Iron-steel interface iron water temperature drop prediction system and method with real-time correction function

Abstract

The invention discloses an iron-steel interface molten iron temperature drop prediction system and method with a real-time correction function. The system comprises a resume data unit, a real-time correction data unit and a molten iron temperature drop prediction model unit. According to the method, the shell temperature, the environment temperature, the rain and snow amount and the light intensity of the iron-steel interface torpedo ladle or iron ladle are obtained through the infrared temperature measuring sensor, the environment temperature sensor, the rain and snow sensor and the photosensitive sensor; then, according to the real-time process data of the iron-steel interface, the molten iron temperature drop of the iron-steel interface is predicted through a molten iron temperature drop prediction model unit, and uncorrected molten iron temperature drop information is obtained; and finally, the shell temperature of the torpedo ladle or the iron ladle, the environment temperature, the rain and snow amount and the light intensity are embedded into the molten iron temperature drop prediction model unit in real time, the influence factor coefficient of the molten iron temperature drop model is corrected in real time through the shell temperature, the environment temperature, the rain and snow amount and the light intensity, and high-precision molten iron temperature drop information is obtained.

Description

Technical Field

[0001] This invention relates to the field of iron and steel smelting technology, and in particular to a system and method for predicting the temperature drop of molten iron at the iron-steel interface with real-time correction. Background Technology

[0002] The process interface of the blast furnace-converter section (also known as the iron-steel interface) refers to a relatively flexible interface that exists between the two relatively rigid processes of ironmaking and steelmaking. It includes the connection, transmission, matching and buffering of multidimensional material flow between the ironmaking and steelmaking processes. Its tasks are to receive, transport, pre-treat, store and buffer molten iron, accurately control the amount of molten iron added, and ensure the rapid turnover of the iron-receiving container.

[0003] The temperature drop of molten iron is one of the core factors of concern at the steel-iron interface. Studies have shown that replacing molten iron with scrap steel can significantly reduce the overall energy consumption per ton of steel in integrated steel processing plants, with each ton of scrap steel replacing molten iron reducing the consumption by 0.45 tons of standard coal. If the temperature of the molten iron entering the converter increases by 20°C, the iron-to-iron ratio can be reduced by 0.8%. Therefore, predicting the temperature drop of molten iron in advance allows for real-time adjustment of the iron-to-iron ratio, maximizing the use of scrap steel. This not only helps reduce energy consumption and carbon emissions but also lowers production costs.

[0004] To predict the temperature drop of molten iron, patent document No. 202011390081.2 discloses a method, apparatus, and terminal device for predicting the temperature drop of molten iron at the iron-steel interface. The method includes: constructing a data mart for the iron-steel interface; obtaining training samples from the data mart to train a target model and obtain a prediction model for the temperature drop of molten iron at the iron-steel interface; performing segmented mechanism modeling based on process information to obtain a data-driven mechanism model for the temperature drop of molten iron at the iron-steel interface; coupling the data-driven prediction model for the temperature drop of molten iron at the iron-steel interface and the mechanism model for the temperature drop of molten iron at the iron-steel interface to obtain a coupled model for the temperature drop of molten iron at the iron-steel interface; and using the coupled model for the temperature drop of molten iron at the iron-steel interface to predict the temperature drop information of molten iron.

[0005] The above-mentioned technical solution can accurately predict the temperature change of molten iron online, which facilitates the subsequent control of the temperature drop of molten iron and provides molten iron with stable composition and temperature. It is of great significance for improving the energy efficiency of the iron-steel interface, reducing costs and increasing efficiency, improving production efficiency and clean production.

[0006] However, the above-mentioned technical solutions, which correct the temperature coefficient through offline mechanisms or data-driven methods, still have the following shortcomings:

[0007] 1) Lack of real-time analysis of factors affecting temperature drop (especially weather factors). Molten iron ladles or torpedo ladles at the iron-steel interface need to be transported outdoors. Weather changes have a significant impact on the temperature drop of molten iron, such as the temperature difference between summer and winter, the temperature drop between sunny and rainy / snowy days, and the temperature difference between day and night.

[0008] 2) The prediction of molten iron temperature drop is based solely on the coupling of mechanism simulation and data-driven methods, lacking online measurement correction.

[0009] Therefore, it is necessary to improve the existing technology to overcome the aforementioned defects. Summary of the Invention

[0010] The purpose of this invention is to provide a system and method for predicting the temperature drop of molten iron at the iron-steel interface with real-time correction, so as to solve the problems existing in the prior art.

[0011] The above-mentioned technical objective of the present invention is achieved through the following technical solution:

[0012] A real-time corrected system for predicting the temperature drop of molten iron at the iron-steel interface, comprising:

[0013] A history data unit is used to construct a history database for torpedo cans or ladles with steel-to-iron interfaces.

[0014] The real-time correction data unit acquires external environmental data in real time through infrared temperature sensors or pre-embedded thermocouples, ambient temperature sensors, rain and snow sensors, and photosensors.

[0015] The hot metal temperature drop prediction model unit adopts a prediction model based on mechanism, a prediction model driven by data, or a prediction model coupled with mechanism and data driven by data. It uses external environmental data to correct the prediction accuracy of the prediction model in real time, thereby obtaining high-precision hot metal temperature drop information.

[0016] Furthermore, the history data unit is used to provide history data on the changes in influencing factors required by the molten iron temperature drop prediction model unit. The history data includes the history of the torpedo ladle or iron ladle, including real-time scheduling and process planning, real-time operation process, small, medium and large cycle maintenance and baking data.

[0017] Furthermore, the infrared sensor or thermocouple of the real-time correction data unit is used to acquire temperature data in real time during the operation of the torpedo canister or iron ladle; the ambient temperature sensor, rain and snow sensor and photosensitive sensor are used to acquire the temperature, amount of rain and snow and light intensity of the environment in which the torpedo canister or iron ladle is located in real time.

[0018] Furthermore, the mechanism model includes the temperature drop mechanism of the blast furnace tapping trough, the temperature drop mechanism of the torpedo ladle or iron ladle receiving iron, the temperature drop mechanism of three-stage or single-stage desulfurization, the temperature drop mechanism of the torpedo ladle or iron ladle adding iron, and the temperature drop mechanism of the torpedo ladle or iron ladle adding scrap steel or metal oxides; the data-driven algorithm model includes statistical machine learning algorithms and deep learning algorithms.

[0019] A method for predicting the temperature drop of molten iron at the iron-steel interface with real-time correction includes the following steps:

[0020] 1) The outer shell temperature of the torpedo canister or iron ladle at the iron-steel interface, the ambient temperature, the amount of rain and snow, and the light intensity are obtained through infrared temperature sensors, ambient temperature sensors, rain and snow sensors, and photosensors.

[0021] 2) Based on the real-time process data of the iron-steel interface, the molten iron temperature drop at the iron-steel interface is predicted by the mechanism prediction model, the data-driven algorithm mechanism prediction model, or the mechanism and data-driven coupled molten iron temperature drop prediction model, so as to obtain the uncorrected molten iron temperature drop information.

[0022] 3) The outer shell temperature of the torpedo ladle or iron ladle, the ambient temperature, the amount of rain and snow and the light intensity are embedded in the mechanism prediction model, the data-driven algorithm mechanism prediction model or the mechanism and data-driven coupled molten iron temperature drop prediction model in real time. The influence factor coefficients of the molten iron temperature drop model are corrected in real time using the outer shell temperature, the ambient temperature, the amount of rain and snow and the light intensity to obtain high-precision molten iron temperature drop information.

[0023] Furthermore, the infrared temperature sensor is arranged on both sides of the running track or path of the torpedo canister or iron ladle to obtain the outer shell temperature of the torpedo canister or iron ladle.

[0024] The ambient temperature sensor, rain and snow sensor, and light sensor are arranged in the area near the torpedo canister or ladle to obtain the ambient temperature, rain and snow amount, and light intensity in a timely manner; if the ambient temperature is obtained from the network system, the ambient temperature sensor is removed.

[0025] Furthermore, the mechanism prediction model, data-driven algorithm prediction model, or mechanism-data-driven algorithm coupled prediction model are all online models, which utilize existing molten iron management data systems and server-collected data, or construct corresponding datasets to meet the data required by the prediction model.

[0026] Furthermore, the influence factor variables obtained by the infrared temperature sensor, ambient temperature sensor, rain and snow sensor, and photosensitive sensor are used to correct the boundary conditions, ambient temperature, shell temperature, thermal convection, and thermal radiation coefficients involved in the mechanism prediction model in real time, thereby obtaining more accurate information on the temperature drop of molten iron.

[0027] Furthermore, the influencing factor variables obtained by using infrared temperature sensors, ambient temperature sensors, rain and snow sensors, and photosensors are added to the input variables of the data-driven prediction model to correct the coefficients of the influencing factor variables in historical experience and mathematical statistical methods in real time, thereby obtaining high-precision information on the temperature drop of molten iron.

[0028] Furthermore, the influencing factor variables obtained by the ambient temperature sensor, rain and snow sensor and photosensitive sensor are introduced into the iron temperature drop prediction model that couples mechanism and data, and the boundary conditions and coefficients are corrected in real time to obtain high-precision iron temperature drop information.

[0029] In summary, the present invention has the following beneficial effects:

[0030] 1) Overcoming the shortcomings of existing methods in terms of factors affecting the temperature drop of molten iron, this method provides real-time changes in the temperature of the torpedo ladle / iron ladle body, external ambient temperature, rainfall and snowfall, and light intensity, reflecting the impact of changes in the torpedo ladle / iron ladle body and external conditions on the temperature drop of molten iron, especially reflecting changes such as thunderstorms, winter / summer, and day / night cycles.

[0031] 2) Overcoming the shortcomings of existing systems in real-time verification, a system and method for real-time verification of molten iron temperature drop prediction models are provided, which improves the accuracy of molten iron temperature drop prediction models and provides strong support for improving the operating efficiency and energy conservation and carbon reduction of steel interfaces.

[0032] 3) The improved accuracy of the iron-steel interface molten iron temperature drop prediction model is conducive to the dynamic and more precise adjustment of the molten iron ratio in steelmaking converters, thereby improving the production efficiency of steelmaking plants. At the same time, it creates favorable conditions for adding more scrap steel and realizes the intelligentization and low-carbonization of converters. Attached Figure Description

[0033] Figure 1 This is a schematic diagram of the iron-steel interface molten iron temperature drop prediction system described in this invention.

[0034] Figure 2 This is a schematic diagram of the iron-steel interface molten iron temperature drop prediction system described in this invention.

[0035] Figure 3 This is a flowchart of the iron-steel interface molten iron temperature drop prediction system described in this invention. Detailed Implementation

[0036] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below with reference to the figures and specific embodiments.

[0037] like Figure 1 and Figure 2As shown, the present invention proposes a real-time correction system for predicting the temperature drop of molten iron at the steel-iron interface, comprising a torpedo ladle / ladle history data unit 1, a real-time correction data unit 2, and a molten iron temperature drop prediction model unit 3. First, the torpedo ladle / ladle history information is obtained from the data system or data mart of the torpedo ladle / ladle history data unit 1. Then, the torpedo ladle / ladle real-time correction data unit 2 is established. The torpedo ladle / ladle history information and the real-time correction data are imported into the molten iron temperature drop prediction model 3, ultimately establishing a real-time correction system 4 for the molten iron temperature drop prediction model at the steel-iron interface.

[0038] The resume data unit 1 includes an existing data system 101 and a corresponding data mart 102 built according to requirements.

[0039] The real-time correction data unit 2 includes temperature data 201 of the torpedo canister / iron bag body, ambient temperature data 202, rain and snow data 203, and light intensity data 204.

[0040] The real-time correction data unit 2 includes an infrared temperature / thermocouple sensor 2011, a temperature acquisition unit 2012, and a main control unit 2013 installed on the torpedo can / ladle body, transmitting temperature data to the real-time correction data unit; an ambient temperature sensor 2021, a temperature acquisition unit 2022, and a main control unit 2023 installed near the torpedo can / ladle operating area, transmitting ambient temperature data to the real-time correction data unit 2; and a rain / snow volume sensor 2031 and a light sensor 2041, a rain / snow volume acquisition unit 2032 and a light intensity acquisition unit 2042, and main control units 2033 and 2043 installed near the torpedo can / ladle operating area, transmitting rain / snow volume and light intensity data to the real-time correction data unit.

[0041] The iron temperature drop prediction model unit 3 includes a mechanism model 301, an intelligent algorithm model 302, and a mechanism and intelligent algorithm coupled model 303.

[0042] See Figure 3 The present invention provides a method for predicting the temperature drop of molten iron at the iron-steel interface with real-time correction, comprising the following steps:

[0043] 1) Set up infrared temperature / thermocouple sensors, ambient temperature sensors, rain and snow sensors and photosensitive sensors to acquire variables S1 in real time, such as the temperature of the torpedo can / iron ladle shell or refractory material at the iron-steel interface, the temperature of the external environment, the amount of rain and snow and the light intensity.

[0044] 2) Based on the real-time process data S2 of the iron-steel interface, the uncorrected molten iron temperature drop information is obtained by using models S3 such as the iron-steel interface molten iron temperature drop mechanism prediction, the molten iron temperature drop prediction of data-driven intelligent algorithms, or the molten iron temperature drop prediction of mechanism coupled with data-driven methods.

[0045] 3) The real-time data S1 of the torpedo ladle / iron ladle body temperature, ambient temperature, rain and snow amount and light intensity are embedded into the molten iron temperature drop mechanism prediction model, the data-driven molten iron temperature drop prediction model or the molten iron temperature drop prediction model S3 coupled with mechanism and data driving, and the molten iron temperature drop prediction model is corrected in real time to obtain higher precision molten iron temperature drop information S5.

[0046] The aforementioned real-time correction method for predicting molten iron temperature drop at the iron-steel interface includes: real-time data S1 of the shell or refractory material temperature, ambient temperature, rain and snow amount, and light intensity of the torpedo ladle or ladle during operation, obtained by infrared temperature sensors, ambient temperature sensors, rain and snow sensors, and photosensitive sensors, is imported into a real-time correction data unit. This data serves as the influencing factor required for the molten iron temperature drop mechanism prediction model, the molten iron temperature drop prediction model of a data-driven intelligent algorithm, or the prediction model that couples the mechanism with a data-driven intelligent algorithm. The coefficients of the influencing factors are corrected in real time according to the shell temperature or refractory material temperature of the torpedo ladle / ladle and environmental changes, thereby obtaining high-precision molten iron temperature drop information S5.

[0047] The real-time correction method for predicting the temperature drop of molten iron at the iron-steel interface includes: using the influencing factor variable S1 obtained by infrared temperature / thermocouple sensor, ambient temperature sensor, rain and snow sensor and photosensitive sensor to correct the boundary condition ambient temperature, shell temperature, thermal convection and thermal radiation coefficient involved in the molten iron temperature drop mechanism prediction model S301 in real time, so as to obtain high-precision molten iron temperature drop information S5.

[0048] The real-time correction method for predicting the temperature drop of molten iron at the iron-steel interface includes: using the influencing factor variable S1 obtained by infrared temperature / thermocouple sensor, ambient temperature sensor, rain and snow sensor and photosensitive sensor, adding it to the input variable of the molten iron temperature drop prediction model S302 of data-driven intelligent algorithm, and correcting the coefficients of the influencing factor variable in historical experience and mathematical statistical methods in real time to obtain high-precision molten iron temperature drop information S5.

[0049] The real-time correction method for predicting the temperature drop of molten iron at the iron-steel interface includes: using the influencing factor variables S1 obtained by infrared temperature / thermocouple sensors, ambient temperature sensors, rain and snow sensors and photosensitive sensors, introducing the above influencing factor variables into the coupled model S3 of molten iron temperature drop prediction based on the molten iron temperature drop mechanism and data-driven intelligent algorithm, and correcting the boundary conditions and coefficients in real time to obtain high-precision molten iron temperature drop information S5.

[0050] In this document, the terms "upper," "lower," "front," "back," "left," "right," "top," "bottom," "inner," "outer," "vertical," and "horizontal," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only used for the clarity of expressing the technical solution and for the convenience of description, and therefore should not be construed as limiting the present invention.

[0051] In this document, the terms “comprising,” “including,” or any other variations thereof are intended to cover non-exclusive inclusion, which includes not only the elements listed but also other elements not expressly listed.

[0052] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A real-time corrected system for predicting the temperature drop of molten iron at the iron-steel interface, characterized in that, include A history data unit is used to construct a history database for torpedo cans or ladles with steel-to-iron interfaces. The real-time correction data unit acquires external environmental data in real time through infrared temperature sensors or pre-embedded thermocouples, ambient temperature sensors, rain and snow sensors, and photosensors. The hot metal temperature drop prediction model unit adopts a prediction model based on mechanism, a prediction model driven by data, or a prediction model coupled with mechanism and data driven by data. It uses external environmental data to correct the prediction accuracy of the prediction model in real time, thereby obtaining high-precision hot metal temperature drop information.

2. The real-time corrected iron-steel interface molten iron temperature drop prediction system according to claim 1, characterized in that, The history data unit is used to provide history data on the changes in influencing factors required by the molten iron temperature drop prediction model unit. The history data includes the history of the torpedo ladle or iron ladle, including real-time scheduling and process planning, real-time operation process, small, medium and large cycle maintenance and baking data.

3. The real-time corrected iron-steel interface molten iron temperature drop prediction system according to claim 1, characterized in that, The infrared sensor or thermocouple of the real-time correction data unit is used to acquire temperature data in real time during the operation of the torpedo canister or iron ladle; the ambient temperature sensor, rain and snow sensor and photosensitive sensor are used to acquire the temperature, amount of rain and snow and light intensity of the environment in which the torpedo canister or iron ladle is located in real time.

4. The real-time corrected iron-steel interface molten iron temperature drop prediction system according to claim 1, characterized in that, The mechanistic models include the temperature drop mechanism of the blast furnace tapping trough, the temperature drop mechanism of the torpedo ladle or iron ladle receiving iron, the temperature drop mechanism of three-stage or single-stage desulfurization, the temperature drop mechanism of the torpedo ladle or iron ladle adding iron, and the temperature drop mechanism of the torpedo ladle or iron ladle adding scrap steel or metal oxides; the data-driven algorithm models include statistical machine learning algorithms and deep learning algorithms.

5. A method for predicting the temperature drop of molten iron at the iron-steel interface with real-time correction, characterized in that, Includes the following steps: 1) The outer shell temperature of the torpedo canister or iron ladle at the iron-steel interface, the ambient temperature, the amount of rain and snow, and the light intensity are obtained through infrared temperature sensors, ambient temperature sensors, rain and snow sensors, and photosensors. 2) Based on the real-time process data of the iron-steel interface, the molten iron temperature drop at the iron-steel interface is predicted by the mechanism prediction model, the data-driven algorithm mechanism prediction model, or the mechanism and data-driven coupled molten iron temperature drop prediction model, so as to obtain the uncorrected molten iron temperature drop information. 3) The outer shell temperature of the torpedo ladle or iron ladle, the ambient temperature, the amount of rain and snow and the light intensity are embedded in the mechanism prediction model, the data-driven algorithm mechanism prediction model or the mechanism and data-driven coupled molten iron temperature drop prediction model in real time. The influence factor coefficients of the molten iron temperature drop model are corrected in real time using the outer shell temperature, the ambient temperature, the amount of rain and snow and the light intensity to obtain high-precision molten iron temperature drop information.

6. The method for predicting the temperature drop of molten iron at the iron-steel interface with real-time correction according to claim 4, characterized in that, The infrared temperature sensor is arranged on both sides of the running track or path of the torpedo canister or iron ladle to obtain the outer shell temperature of the torpedo canister or iron ladle. The ambient temperature sensor, rain and snow sensor, and light sensor are arranged in the area near the torpedo canister or ladle to obtain the ambient temperature, rain and snow amount, and light intensity in a timely manner; if the ambient temperature is obtained from the network system, the ambient temperature sensor is removed.

7. The method for predicting the temperature drop of molten iron at the iron-steel interface with real-time correction according to claim 4, characterized in that, The aforementioned mechanism prediction model, data-driven algorithm prediction model, or mechanism-data-driven algorithm coupled prediction model are all online models. They utilize existing molten iron management data systems and server-collected data, or construct corresponding datasets to meet the data requirements of the prediction model.

8. The method for predicting the temperature drop of molten iron at the iron-steel interface with real-time correction according to claim 4, characterized in that, The influence factor variables obtained by infrared temperature sensors, ambient temperature sensors, rain and snow sensors, and photosensitive sensors are used to correct the boundary conditions, ambient temperature, shell temperature, thermal convection, and thermal radiation coefficients involved in the mechanism prediction model in real time, thereby obtaining more accurate information on the temperature drop of molten iron.

9. The method for predicting the temperature drop of molten iron at the iron-steel interface with real-time correction according to claim 4, characterized in that, The influencing factor variables obtained by using infrared temperature sensors, ambient temperature sensors, rain and snow sensors, and photosensors are added to the input variables of the data-driven prediction model. The coefficients of the influencing factor variables in historical experience and mathematical statistical methods are corrected in real time to obtain high-precision information on the temperature drop of molten iron.

10. The method for predicting the temperature drop of molten iron at the iron-steel interface with real-time correction according to claim 4, characterized in that, The influencing factor variables obtained by the ambient temperature sensor, rain and snow sensor and photosensitive sensor are introduced into the iron temperature drop prediction model that couples mechanism and data, and the boundary conditions and coefficients are corrected in real time to obtain high-precision iron temperature drop information.

Images