A (en) * | 2021-03-15 | 2021-07-30 | 山东钢铁股份有限公司 | Refining furnace desulfurization terminal point forecasting method and system |
| CN113033705A (en) * | 2021-04-22 | 2021-06-25 | 江西理工大学 | Intelligent judgment and verification method for copper converter blowing slagging period end point based on pattern recognition |
| CN114638555A (en) * | 2022-05-18 | 2022-06-17 | 国网江西综合能源服务有限公司 | Power consumption behavior detection method and system based on multilayer regularization extreme learning machine |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109359723A (en) | Based on the converter terminal manganese content prediction technique for improving regularization extreme learning machine | |
| CN108676955B (en) | A kind of BOF Steelmaking Endpoint carbon content and temprature control method | |
| CN103882176B (en) | The online dynamic control method of a kind of convertor steelmaking process based on data-driven | |
| Wang et al. | Applying input variables selection technique on input weighted support vector machine modeling for BOF endpoint prediction | |
| CN111353656B (en) | Steel enterprise oxygen load prediction method based on production plan | |
| CN104630410A (en) | Real-time dynamic converter steelmaking quality prediction method based on data analysis | |
| CN104651559B (en) | Blast furnace liquid iron quality online forecasting system and method based on multivariable online sequential extreme learning machine | |
| CN106096788B (en) | Converter steelmaking process cost control method and system based on PSO _ ELM neural network | |
| CN106054836B (en) | Converter steelmaking process cost control method and system based on GRNN | |
| Wang et al. | Prediction model of end-point manganese content for BOF steelmaking process | |
| CN114611844B (en) | Method and system for determining alloy addition amount in converter tapping process | |
| CN110378064A (en) | A method of the prediction dead stock column temperature of blast furnace furnace core | |
| CN102163261A (en) | Case-reasoning-based molten steel temperature prediction method | |
| CN115522012B (en) | Method for controlling TSC stage oxygen blowing amount of large converter | |
| CN113962050A (en) | Oxygen scheduling calculation method combining production consumption prediction and pipe network calculation | |
| Xin et al. | Predicting the alloying element yield in a ladle furnace using principal component analysis and deep neural network | |
| Shi et al. | Key issues and progress of industrial big data-based intelligent blast furnace ironmaking technology | |
| Gu et al. | An improved CBR model using time-series data for predicting the end-point of a converter | |
| Wang et al. | An integrated CBR model for predicting endpoint temperature of molten steel in AOD | |
| CN103276136A (en) | Converter-steelmaking molten steel phosphorus-determination method based on sublance system | |
| CN116469481B (en) | LF refined molten steel composition forecasting method based on XGBoost algorithm | |
| CN111126550B (en) | Neural network molten steel temperature forecasting method based on Monte Carlo method | |
| CN116384254A (en) | Error correction method and system for predicting molten steel temperature of LF refining furnace | |
| Han et al. | BOF oxygen control by mixed case retrieve and reuse CBR | |
| CN109234491A (en) | A kind of BOF Steelmaking Endpoint manganese content prediction technique based on extreme learning machine |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190219 |