Converter molten steel temperature measurement and carbon determination method adopting multispectral infrared detection technology

Abstract

The invention provides a converter molten steel temperature measurement and carbon determination method adopting a multispectral infrared detection technology, which belongs to the technical field of steelmaking and comprises the following steps: detecting molten steel in a converter by using a multispectral thermodetector; collecting infrared spectrum data of the molten steel; according to the Hough law and known spectral line characteristics, the temperature of the molten steel is reversely deduced through spectral data; through multi-channel data processing and algorithm optimization, the molten steel temperature is calculated in real time, and the measurement precision is improved. The sublance system directly extracts a high-temperature molten steel sample in the converter smelting process and pretreats the high-temperature molten steel sample; irradiating the pretreated sample by using an excitation light source, and exciting the carbon element in the sample to emit a spectral signal with a specific wavelength; selecting a specific wavelength in the molten steel for spectral analysis; acquiring the concentration of carbon in the molten steel according to the emission or absorption characteristics of gas molecules in the molten steel; the method comprises the following steps: establishing a model by utilizing multispectral data, and establishing a relationship between spectral characteristics and carbon content through a regression analysis or machine learning algorithm;

Description

Technical Field

[0001] The invention relates to the technical field of steelmaking, in particular to a method for measuring temperature and determining carbon of molten steel in a converter by adopting a multi-spectrum infrared detection technology. Background Art

[0002] With the advent of the Industrial 4.0 era, artificial intelligence technology is being widely used in various industrial scenarios to improve production efficiency and product quality. As the core link of steel production, converter steelmaking has great significance for the introduction of AI technology for detection and control.

[0003] During the converter steelmaking process, the reaction inside the furnace is complex, and traditional control methods make it difficult to achieve accurate control of key parameters such as temperature and composition. Currently, the surface temperature of the continuous casting billet cannot be measured online continuously and the temperature measurement accuracy is not high. There are problems such as electromagnetic interference, water vapor interference, high temperature environment, unknown emissivity, and dust blocking the lens.

[0004] Therefore, this field urgently needs a technical solution based on AI-based special detection and control technology to improve the automation and intelligence level of the steelmaking process.

[0005] The information disclosed in this background technology section is only intended to enhance the understanding of the overall background of the invention and should not be regarded as an acknowledgment or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a method for measuring temperature and determining carbon of molten steel in a converter by adopting a multi-spectrum infrared detection technology.

[0007] To achieve the above object, the present invention provides the following solutions:

[0008] A method for measuring temperature and determining carbon content of molten steel in a converter using multi-spectral infrared detection technology, comprising:

[0009] Calibrate with standard samples of known temperature to obtain the relationship between temperature and infrared radiation intensity;

[0010] Use a multi-spectral pyrometer to detect molten steel in the converter;

[0011] Collect infrared spectrum data of molten steel;

[0012] According to Hott's law and known spectral line characteristics, the temperature of molten steel can be inferred from the spectral data.

[0013] Through multi-channel data processing and algorithm optimization, the molten steel temperature is calculated in real time to improve the measurement accuracy;

[0014] Optionally, also include:

[0015] The auxiliary gun system directly extracts high-temperature molten steel samples during the converter smelting process and pre-treats them;

[0016] The pretreated sample is irradiated with an excitation light source to excite the carbon element in the sample to emit a spectral signal of a specific wavelength;

[0017] Select specific wavelengths in molten steel for spectral analysis;

[0018] Obtain the carbon concentration in molten steel based on the emission or absorption characteristics of gas molecules in molten steel;

[0019] Use multispectral data to build models and establish relationships between spectral features and carbon content through regression analysis or machine learning algorithms;

[0020] Update the model in real time to ensure accuracy and timeliness;

[0021] The measured temperature and carbon content are displayed in real time on the operation interface for the operator's reference;

[0022] Combine historical data and trend analysis to provide relevant recommendations.

[0023] Optionally, the mechanical structure of the sub-gun is used to detect the molten steel with a dual fiber optic detector, with a silicon carbide protection tube protected by sapphire glass at the front end, to measure the endpoint temperature and endpoint carbon respectively.

[0024] Optionally, the specific wavelength includes characteristic bands related to CO and C.

[0025] Optionally, the method of using multi-spectral data to establish a model and establishing a relationship between spectral features and carbon content through regression analysis or machine learning algorithm specifically includes:

[0026] The collected multispectral data are divided into a training set and a data set;

[0027] Apply the selected regression analysis model or machine learning algorithm on the training set to train the model;

[0028] Evaluate model performance on the test set and use metrics to measure model accuracy

[0029] Use cross-validation methods to adjust the model's hyperparameters to improve performance;

[0030] Based on the importance assessment of the model, the best feature set is selected to reduce computational complexity and improve prediction performance;

[0031] Use scatter plots and regression curves to show the prediction effect of the model and judge the model fit;

[0032] Analyze the contribution of different spectral features to carbon content prediction to guide subsequent data collection and experimental design;

[0033] Integrate the trained model into the online monitoring system to estimate the carbon content in molten steel in real time;

[0034] Compare the model's predictions with the actual analysis results and continuously improve the model.

[0035] Optionally, the regression analysis model includes: linear regression, ridge regression or Lasso regression;

[0036] The machine learning algorithms include: support vector machine regression, decision tree regression, random forest regression, gradient boosting tree and deep learning algorithms.

[0037] Optionally, the indicators include: root mean square error and determination coefficient.

[0038] Optionally, the dual-fiber detector includes an optical lens, an optical fiber, a fiber connector, an incident slit, a grating, a detector, a focusing reflector and a collimating mirror; the optical lens is connected to the fiber connector via an optical fiber, and the fiber connector is connected to the incident slit.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] From the perspective of accuracy, multi-spectral temperature measurement technology achieves high-precision temperature measurement by capturing and analyzing the radiation characteristics of the target object in different spectral bands. This non-contact measurement method effectively avoids the errors that may be introduced by traditional contact temperature measurement, such as the influence of probe heat conduction and the limitations of temperature measurement point selection. At the same time, the spectrometer carbon and oxygen determination technology uses the principle of spectral analysis to directly quantitatively determine the carbon and oxygen content in the converter molten steel. Its high sensitivity and high resolution characteristics ensure the accuracy of the measurement results, far exceeding traditional chemical analysis or sampling methods, and providing more reliable process parameter support for the steel smelting process.

[0041] The improvement of efficiency is a highlight of the new technology that cannot be ignored. Multi-spectral temperature measurement technology can achieve rapid response and real-time measurement, greatly shortening the waiting time in traditional temperature measurement methods and improving the continuity of the production process. The spectrometer carbon and oxygen determination technology significantly accelerates the data processing speed through the automated data collection and analysis system, allowing operators to obtain key information in real time and quickly adjust process parameters, thereby optimizing the smelting cycle and improving production efficiency.

[0042] In terms of cost, although the initial equipment purchase cost may be high, in the long run, the economic advantages of the new technology gradually emerge. High-precision measurement reduces the waste of raw materials and waste caused by errors, reducing production costs; improved efficiency means shorter smelting cycles and higher capacity utilization, indirectly reducing the fixed cost per unit product. The new technology has a high level of intelligence and automation, which reduces manual intervention and operational errors, and reduces labor costs and maintenance costs.

[0043] Finally, from the perspective of environmental protection and safety, the application of multi-spectral temperature measurement and spectrometer carbon and oxygen determination technology effectively reduces the environmental pollution and safety hazards that may be caused by traditional methods. For example, it reduces thermal radiation pollution to the environment and the emission of harmful gases; at the same time, non-contact measurement reduces the direct exposure risk of operators and improves the safety of the working environment. Accurate process control also helps to reduce safety accidents caused by improper operation or fluctuations in process parameters, and improves the safety and stability of overall production.

[0044] The deep integration of multi-spectral temperature measurement and spectrometer carbon and oxygen determination technology into the converter auxiliary gun system is an important step in the intelligent upgrade of the steel industry. The implementation of this technical solution will not only help improve the level of steelmaking automation, optimize the production process, and reduce errors and risks caused by human factors, but also significantly improve product quality and enhance the market competitiveness of enterprises. With the continuous maturity and application promotion of technology, the steel industry will usher in a more efficient, green and intelligent development future.

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0046] Figure 1 A schematic flow chart of a method for measuring the temperature of molten steel in a converter using multi-spectral infrared detection technology provided in an embodiment of the present invention.

[0047] Figure 2 A schematic flow chart of a method for determining carbon content in converter molten steel using multi-spectral infrared detection technology provided in an embodiment of the present invention.

[0048] Figure 3 A schematic structural diagram of a converter auxiliary gun system provided in an embodiment of the present invention.

[0049] Figure 4 A schematic diagram of the structure of a dual-fiber detector provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] The purpose of the present invention is to provide a technical solution based on AI-based special detection and control technology to improve the automation and intelligence level of the steelmaking process.

[0052] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0053] Embodiment 1:

[0054] In order to realize automatic, continuous and sustained measurement of endpoint carbon content and endpoint temperature, the spectral information model is used to realize the measurement of endpoint temperature and endpoint carbon content. The spectrometer is used to obtain the spectral information of molten steel, and the influence of complex environment on the measurement is avoided through structural design and algorithm research.

[0055] In terms of algorithm, the correspondence between the radiation spectrum of molten steel and its composition is first established, and the endpoint carbon content is calculated through the characteristic peak of the emission spectrum. Secondly, the endpoint temperature of the molten steel is calculated based on the non-peak area of ​​the emission spectrum combined with the multi-spectral radiation temperature measurement algorithm.

[0056] Auxiliary gun spectrum detection: Place the optical fiber into the original auxiliary gun, lead out the spectrum, and then separate two spectrum beams to the multi-spectrum thermometer and spectrum carbon oxygen analyzer respectively, so as to analyze the temperature, C, and O2 in the molten steel in real time, replacing the auxiliary gun.

[0057] This embodiment provides a method for measuring temperature and determining carbon content of molten steel in a converter using multi-spectral infrared detection technology. Figure 1 As shown, including:

[0058] Calibrate with standard samples of known temperature to obtain the relationship between temperature and infrared radiation intensity;

[0059] Use a multi-spectral pyrometer to detect molten steel in the converter;

[0060] Collect infrared spectrum data of molten steel;

[0061] According to Hott's law and known spectral line characteristics, the temperature of molten steel can be inferred from the spectral data.

[0062] Through multi-channel data processing and algorithm optimization, the molten steel temperature is calculated in real time to improve the measurement accuracy;

[0063] Further, such as Figure 2 As shown, it also includes:

[0064] The auxiliary gun system directly extracts high-temperature molten steel samples during the converter smelting process and pre-treats them;

[0065] The pretreated sample is irradiated with an excitation light source to excite the carbon element in the sample to emit a spectral signal of a specific wavelength;

[0066] Select specific wavelengths in molten steel for spectral analysis;

[0067] Obtain the carbon concentration in molten steel based on the emission or absorption characteristics of gas molecules in molten steel;

[0068] Use multispectral data to build models and establish relationships between spectral features and carbon content through regression analysis or machine learning algorithms;

[0069] Update the model in real time to ensure accuracy and timeliness;

[0070] The measured temperature and carbon content are displayed in real time on the operation interface for the operator's reference;

[0071] Combine historical data and trend analysis to provide relevant recommendations.

[0072] Further, such as Figure 3 As shown, the mechanical structure of the auxiliary gun is utilized to detect the molten steel with a dual optical fiber detector, and a silicon carbide protection tube protected by sapphire glass is provided at the front end to measure the endpoint temperature and endpoint carbon respectively.

[0073] Furthermore, the specific wavelength includes characteristic bands related to CO and C.

[0074] Furthermore, the use of multi-spectral data to establish a model and establish a relationship between spectral features and carbon content through regression analysis or machine learning algorithms specifically includes:

[0075] The collected multispectral data are divided into a training set and a data set;

[0076] Apply the selected regression analysis model or machine learning algorithm on the training set to train the model;

[0077] Evaluate model performance on the test set and use metrics to measure model accuracy

[0078] Use cross-validation methods to adjust the model's hyperparameters to improve performance;

[0079] Based on the importance assessment of the model, the best feature set is selected to reduce computational complexity and improve prediction performance;

[0080] Use scatter plots and regression curves to show the prediction effect of the model and judge the model fit;

[0081] Analyze the contribution of different spectral features to carbon content prediction to guide subsequent data collection and experimental design;

[0082] Integrate the trained model into the online monitoring system to estimate the carbon content in molten steel in real time;

[0083] Compare the model's predictions with the actual analysis results and continuously improve the model.

[0084] Furthermore, the regression analysis model includes: linear regression, ridge regression or Lasso regression;

[0085] The machine learning algorithms include: support vector machine regression, decision tree regression, random forest regression, gradient boosting tree and deep learning algorithms.

[0086] Furthermore, the indicators include: root mean square error and determination coefficient.

[0087] Further, such as Figure 4 As shown, the dual-fiber detector includes an optical lens, an optical fiber, a fiber connector, an incident slit, a grating, a detector, a focusing reflector and a collimating mirror; the optical lens is connected to the fiber connector through an optical fiber, and the fiber connector is connected to the incident slit.

[0088] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0089] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for measuring temperature and determining carbon content of molten steel in a converter using multi-spectral infrared detection technology, characterized in that: include: Calibrate with standard samples of known temperature to obtain the relationship between temperature and infrared radiation intensity; Use a multi-spectral thermometer to detect molten steel in the converter; Collect infrared spectrum data of molten steel; According to Hott's law and known spectral line characteristics, the temperature of molten steel can be inferred from the spectral data; Through multi-channel data processing and algorithm optimization, the molten steel temperature is calculated in real time to improve measurement accuracy.

2. The method for measuring temperature and determining carbon content of molten steel in a converter using multi-spectral infrared detection technology according to claim 1, characterized in that: Also includes: The auxiliary gun system directly extracts high-temperature molten steel samples during the converter smelting process and pre-treats them; The pretreated sample is irradiated with an excitation light source to excite the carbon element in the sample to emit a spectral signal of a specific wavelength; Select specific wavelengths in molten steel for spectral analysis; Obtain the carbon concentration in molten steel based on the emission or absorption characteristics of gas molecules in molten steel; Use multispectral data to build models and establish relationships between spectral features and carbon content through regression analysis or machine learning algorithms; Update the model in real time to ensure accuracy and timeliness; The measured temperature and carbon content are displayed in real time on the operation interface for the operator's reference; Combine historical data and trend analysis to provide relevant recommendations.

3. The method for measuring temperature and determining carbon content of molten steel in a converter using multi-spectral infrared detection technology according to claim 1, characterized in that: Utilizing the mechanical structure of the auxiliary gun, a dual-fiber optical detector is used to detect the molten steel, with a silicon carbide protection tube protected by sapphire glass at the front end to measure the endpoint temperature and endpoint carbon respectively.

4. The method for measuring temperature and determining carbon content of molten steel in a converter using multi-spectral infrared detection technology according to claim 1, characterized in that: The specific wavelengths include characteristic bands associated with CO and C.

5. The method for measuring temperature and determining carbon content of molten steel in a converter using multi-spectral infrared detection technology according to claim 1, characterized in that: The method of using multi-spectral data to establish a model and establishing a relationship between spectral features and carbon content through regression analysis or machine learning algorithm specifically includes: The collected multispectral data are divided into a training set and a data set; Apply the selected regression analysis model or machine learning algorithm on the training set to train the model; Evaluate model performance on the test set and use metrics to measure model accuracy Use cross-validation methods to adjust the model's hyperparameters to improve performance; Based on the importance assessment of the model, the best feature set is selected to reduce computational complexity and improve prediction performance; Use scatter plots and regression curves to show the prediction effect of the model and judge the model fit; Analyze the contribution of different spectral features to carbon content prediction to guide subsequent data collection and experimental design; Integrate the trained model into the online monitoring system to estimate the carbon content in molten steel in real time; Compare the model's predictions with the actual analysis results and continuously improve the model.

6. The method for measuring temperature and determining carbon content of molten steel in a converter using multi-spectral infrared detection technology according to claim 4, characterized in that: The regression analysis model includes: linear regression, ridge regression or Lasso regression; The machine learning algorithms include: support vector machine regression, decision tree regression, random forest regression, gradient boosting tree and deep learning algorithms.

7. The method for measuring temperature and determining carbon content of molten steel in a converter using multi-spectral infrared detection technology according to claim 4, characterized in that: The indicators include: root mean square error and determination coefficient.

8. The method for measuring temperature and determining carbon content of molten steel in a converter using multi-spectral infrared detection technology according to claim 2, characterized in that: The dual-fiber detector comprises an optical lens, an optical fiber, an optical fiber connector, an incident slit, a grating, a detector, a focusing reflector and a collimating mirror; the optical lens is connected to the optical fiber connector through the optical fiber, and the optical fiber connector is connected to the incident slit.

Images