CN109541143B - Prediction method for actual components and physical property of slag containing volatile components along with time change - Google Patents
Prediction method for actual components and physical property of slag containing volatile components along with time change Download PDFInfo
- Publication number
- CN109541143B CN109541143B CN201811436920.2A CN201811436920A CN109541143B CN 109541143 B CN109541143 B CN 109541143B CN 201811436920 A CN201811436920 A CN 201811436920A CN 109541143 B CN109541143 B CN 109541143B
- Authority
- CN
- China
- Prior art keywords
- slag
- components
- neural network
- actual
- training
- 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
- 239000002893 slag Substances 0.000 title claims abstract description 141
- 230000000704 physical effect Effects 0.000 title claims abstract description 35
- 238000000034 method Methods 0.000 title claims abstract description 34
- 230000008859 change Effects 0.000 title claims abstract description 15
- 238000003062 neural network model Methods 0.000 claims abstract description 31
- 238000005259 measurement Methods 0.000 claims abstract description 11
- 238000001514 detection method Methods 0.000 claims abstract description 6
- 238000010791 quenching Methods 0.000 claims abstract description 5
- 230000000171 quenching effect Effects 0.000 claims abstract description 5
- 238000012549 training Methods 0.000 claims description 37
- 238000010438 heat treatment Methods 0.000 claims description 31
- 238000013528 artificial neural network Methods 0.000 claims description 16
- 238000002844 melting Methods 0.000 claims description 14
- 230000008018 melting Effects 0.000 claims description 14
- 238000012360 testing method Methods 0.000 claims description 13
- 239000000203 mixture Substances 0.000 claims description 10
- 230000004913 activation Effects 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- KKCBUQHMOMHUOY-UHFFFAOYSA-N sodium oxide Chemical compound [O-2].[Na+].[Na+] KKCBUQHMOMHUOY-UHFFFAOYSA-N 0.000 claims description 4
- VEXZGXHMUGYJMC-UHFFFAOYSA-M Chloride anion Chemical compound [Cl-] VEXZGXHMUGYJMC-UHFFFAOYSA-M 0.000 claims description 3
- KRHYYFGTRYWZRS-UHFFFAOYSA-M Fluoride anion Chemical compound [F-] KRHYYFGTRYWZRS-UHFFFAOYSA-M 0.000 claims description 3
- 229910001948 sodium oxide Inorganic materials 0.000 claims description 3
- 230000000875 corresponding effect Effects 0.000 description 19
- 230000008569 process Effects 0.000 description 11
- 238000003723 Smelting Methods 0.000 description 5
- 229910052681 coesite Inorganic materials 0.000 description 5
- 229910052906 cristobalite Inorganic materials 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 239000000377 silicon dioxide Substances 0.000 description 5
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N silicon dioxide Inorganic materials O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 5
- 229910052682 stishovite Inorganic materials 0.000 description 5
- 229910052905 tridymite Inorganic materials 0.000 description 5
- 239000000126 substance Substances 0.000 description 4
- YCKRFDGAMUMZLT-UHFFFAOYSA-N Fluorine atom Chemical compound [F] YCKRFDGAMUMZLT-UHFFFAOYSA-N 0.000 description 2
- 229910001634 calcium fluoride Inorganic materials 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 229910052731 fluorine Inorganic materials 0.000 description 2
- 239000011737 fluorine Substances 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000004321 preservation Methods 0.000 description 2
- 229910002974 CaO–SiO2 Inorganic materials 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- WUKWITHWXAAZEY-UHFFFAOYSA-L calcium difluoride Chemical compound [F-].[F-].[Ca+2] WUKWITHWXAAZEY-UHFFFAOYSA-L 0.000 description 1
- 238000005266 casting Methods 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 238000009749 continuous casting Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000004907 flux Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 238000005272 metallurgy Methods 0.000 description 1
- 239000000843 powder Substances 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000011734 sodium Substances 0.000 description 1
- 230000036962 time dependent Effects 0.000 description 1
- 230000004580 weight loss Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
Abstract
The invention relates to a method for predicting the change of the actual components and physical properties of slag containing volatile components along with time, which comprises the following steps: detecting and analyzing the initial components of the slag containing the volatile components; carrying out multi-group physical property detection on the slag containing the volatile component under certain temperature and time conditions to obtain slag physical property parameters under different conditions; quenching the slag to room temperature, and detecting and analyzing the actual components of the slag; and establishing a BP neural network model according to the corresponding relation between the initial components of the slag, the high-temperature physical parameters of the slag under different conditions and the actual components of the slag, wherein the input layer of the BP neural network model is the initial components of the slag, the temperature and the time, and the output layer is the actual components and the physical properties of the slag at the corresponding time. The prediction method takes time as a variable condition, solves the problem of inaccurate measurement of actual components when the performance of the slag containing volatile components is measured at high temperature, and obtains the corresponding relation between the physical property parameters of the slag and the actual components at different moments under certain temperature conditions.
Description
Technical Field
The invention relates to the technical field of metallurgy, in particular to a method for predicting actual components of slag containing volatile components and the time-dependent change rule of high-temperature physical property parameters of the corresponding slag by utilizing a BP neural network model.
Background
The metallurgical slag components and the physical and chemical properties thereof are closely related to the temperature control, the interface reaction characteristics, the inclusion removal and even the forward running of the smelting process in the smelting process. However, part of metallurgical slag components have the characteristic of easy volatilization, and the easy volatilization components can continuously volatilize along with time under the high-temperature condition, so that the slag components are continuously changed in smelting, the physical parameters of the slag are further changed, and the smooth smelting production is influenced. In the determination of the high-temperature properties of the slag, too, the finally determined slag properties do not actually correspond to the set initial slag composition due to the large volatilization of the volatile constituents. Such problems are prevalent in the melting and performance measurements of volatile component-containing slags. In recent years, with the maturation of new direct lead smelting process, researchers have focused on PbO-FeOx-CaO-SiO2The attention on the physicochemical properties and the reduction characteristics of ZnO polynary systems is multiplied, and a plurality of very successful research results are obtained successively. When the high-temperature physical property of the slag containing PbO and other volatile components is measured, the volatile components are inevitably volatilized, so that the components of the slag are changed to different degrees, the measured data is actually the physical property parameters of the slag after the change of the components of the slag, but not the initial set values of the components of the slag, and the changed actual components are not detected and analyzed; in addition, Na is generally added into the fluorine-containing continuous casting crystallizer casting powder2O、CaF2、Li2Additives such as O, MgO, MnO and the like to improve the physical and chemical properties of the mold flux, Na2O、CaF2The method plays a role in reducing the melting temperature and viscosity in the covering slag, but has the defects that volatile substances are easily generated by reaction at high temperature to change the slag components, so that the deviation of the slag components and the initial components occurs, and the slag performance is influenced; similarly, the high temperature of the slag system containing medium and high fluorine in the electroslag remelting processCaF in the tailings2Can react with other components to generate volatile gas, and influences the actual slag performance and the smooth production in the electroslag remelting process. In summary, the problem of the change of physical properties and components of the slag with time cannot be solved by using the existing slag property measuring method for the slag containing volatile components, and the longer the high-temperature retention time is, the larger the deviation of the physical properties and the components of the slag is, and an effective solution is not available at present.
An Artificial Neural Network (Artificial Neural Network) is called Neural Network (NN) for short, and is an information processing system simulating the structure and function of a physiological Neural Network. Among them, the BP back propagation algorithm is the most widely and well-developed algorithm applied in many neural network models, and is widely applied in the metallurgical industry due to its many advantages, and good results are obtained. Therefore, the method aims to predict the corresponding relation between the actual components of the slag containing the volatile components and the related high-temperature physical property parameters thereof by using the BP neural network model under the conditions of a certain temperature range and heating time (heat preservation duration).
Disclosure of Invention
The invention aims to provide a method for predicting the change of actual components and physical properties of actual slag containing volatile components along with time, aiming at solving the problem that the change of the components and the properties of the actual slag containing the volatile components along with time is difficult to determine when the high-temperature performance of the slag is measured by utilizing the prior art.
In order to achieve the technical purpose, the technical scheme adopted by the invention is as follows:
a method for predicting the change of actual components and physical properties of slag containing volatile components along with time comprises the following steps:
(1) detecting and analyzing initial components of the slag containing the volatile components;
(2) under the conditions that the heating temperature is 600-1600 ℃ and the heating time is 0-300 minutes, carrying out multi-group physical property detection on the slag containing the volatile components in the step one to obtain the physical property parameter measurement results of the slag under the conditions of different heating temperatures and different heating times;
(3) quenching the related slag in the second step to room temperature, and detecting and analyzing the actual components of the slag;
(4) according to the initial components of the slag determined in the step (1) and the physical parameters and actual components of the slag obtained by the slag in the steps (2) and (3) under the conditions of different heating temperatures and heating times, establishing a BP neural network model to predict the corresponding relation between the actual components and the physical parameters of the slag at the corresponding time, wherein the BP neural network model is established according to the following steps:
(41) acquiring slag physical property parameters and actual components under different heating temperature and heating time conditions in the steps (2) and (3) as sample data, performing normalization processing, and dividing the sample data into a training sample and a test sample;
(42) taking initial components, heating temperature and heating time of the slag as input layers of the BP neural network model, and taking actual components and physical parameters of the slag to be predicted as output layers of the BP neural network model at corresponding time of the slag; a hidden layer is arranged between the input layer and the output layer, and the number of hidden layer units is determined;
(43) selecting a random number as an initial weight to initialize the BP neural network model, and selecting an activation function;
(44) inputting the training sample into a BP neural network model for training, calculating a prediction result of an output layer, then calculating an error between the prediction result and the result measured in the step (3), judging whether the output error meets the requirement or the training frequency reaches the maximum training frequency, and if any one of the conditions is met, finishing the training; otherwise, adjusting the weight of each node and then executing next training;
(45) inputting a test sample for testing after training is finished, and calculating and recording a prediction result of the neural network and an error between the prediction result and an actual numerical value;
(46) adjusting the number of nodes of the hidden layer, reconstructing a BP neural network model, and repeating the steps from (43) to (45);
(47) and comparing the prediction performances, and selecting an optimal BP neural network model for prediction.
The slag comprises lead-containing slag, sodium oxide-containing slag, volatile fluoride-containing slag and volatile chloride-containing slag.
Further defined, the physical property parameters include melting point, viscosity, density, surface tension, electrical conductivity, and thermal conductivity.
Further defined, the activation function is selected
Further defined, the training samples are trained using the trainGDM algorithm.
In the measurement of the actual components of the slag containing the volatile components, the slag is subjected to the processes of temperature rise and heat preservation, so that the volatile components inevitably and continuously volatilize, the actual components of the slag are continuously changed, the longer the holding time under the high-temperature condition is, the greater the volatilization of the components is, and correspondingly, the performance measurement deviation is larger and larger.
According to the prediction method provided by the application, time is taken as a variable condition, accurate actual slag components are measured by quenching treatment and detection analysis of the volatilized high-temperature slag, and a neural network model is established, so that the problem of inaccurate measurement in high-temperature performance measurement caused by continuous change of high-temperature slag components containing volatile components along with time is solved, and the corresponding relation between the actual slag components and physical property parameters at corresponding moments under a certain temperature condition is obtained.
Drawings
FIG. 1 is a BP neural network performance parameter diagram of a PbO content prediction model after lead-containing slag volatilization;
FIG. 2 is a diagram of performance parameters of a lead-containing slag melting point prediction model BP neural network;
FIG. 3 is a BP neural network performance parameter diagram of a lead-containing slag viscosity prediction model;
FIG. 4 is a comparison graph of a predicted value and an experimental value of a PbO content prediction model at 180min after lead-containing slag volatilization;
FIG. 5 is a comparison graph of a predicted value and an experimental value of a PbO weight loss rate prediction model at 180min after lead-containing slag volatilization;
FIG. 6 is a graph showing a comparison between a predicted value and an experimental value of a model for predicting the melting point of lead-containing slag;
FIG. 7 is a comparison graph of a predicted value and an experimental value of a lead-containing slag viscosity prediction model at 180 min.
Detailed Description
In order that those skilled in the art can better understand the present invention, the following technical solutions are further described with reference to the accompanying drawings and examples.
The existing high-temperature physical property detection means has great result deviation in the process of measuring the slag containing volatile components, because the volatile components in the slag continuously volatilize at high temperature to cause the actual components of the slag to be always changed, the measured physical property parameter data actually correspond to the changed slag components, and the actual components of the slag are not further detected and analyzed after the performance is measured. Usually the measured physical property parameters are correlated with the initial slag composition, which presents a problem of incompatibility between the slag composition and the property measurements.
Aiming at the problem, the application provides a method for predicting the change of the actual components and physical properties of the slag containing volatile components along with time by using a BP neural network, which mainly comprises the following steps:
s1: the initial components of the slag containing volatile components (mainly including lead-containing slag, sodium oxide-containing slag, volatile fluoride-containing slag, chloride-containing slag and the like) are detected and analyzed.
S2: and (3) carrying out multiple groups of physical property detection on the volatile component-containing slag in the S1 under the conditions of heating temperature of 600-1600 ℃ and heating time of 0-300 minutes to obtain the measurement results of physical property parameters of the slag under different specified temperature and continuous heating time conditions, wherein the physical property parameters can be the melting point, viscosity, density, surface tension, electrical conductivity, thermal conductivity and the like of the slag.
S3: and (4) immediately quenching the related high-temperature slag in the S2 to room temperature (20-25 ℃), and detecting and analyzing actual components of the slag.
S4: and establishing a BP neural network model to predict the corresponding relation between the actual slag components and the physical parameters at corresponding time according to the initial slag components determined in the step S1 and the corresponding relation between the high-temperature physical parameters and the actual slag components of the slag obtained under the conditions of different heating temperatures and heating times in the step S2 and the step S3.
The BP neural network model establishing method comprises the following steps:
s41: and (4) acquiring the physical parameters and actual components of the slag under the conditions of certain heating temperature and heating time in the steps S2 and S3 as sample data, performing normalization processing, and dividing the sample data into a training sample and a test sample.
S42: the BP neural network model is provided with an input layer and an output layer, the initial components, the heating temperature and the heating time of the slag are used as the input layer of the BP neural network model, and the predicted actual components and physical parameters of the slag are used as the output layer of the BP neural network model at the corresponding time of the slag; and a hidden layer is arranged between the input layer and the output layer, and the number of hidden layer units is determined.
S43: and initializing the BP neural network model by using a random number as an initial weight, and selecting an activation function.
S44: inputting the training samples into a BP neural network model for training, calculating a prediction result of an output layer, then calculating an error between the prediction result and the result measured in the step S3, judging whether the output error meets the requirement or the training frequency reaches the maximum training frequency, if so, finishing the training, otherwise, adjusting the weight of each node and executing the next training.
S45: and after training is finished, inputting a test sample for testing, calculating and recording a prediction result of the neural network at this time and an error between the prediction result and an actual numerical value.
S46: and adjusting the number of nodes of the hidden layer, reconstructing a BP neural network model, and repeating S43 to S45.
S47: and comparing the prediction performances, and selecting an optimal BP neural network model for prediction.
The prediction method can be suitable for predicting the high-temperature physical and chemical properties and actual components of the slag containing the volatile components at different moments.
The invention is further described below by means of specific examples:
the actual composition of the high-temperature slag in a certain period of time, the corresponding melting point (hemispheric method) and the viscosity in a molten state at 1250 ℃ are measured as examples. A large number of experiments prove that the high-temperature volatilization process of the lead-containing slag is mainly the loss of lead element, so that only the volatilization of PbO is considered in the actual operation process.
For different FeO/SiO2And CaO/SiO2The initial composition of the high lead slag is measured together with the actual composition of the high lead slag at different times and the corresponding melting point (hemispherical melting temperature), as shown in table 1 below.
TABLE 1
In the testing process, A1-A9 groups of data are selected as training samples, and A10-A12 groups of data are selected as testing samples.
For different FeO/SiO2And CaO/SiO2The initial composition of the high lead slag of (2) was measured together with the actual composition of the high lead slag at different times and the viscosity in a molten state at 1250 ℃ corresponding to the measurement, as shown in Table 2 below.
TABLE 2
Wherein the groups B1, B2, B3 and B5 are used as training samples, and the groups B4, B6 and B7 are used as testing samples.
Establishing a model through the data of the two tables:
selecting an input layer and an output layer: according to the experimental scheme, the input layer group of the lead-containing slag is selected to be PbO, ZnO, FeO, CaO and SiO2Heating time t, and predicting the actual components of the high-temperature furnace slag at different moments as output; for the corresponding melting point of the hemispherical point, the viscosity of the molten state at 1250 ℃ at different times is taken as an output layerAnd (6) performing prediction.
Selecting the number of hidden layer units: after multiple tests, the number of lead slag hidden layer units is set to be 10.
Selection of an activation function:
selecting a learning rate: generally, the learning rate is selected to be 0.01-0.8, and the learning rate in the training process of the lead-containing slag is selected to be 0.1.
Determining an initial weight: random numbers are selected as initial weights.
Selection of the expected error: the error is generally selected based on the comparison training results. In general, we can train by comparing networks with different expected error values, and determine by comprehensive consideration after comparison, the method selects 1.0 × 10-9。
Normalization processing of the sample: and performing dimensionless normalization processing on the sample data to improve the prediction precision of the network and process the trained sample data to be between 0 and 1.
X’＝(Xi-Xi Min)/(Xi Max-Xi Min)
Wherein, Xi' is a value after treatment; xiIs the original data value; xi MaxIs the maximum of the data; xi MinIs the minimum value of the data.
The neural network weight array is trained by adopting a trainGDM algorithm, the number of hidden layer units of the network is determined to be 10, and the total error is set to be 1.0 multiplied by 10-9。
FIG. 1 is a BP neural network performance parameter diagram of a PbO content prediction model after 221 iterations of lead-containing slag volatilization, FIGS. 4 and 5 show comparison diagrams of predicted values and experimental values after corresponding iterative learning, and final training errors are 4.2469 × 10-5。
FIG. 2 is a BP neural network performance parameter graph of a lead-containing slag melting point prediction model after 6000 iterations, FIG. 6 shows a comparison graph of a predicted value and an experimental value after corresponding iterative learning, and finally training errors are respectively corresponding to the final training errorsIs 1.4524X 10-5。
FIG. 3 is a BP neural network performance parameter graph of a lead-containing slag viscosity prediction model after 1947 iterations, and FIG. 7 shows that the final training errors of a comparison graph of a predicted value and an experimental value after corresponding iterative learning are 1.5851 × 10-9。
The following is the language of the settings for the trainGDM algorithm model parameters:
net, train param, show 20; // show-display training iterative process interval times;
net.trainparam.lr ═ 0.1; // lr-learning rate;
net.trainparam.mc ═ 0.9; v/mc-momentum factor;
net, trainparam, epochs 7000; // epochs-maximum number of training sessions;
net.trainParam.goal＝1e-9(ii) a // coarse — target error;
the predicted results were compared to the actual results, as shown in table 3 below:
table 3 prediction results of lead slag melting point and viscosity based on improved BP algorithm
As can be seen from Table 3, the maximum error of the predicted value of the PbO content value after volatilization is 3.29 percent, and the minimum error is 0.98 percent; the maximum error of the predicted value of the melting point is 0.29 percent, the minimum error is 0.04 percent, the maximum error of the predicted viscosity is 2.70 percent, the minimum error is 0.98 percent, and the effect is better.
From the analysis, the prediction method can better predict the actual components and physical parameters of the high-temperature slag containing the volatile components at different moments, and if the sample size can be increased to improve the training effect, the prediction error can be further reduced.
The method for predicting the change of the actual components and physical properties of the slag containing the volatile components along with time provided by the invention is described in detail above. The description of the specific embodiments is only intended to facilitate an understanding of the method of the invention and its core ideas. It should be noted that, for those skilled in the art, it is possible to make various improvements and modifications to the present invention without departing from the principle of the present invention, and those improvements and modifications also fall within the scope of the claims of the present invention.
Claims (4)
1. A method for predicting the change of actual components and physical properties of slag containing volatile components along with time is characterized by comprising the following steps:
(1) detecting and analyzing initial components of the slag containing the volatile components;
(2) under the conditions that the heating temperature is 600-1600 ℃ and the heating time is 0-300 minutes, carrying out multi-group physical property detection on the slag containing the volatile components in the step one to obtain the physical property parameter measurement results of the slag under the conditions of different heating temperatures and heating times, wherein the physical property parameters comprise a melting point, viscosity, density, surface tension, electrical conductivity and thermal conductivity;
(3) quenching the related slag in the second step to room temperature, and detecting and analyzing the actual components of the slag;
(4) according to the initial components of the slag determined in the step (1) and the physical parameters and actual components of the slag obtained by the slag in the steps (2) and (3) under the conditions of different heating temperatures and heating times, establishing a BP neural network model to predict the corresponding relation between the actual components and the physical parameters of the slag at the corresponding time, wherein the BP neural network model is established according to the following steps:
(41) acquiring slag physical property parameters and actual components under different heating temperature and heating time conditions in the steps (2) and (3) as sample data, performing normalization processing, and dividing the sample data into a training sample and a test sample;
(42) taking initial components, heating temperature and heating time of the slag as input layers of the BP neural network model, and taking actual components and physical parameters of the slag to be predicted as output layers of the BP neural network model at corresponding time of the slag; a hidden layer is arranged between the input layer and the output layer, and the number of hidden layer units is determined;
(43) selecting a random number as an initial weight to initialize the BP neural network model, and selecting an activation function;
(44) inputting the training sample into a BP neural network model for training, calculating a prediction result of an output layer, then calculating an error between the prediction result and the result measured in the step (3), judging whether the output error meets the requirement or the training frequency reaches the maximum training frequency, and if any one of the conditions is met, finishing the training; otherwise, adjusting the weight of each node and then executing next training;
(45) inputting a test sample for testing after training is finished, and calculating and recording a prediction result of the neural network and an error between the prediction result and an actual numerical value;
(46) adjusting the number of nodes of the hidden layer, reconstructing a BP neural network model, and repeating the steps from (43) to (45);
(47) and comparing the prediction performances, and selecting an optimal BP neural network model for prediction.
2. The method for predicting the change of the actual composition and physical properties of the slag containing the volatile components with time according to claim 1, wherein the slag comprises lead-containing slag, sodium oxide-containing slag, volatile fluoride-containing slag and volatile chloride-containing slag.
3. The method as claimed in any one of claims 1-2, wherein the activation function is selected from the group consisting of
4. The method for predicting the actual composition and physical properties of the slag containing the volatile components according to any one of claims 1 to 2, wherein the training samples are trained by a trainGDM algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811436920.2A CN109541143B (en) | 2018-11-28 | 2018-11-28 | Prediction method for actual components and physical property of slag containing volatile components along with time change |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811436920.2A CN109541143B (en) | 2018-11-28 | 2018-11-28 | Prediction method for actual components and physical property of slag containing volatile components along with time change |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109541143A CN109541143A (en) | 2019-03-29 |
| CN109541143B true CN109541143B (en) | 2021-07-06 |
Family
ID=65852263
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811436920.2A Active CN109541143B (en) | 2018-11-28 | 2018-11-28 | Prediction method for actual components and physical property of slag containing volatile components along with time change |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109541143B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110276490A (en) * | 2019-06-20 | 2019-09-24 | 华北理工大学 | Continuous casting slag intelligent Forecasting based on LSTM model |
| CN111047474B (en) * | 2019-10-21 | 2024-03-22 | 贝壳技术有限公司 | Indoor harmful substance volatilization time estimation method, device and storage medium |
| CN111210877B (en) * | 2020-01-03 | 2023-11-24 | 山东天岳先进科技股份有限公司 | Method and device for deducing physical parameters |
| CN111737927A (en) * | 2020-06-24 | 2020-10-02 | 中国恩菲工程技术有限公司 | Melt composition analysis method and device for ladle refining process |
| CN112147175A (en) * | 2020-08-28 | 2020-12-29 | 西安建筑科技大学 | Method for measuring high-temperature physicochemical property of lead slag |
| CN112362842B (en) * | 2020-10-29 | 2022-04-29 | 华中科技大学 | Method for predicting release characteristics of harmful elements in fire coal by support vector regression method |
| CN114137017B (en) * | 2021-11-25 | 2024-01-23 | 西安建筑科技大学 | Evaluation method for influence of slag volatilization containing volatile components on melting point |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH09256021A (en) * | 1996-03-19 | 1997-09-30 | Kobe Steel Ltd | Device for calculating charging quantity of auxiliary raw material into converter |
| JPH10132462A (en) * | 1996-10-31 | 1998-05-22 | Kobe Steel Ltd | Controlling device of melting furnace |
| CN101793886A (en) * | 2010-03-24 | 2010-08-04 | 北京林业大学 | Sensory evaluation method for predicting fermented yogurt based on BP neural network |
| CN102004088A (en) * | 2010-11-09 | 2011-04-06 | 清华大学 | Method for measuring coal property on line based on neural network |
| CN103679268A (en) * | 2012-09-14 | 2014-03-26 | 宝钢不锈钢有限公司 | Blast furnace slag viscosity prediction method |
| CN103886224A (en) * | 2014-04-15 | 2014-06-25 | 江西理工大学 | Method for predicting melting temperature of metallurgical slag |
| CN104991051A (en) * | 2015-06-30 | 2015-10-21 | 华侨大学 | Method for predicting concrete strength based on hybrid model |
| CN105352918A (en) * | 2015-11-13 | 2016-02-24 | 湖南大学 | SVR-based real-time element concentration monitoring method and SVR-based real-time element concentration monitoring device for laser-aided direct metal deposition processes |
| CN105445138A (en) * | 2015-11-27 | 2016-03-30 | 西安建筑科技大学 | Slag actual ingredient performance detection method for slag containing volatile components |
| CN106248707A (en) * | 2016-07-21 | 2016-12-21 | 金川集团股份有限公司 | Ni, Cu, Fe, S, CaO, MgO, SiO in a kind of quickly translocation flash smelting slag2, the analysis method of Pb, Zn, As, Sb, Bi |
| CN107045658A (en) * | 2017-04-15 | 2017-08-15 | 西安科技大学 | A kind of Forecasting Methodology of coal ash viscosity-temperature characteristic |
| CN107287380A (en) * | 2017-07-04 | 2017-10-24 | 安徽工业大学 | A kind of slag composition on-line prediction method |
| CN108241789A (en) * | 2018-01-12 | 2018-07-03 | 内蒙古电力勘测设计院有限责任公司 | The carbon containing measuring method of boiler slag and equipment based on adaptive weighted AWLS-SVR |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180230390A1 (en) * | 2017-02-10 | 2018-08-16 | United States Department Of Energy | Slag management toolset for determining optimal gasification temperatures |
-
2018
- 2018-11-28 CN CN201811436920.2A patent/CN109541143B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH09256021A (en) * | 1996-03-19 | 1997-09-30 | Kobe Steel Ltd | Device for calculating charging quantity of auxiliary raw material into converter |
| JPH10132462A (en) * | 1996-10-31 | 1998-05-22 | Kobe Steel Ltd | Controlling device of melting furnace |
| CN101793886A (en) * | 2010-03-24 | 2010-08-04 | 北京林业大学 | Sensory evaluation method for predicting fermented yogurt based on BP neural network |
| CN102004088A (en) * | 2010-11-09 | 2011-04-06 | 清华大学 | Method for measuring coal property on line based on neural network |
| CN103679268A (en) * | 2012-09-14 | 2014-03-26 | 宝钢不锈钢有限公司 | Blast furnace slag viscosity prediction method |
| CN103886224A (en) * | 2014-04-15 | 2014-06-25 | 江西理工大学 | Method for predicting melting temperature of metallurgical slag |
| CN104991051A (en) * | 2015-06-30 | 2015-10-21 | 华侨大学 | Method for predicting concrete strength based on hybrid model |
| CN105352918A (en) * | 2015-11-13 | 2016-02-24 | 湖南大学 | SVR-based real-time element concentration monitoring method and SVR-based real-time element concentration monitoring device for laser-aided direct metal deposition processes |
| CN105445138A (en) * | 2015-11-27 | 2016-03-30 | 西安建筑科技大学 | Slag actual ingredient performance detection method for slag containing volatile components |
| CN106248707A (en) * | 2016-07-21 | 2016-12-21 | 金川集团股份有限公司 | Ni, Cu, Fe, S, CaO, MgO, SiO in a kind of quickly translocation flash smelting slag2, the analysis method of Pb, Zn, As, Sb, Bi |
| CN107045658A (en) * | 2017-04-15 | 2017-08-15 | 西安科技大学 | A kind of Forecasting Methodology of coal ash viscosity-temperature characteristic |
| CN107287380A (en) * | 2017-07-04 | 2017-10-24 | 安徽工业大学 | A kind of slag composition on-line prediction method |
| CN108241789A (en) * | 2018-01-12 | 2018-07-03 | 内蒙古电力勘测设计院有限责任公司 | The carbon containing measuring method of boiler slag and equipment based on adaptive weighted AWLS-SVR |
Non-Patent Citations (3)
| Title |
|---|
| Predicting coal ash slag flow characteristics (viscosity model for the Al203-Ca0-‘Fe0’-Si02 system);Alex Kondratiev 等;《FUEL》;20010407;第1989-2000页 * |
| 基于神经网络的多组分保护渣高温物理性能预测模型研究;向嵩;《中国优秀博硕士学位论文全文数据库 (硕士) 工程科技Ⅰ辑》;20031215(第04期);第1-65页 * |
| 镍基高温合金中丫和守相点阵常数的简捷人工神经网络预测法;骆宇时 等;《金属学报》;20030930;第3 9 卷(第9 期);第897-902页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109541143A (en) | 2019-03-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109541143B (en) | Prediction method for actual components and physical property of slag containing volatile components along with time change | |
| CN107526927B (en) | Blast furnace molten iron quality online robust soft measurement method | |
| CN108311544B (en) | A kind of roll-force Parameter Self-learning method and device | |
| CN110598958A (en) | Steel ladle grading management analysis method and system | |
| TW201145186A (en) | Method for buliding adaptive soft sensor | |
| CN108897354B (en) | Aluminum smelting process hearth temperature prediction method based on deep belief network | |
| CN111310348A (en) | Material constitutive model prediction method based on PSO-LSSVM | |
| CN114611844A (en) | Method and system for determining alloy addition amount in converter tapping process | |
| WO2023087917A1 (en) | Cognitive decision-making evaluation method and system based on multi-dimensional hierarchical drift diffusion model | |
| CN106991074B (en) | Accelerated degradation test optimization design method based on principle of acceleration factor invariance | |
| JP2007052739A (en) | Method and device for generating model, method and device for predicting state, and method and system for adjusting state | |
| CN114764547A (en) | Heating furnace steel billet temperature prediction method and device integrating mechanism and deep learning | |
| CN117148169B (en) | Battery service time prediction method, system, equipment and medium based on big data | |
| CN110610140B (en) | Training method, device and equipment of face recognition model and readable storage medium | |
| JP5582487B2 (en) | Process state prediction method | |
| Xu et al. | Influence of short-range order on diffusion in multiprincipal element alloys from long-time atomistic simulations | |
| CN110956014A (en) | Method for predicting service life of PCB under action of different wire spacing and voltage | |
| CN112509642B (en) | Online prediction method for blast furnace slag viscosity | |
| CN115938496A (en) | Quality Estimation Method Based on XGBoost Algorithm | |
| JP4365536B2 (en) | Steel production process abnormality judgment device | |
| CN112992284A (en) | Sintered mineral quality forecasting method based on PSO-ELM algorithm | |
| CN107977742B (en) | Construction method of medium-long term power load prediction model | |
| CN117553921B (en) | Converter molten steel temperature prediction method, system, terminal and storage medium | |
| Yu et al. | Research on quality prediction of iron ore green pellets based on optimized neural network | |
| JP6933196B2 (en) | Blast furnace unloading speed prediction model learning method, blast furnace unloading speed prediction method, blast furnace operation guidance method, blast furnace unloading speed control method, hot metal manufacturing method, blast furnace operation method, and blast furnace unloading speed prediction Model learning device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20231016 Address after: 710119 No.1, north section of development avenue, new industrial park, high tech Zone, Xi'an City, Shaanxi Province Patentee after: Xi'an Huijin Technology Co.,Ltd. Address before: 710055 Yanta Road 13, Xi'an City, Shaanxi Province Patentee before: XIAN University OF ARCHITECTURE AND TECHNOLOG |
|
| TR01 | Transfer of patent right |