Crystallizer heat flux inverse calculation method, system and medium based on optical fiber temperature measurement

Abstract

The application discloses a crystallizer heat flux inverse calculation method and system based on optical fiber temperature measurement, a medium and equipment, which realizes real-time acquisition of temperature measurement data through optical fiber sensors arranged in a slab continuous casting crystallizer of a steel plant, and combines with the running process conditions (steel composition and casting speed). First, the on-site temperature measurement data is collected and stored, then the noise data is processed by using a smoothing exponential algorithm according to the data characteristics, then a stable data source is selected for big data empirical formula fitting to determine the empirical formula parameters, thereby providing a reference for setting the initial value of iterative calculation and serving as a way to test the accuracy of model calculation in the future. According to the arrangement characteristics of the optical fiber temperature measurement sensor, heat transfer calculation is carried out by using a longitudinal modeling method. In addition, a man-machine interface based on a Web front-end visualization technology is further developed to more intuitively and effectively monitor the temperature of the crystallizer part. The method solves the problems of poor accuracy and sensitivity of the previous temperature measurement method relative to optical fiber temperature measurement, the data noise processing is adopted for the large amount of data of optical fiber temperature measurement, and the direct inverse calculation method of the longitudinal heat flux of the crystallizer is adopted, so that the calculation is simple and the result is accurate.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of crystallizers, in particular to a crystallizer heat flux inverse calculation method, system and medium based on optical fiber temperature measurement. BACKGROUND

[0002] The crystallizer is called the "heart" of continuous casting, and is the starting point for controlling the purity of continuous casting billets and ensuring the quality of the cast billets. Many surface defects found in the final rolled products can usually be traced back to the heat transfer in the continuous casting crystallizer during the early stage of solidification of the molten steel. According to statistics, 80% of the surface defects of continuous casting billets are derived from the crystallizer. The crystallizer heat flow, especially in the meniscus region of the crystallizer, is highly complex due to the penetration of the lubricant liquid protective slag, the strong fluid flow and the transient nature of the crystallizer vibration. This complexity poses a major challenge to clearly and comprehensively understand all the dynamics within the system. Fluctuations in the crystallizer heat flux reflect the solidification process within the crystallizer, and accurate measurement of the crystallizer heat flux is crucial to ensure the quality of the continuous casting billets. Temperature measurement is usually used to monitor the heat flux, which is very important for improving the cooling system, casting practice and process control. For example, too low a heat extraction rate will cause cracking, and too high a heat extraction rate will cause longitudinal cracks. In order to measure the heat flux distribution on the entire mold surface, a thermocouple array is usually embedded in the mold wall. Although this method is accurate and easy to use, it can be invasive and can affect the thermal behavior of the mold.

[0003] With the continuous development and application of optical fiber temperature measurement sensors in recent years, optical fibers have become a promising technology for measuring crystallizer heat flow, providing high precision, fast response and electromagnetic interference resistance. For example, Chinese patent CN109960835A discloses a method for establishing a continuous casting crystallizer heat flow distribution model, which determines the heat flux distribution by simply interpolating the heat flux of several points, which has poor accuracy and cannot process large data. Chinese patent CN101941060A discloses a crystallizer thermal image real-time display method based on multiple rows of measured thermocouple temperatures, which calculates the temperature values at all positions on the same longitudinal section of the crystallizer using the measured thermocouple temperature values by interpolation. Only the temperature is obtained by interpolation, without accurate modeling and calculation, and the amount of measured data is also small. SUMMARY

[0004] Therefore, the present application provides a crystallizer heat flux inverse calculation method, system, medium and equipment based on optical fiber temperature measurement, which solves the problem of poor accuracy of the existing temperature measurement method.

[0005] According to one aspect of the present application, a crystallizer heat flux inverse calculation method based on optical fiber temperature measurement is provided, comprising:

[0006] Step 1: disposing a fiber temperature sensor at a narrow copper plate part of a mold on a continuous casting billet production line, and connecting the fiber temperature sensor and a fiber grating data demodulator;

[0007] Step 2: collecting a measured temperature of the mold in real time by using the fiber temperature sensor on the continuous casting billet production line;

[0008] Step 3: performing noise reduction processing on the measured temperature, and performing empirical formula fitting on the measured temperature after the noise reduction processing, wherein the empirical formula is used to roughly calculate heat flux distribution and set an iteration initial value of the inverse problem model;

[0009] Step 4: establishing a forward problem model of mold heat transfer according to a size of the mold, wherein the forward problem model is used to fit longitudinal distribution of mold heat flux, and a boundary condition of the forward problem model is a result of each heat flux iteration inverse calculation;

[0010] Step 5: constructing an inverse problem model of mold heat transfer according to a known heat transfer condition of the mold and the forward problem model, wherein the inverse problem model is used to calculate heat flux according to the measured temperature;

[0011] Step 6: bringing a preset initial heat flux into the forward problem model, if a termination condition is not met, continuing to iterate the inverse problem model, and bringing heat flux obtained by solving the inverse problem model into the forward problem model, re-determining whether the termination condition is met, until the termination condition is met, and determining that the heat flux longitudinal distribution obtained by the forward problem model in this iteration is a target distribution;

[0012] Step 7: analyzing the target distribution, and visually displaying the target distribution.

[0013] Optionally, the noise reduction processing on the measured temperature comprises:

[0014] using the fiber grating data demodulator to convert the measured temperature into a digital format, and storing the converted measured temperature in a database;

[0015] determining a smoothing coefficient, and using an exponential smoothing method to perform noise reduction processing on the stored measured temperature.

[0016] Optionally, the establishment of the forward problem model of mold heat transfer according to the size of the mold comprises:

[0017] determining a production line parameter, wherein the production line parameter at least comprises a continuous casting machine size parameter on the continuous casting production line;

[0018] According to the production line parameters and the arrangement position of the optical fiber temperature measurement sensor in the crystallizer, a forward problem model of heat transfer of the crystallizer is constructed based on a longitudinal section of the crystallizer.

[0019] Optionally, the termination condition is determined based on an error between the measured temperature and a calculated temperature obtained in this iteration.

[0020] Optionally, the empirical formula fitting of the temperature data comprises:

[0021] An empirical formula of longitudinal heat flux distribution of the crystallizer is obtained, and parameters in the empirical formula are determined by fitting the measured temperature based on the empirical formula.

[0022] Optionally, the real-time visualization of the heat flux calculation result comprises:

[0023] A calculated temperature curve and a measured temperature curve are drawn, and accuracy of the forward problem model and the inverse problem model is determined according to a comparison result between the calculated temperature and the measured temperature.

[0024] A heat flux distribution graph corresponding to the target distribution is displayed on a visual interface by using a Web front-end visualization technology.

[0025] The calculated temperature is subjected to data interpolation processing, and a temperature curve graph and a temperature cloud graph are drawn based on the interpolated calculated temperature by using a Web front-end visualization technology.

[0026] Optionally, the forward problem model comprises a heat transfer control equation of the crystallizer, a heat flux boundary condition between an inner side of a narrow copper plate of the crystallizer and a cast blank, and a forced convection boundary condition between an outer side of the narrow copper plate of the crystallizer and cooling water.

[0027] According to another aspect of the present application, a crystallizer heat flux inverse calculation system based on optical fiber temperature measurement is provided, comprising an interface module, a model module and a controller module.

[0028] The controller module is configured to control the model module to perform the foregoing method, and specifically comprises: collecting, on the continuous casting billet production line, the measured temperature of the mold in real time by using the fiber temperature sensor; performing noise reduction processing on the measured temperature, and performing empirical formula fitting on the measured temperature after the noise reduction processing, wherein the empirical formula is used to roughly calculate the heat flux distribution and set the iteration initial value of the inverse problem model; establishing a forward problem model of mold heat transfer according to the size of the mold, wherein the forward problem model is used to fit the longitudinal distribution of the mold heat flux; constructing an inverse problem model of mold heat transfer according to the known heat transfer condition of the mold and the forward problem model, wherein the inverse problem model is used to calculate the heat flux according to the measured temperature; bringing the preset initial heat flux into the forward problem model, if the termination condition is not met, continuing to iterate the inverse problem model, and bringing the heat flux obtained by solving the inverse problem model into the forward problem model, re-determining whether the termination condition is met, until the termination condition is met, and the heat flux longitudinal distribution obtained by the forward problem model in this iteration is determined as a target distribution; analyzing the target distribution.

[0029] The interface module is configured to visually display the target distribution obtained by the model module.

[0030] Optionally, the controller module is configured to:

[0031] control the model module to convert the measured temperature into a digital format by using the fiber grating data demodulator, and store the converted measured temperature in a database; and determine a smoothing coefficient, and perform noise reduction processing on the stored measured temperature by using an exponential smoothing method.

[0032] Optionally, the controller module is configured to:

[0033] control the model module to determine a production line parameter, wherein the production line parameter at least includes a size parameter of a continuous casting machine on the continuous casting production line; and construct a forward problem model of mold heat transfer based on a longitudinal section of the mold according to the production line parameter and the arrangement position of the fiber temperature sensor in the mold.

[0034] Optionally, the termination condition is determined based on an error between the measured temperature and a calculated temperature obtained in this iteration.

[0035] Optionally, the controller module is configured to:

[0036] control the model module to obtain an empirical formula of the longitudinal heat flux distribution of the mold, and fit the measured temperature based on the empirical formula to determine parameters in the empirical formula.

[0037] Optionally, the interface module is configured to:

[0038] drawing a calculation temperature curve and a measurement temperature curve, determining the accuracy of the forward problem model and the inverse problem model according to a comparison result between the calculation temperature and the measurement temperature;

[0039] displaying a heat flux distribution diagram corresponding to the target distribution on a visual interface by using a Web front-end visualization technology;

[0040] performing data interpolation processing on the calculation temperature, and drawing a temperature curve diagram and a temperature cloud diagram based on the interpolated calculation temperature by using a Web front-end visualization technology.

[0041] Optionally, the forward problem model comprises a crystallizer heat transfer control equation, a heat flux boundary condition between an inner side of a narrow copper plate of the crystallizer and a cast blank, and a forced convection boundary condition between an outer side of the narrow copper plate of the crystallizer and cooling water.

[0042] According to still another aspect of the present application, a medium having a program or instruction stored thereon is provided, the program or instruction being executed by a processor to implement the above-mentioned crystallizer heat flux inverse calculation method based on fiber temperature measurement.

[0043] According to still another aspect of the present application, a device is provided, comprising a storage medium and a processor, the storage medium storing a computer program, and the processor executing the computer program to implement the above-mentioned crystallizer heat flux inverse calculation method based on fiber temperature measurement.

[0044] By means of the above technical solution, the present application fully takes advantage of the characteristics of the fiber temperature measurement sensor, replaces the relatively rough problem of the traditional thermocouple temperature measurement with a more precise and detailed layout and measurement result, and uses a big data algorithm to first determine the undetermined parameters of the heat flux empirical formula as a basis for determining the initial condition and a reliability test of the model calculation result. Then, according to the position characteristics of the fiber temperature measurement sensor layout, the traditional cross-section calculation model is abandoned, and a longitudinal fitting method for obtaining the longitudinal heat flux distribution is used. Instead, a longitudinal model is directly used to calculate the model based on the temperature measurement data of the longitudinal fiber temperature measurement sensor as a reference point, which directly obtains the longitudinal heat flux distribution of the crystallizer, and is more accurate and efficient. Finally, the fiber temperature measurement data is visualized by using a Web front-end visualization human-computer interaction interface language, and the crystallizer heat flux inverse calculation method and visualization are finally realized.

[0045] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the specific embodiments of the present application can be implemented in accordance with the content of the description, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0046] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:

[0047] Figure 1 A flowchart of a crystallizer heat flux inverse calculation method based on fiber temperature measurement provided by an embodiment of the application is shown;

[0048] Figure 2 A distribution position of a fiber temperature measurement sensor on a crystallizer and a heat flux distribution diagram provided by an embodiment of the application are shown;

[0049] Figure 3 A fiber string diagram provided by an embodiment of the application is shown;

[0050] Figure 4 A measured temperature data diagram provided by an embodiment of the application is shown;

[0051] Figure 5 A model grid division diagram provided by an embodiment of the application is shown;

[0052] Figure 6 A comparison diagram of a calculated temperature output and measured temperature data of a fiber temperature measurement sensor provided by an embodiment of the application is shown;

[0053] Figure 7 A crystallizer longitudinal heat flux distribution diagram provided by an embodiment of the application is shown;

[0054] Figure 8 A display flowchart of a crystallizer heat flux inverse calculation system based on fiber temperature measurement provided by an embodiment of the application is shown;

[0055] Figure 9 A structure block diagram of a crystallizer heat flux inverse calculation system based on fiber temperature measurement provided by an embodiment of the application is shown. DETAILED DESCRIPTION

[0056] The application will be described in detail below with reference to the drawings and in conjunction with embodiments. It should be noted that the embodiments in the application and the features in the embodiments can be combined with each other without conflict.

[0057] In the embodiment, a crystallizer heat flux inverse calculation method based on fiber temperature measurement is provided, specifically a crystallizer heat flux inverse calculation method based on fiber temperature measurement, as shown in the figure, the method comprises: Figure 1

[0058] ​Step 1: The optical fiber temperature sensor is arranged at the narrow face copper plate part of the crystallizer on the continuous casting billet production line, and the optical fiber temperature sensor and the optical fiber grating data demodulator are connected.

[0059] In this step, first, the optical fiber temperature sensor is punched into the appropriate part of the crystallizer of the continuous casting machine of the continuous casting production line. Specifically, as shown in FIG. 1, the longitudinal hole is punched in the area 210-410 mm away from the upper part of the 900 mm high crystallizer. The optical fiber temperature sensor is more densely arranged than the traditional crystallizer thermocouple temperature measurement, and the temperature measurement data is more accurate. Then determine the optical fiber temperature measurement area and the distance between each optical fiber temperature measurement point, so as to facilitate the comparison of the measured temperature of each optical fiber temperature measurement point with the calculation result in subsequent model calculation, and analyze the accuracy of the model calculation. At the same time, connect the optical fiber grating data demodulator to facilitate the format conversion of the temperature data by the demodulator. Figure 2

[0060] Step 2: Real-time acquisition of the measured temperature of the crystallizer on the continuous casting billet production line by using the optical fiber temperature sensor.

[0061] In this step, data acquisition is performed on the production line. Specifically, the optical fiber temperature sensor is used to collect the temperature data of the continuous casting production site slab continuous casting crystallizer in real time, and the corresponding production process data and continuous casting crystallizer size are collected. Among them, the production process data includes the drawing speed, the steel composition, the inlet and outlet water temperature of the crystallizer copper plate; the principle of the optical fiber temperature sensor is to use the principle that the spectrum absorbed by part of the substance changes with temperature, and analyze the spectrum transmitted by the optical fiber to understand the real-time temperature.

[0062] Among them, for the sampling rate, if the temperature sampling rate is too low, the temperature data may not be able to capture the full range of temperature changes in the mold. This is because high-frequency temperature changes are incorrectly represented as low-frequency components in the data, leading to inaccurate heat flux estimates. Therefore, when estimating heat flow from observed temperature, the method of inverse heat conduction problem must be used. However, a high sampling rate can increase the computational complexity of inverting heat flux density. In addition, due to sensor measurement error, a high sampling rate can introduce a lot of noise in the temperature data, leading to a decrease in the accuracy of heat flux estimation. Increasing the temperature sampling rate can reduce the accuracy of estimating heat flow through inverse problems. Therefore, it is crucial to study the effect of temperature sampling rate on the accuracy of crystallizer heat flow inversion.

[0063] ​The time step is a key factor in the inverse heat conduction problem. The sampling rate affects the size of the sensitivity coefficients by affecting the time step used to solve the sensitivity coefficient problem. This in turn affects the stability of the algorithm to solve the inverse problem. The instability of the inverse problem refers to the high sensitivity to small changes in the input data noise, which can lead to substantial errors in the output. It can then be inferred that if the temperature is measured too slowly, the heat transfer process can have changed. This difference between the observed temperature and the true temperature can lead to inaccurate measurements.

[0064] In short, it can be observed that when the temperature sampling rate is constantly rising, the accuracy of the heat flux first rises, then stabilizes, and then falls to a relatively stable value. In addition, when using a high sampling rate, the first and second order spatial regularization is more accurate than the zero order spatial regularization. However, increasing the temperature sampling rate significantly increases the CPU time required. The order of spatial regularization has no significant effect on CPU time.

[0065] Step 3: denoising the measured temperature, and fitting the denoised measured temperature with an empirical formula, wherein the empirical formula is used to roughly calculate the heat flux distribution and set the initial value of iteration of the inverse problem model.

[0066] In this step, the temperature data of the crystallizer copper plate collected by the optical fiber temperature sensor and the related process parameters during the continuous casting process are classified and organized, and the temperature measurement point data of the optical fiber temperature sensor are preprocessed for big data noise. The data after preprocessing are fitted with an empirical formula to obtain an empirical formula curve that can roughly reflect the change characteristics of the longitudinal heat flux of the crystallizer. The better the data fitting degree of the empirical formula, the more beneficial it is to the subsequent model calculation efficiency.

[0067] Step 4: establishing a forward problem model of the crystallizer heat transfer according to the size of the crystallizer, wherein the forward problem model is used to fit the longitudinal distribution of the heat flux of the crystallizer, and the boundary condition of the forward problem model is the result of each heat flux iteration inverse calculation.

[0068] In this step, a forward problem model for calculating the heat transfer of the crystallizer copper plate is established according to the size of the crystallizer in the on-site continuous casting production line. Specifically, the model is simplified and assumed as follows:

[0069] (1) The heat flux at the same height of the narrow face of the crystallizer does not change much, so the heat flux on the cross section at the same height is considered equal;

[0070] (2) The copper plate of the crystallizer is subjected to forced convection heat transfer with cooling water. Since the cooling water tank is distributed on the back of the copper plate, it is considered that there is a uniform cooling intensity at the copper plate;

[0071] (3) The cooling water temperature at the same height is considered equal;

[0072] (4) In the stable casting conditions, the observed temperature change at the same position is very small, which can be considered as a steady-state heat conduction process;

[0073] (5) The influence of the bolt holes and the fiber installation holes inside the crystallizer on the temperature distribution is ignored;

[0074] (6) The change of the thermal physical parameters of the copper plate with temperature is ignored, and it is considered as a constant;

[0075] (7) The narrow copper plate exchanges heat with the cast slab on the inner surface and with the cooling water on the outer surface, and the rest is considered as an adiabatic condition.

[0076] According to the arrangement characteristics of the fiber temperature measurement sensor in the crystallizer, the two-dimensional calculation model constructed by the transverse section used in the past is abandoned, and the model is directly established on the longitudinal section, so that the longitudinal heat flux distribution of the crystallizer can be directly fitted.

[0077] Among them, the positive problem model includes the crystallizer heat transfer control equation, the heat flux boundary condition between the inner side of the narrow copper plate of the crystallizer and the cast slab, and the forced convection boundary condition between the outer side of the narrow copper plate of the crystallizer and the cooling water.

[0078] Specifically, the research area of the narrow plate of the crystallizer is two-dimensional steady-state heat conduction, and the crystallizer heat transfer control equation is as follows:

[0079]

[0080] The boundary conditions in the research area are as follows:

[0081] (1) The heat flux boundary condition between the inner side of the narrow copper plate of the crystallizer and the cast slab:

[0082]

[0083] In the formula, k is the thermal conductivity of copper; q is the heat flux between the copper plate and the cast slab.

[0084] (2) The forced convection boundary condition between the outer side of the narrow copper plate of the crystallizer and the cooling water:

[0085]

[0086] In the formula, T k is the temperature of the copper plate of the crystallizer, T w is the water temperature of the cooling water at the corresponding height (i.e. the inlet and outlet water temperature); h w is the forced convection heat transfer coefficient, which can be calculated by the following formula:

[0087]

[0088] In which D wD is the equivalent diameter of the cooling water tank; λ w k is the thermal conductivity of the cooling water; v w v is the flow velocity of the cooling water; μ w μ is the viscosity of the cooling water; C w C is the specific heat capacity of the cooling water; ρ w ρ is the density of the cooling water.

[0089] Step 5: According to the known heat transfer conditions of the crystallizer and the forward problem model, a reverse problem model of the heat transfer of the crystallizer is constructed, wherein the reverse problem model is used to calculate the heat flux according to the measured temperature.

[0090] In this step, according to the known conditions and the established forward problem model, a reverse problem model of the heat transfer calculation of the crystallizer is constructed. Specifically, the reverse problem model is used to calculate the heat flux from the observed temperature, which involves finding the heat flux that minimizes the difference between the calculated temperature T and the observed temperature Y. The model is as follows:

[0091]

[0092]

[0093] T = Tt

[0094] T int <T k <T out

[0095] where j is the time step; T is the temperature measurement of the optical fiber temperature sensor; T k T is the cooling water temperature; T int is the inlet water temperature; T out is the outlet water temperature; R is the regularization term; after considering the accuracy and calculation efficiency, it is determined to use the first order regularization term, which is as follows:

[0096]

[0097] where n1 and n2 represent the two boundary conditions in the aforementioned step 4, respectively.

[0098] The Jacobian matrix of the calculation iteration is as follows:

[0099]

[0100] And then the regularization term is adjusted continuously through the calculation of the Jacobian matrix to continuously optimize the objective function, and finally the optimization purpose is achieved.

[0101] Step 6: The preset initial heat flux is brought into the forward problem model, if the termination condition is not met, the inverse problem model is iterated, and the heat flux obtained by solving the inverse problem model is brought into the forward problem model, and it is judged whether the termination condition is met, until the termination condition is met, and the heat flux longitudinal distribution obtained by the forward problem model in this iteration is determined as the target distribution.

[0102] In this step, a calculation program is written, and the initial heat flux value is preset based on the empirical formula in the foregoing step, and then optimization iteration is performed. Specifically, the crystallizer heat transfer control equation in the foregoing step 4 is integrated:

[0103]

[0104] Then the control equation is discretized to obtain:

[0105]

[0106] Wherein, the termination condition is judged based on the error between the measured temperature and the calculated temperature obtained in this iteration, specifically:

[0107] ∑‖Y i -T i ‖ 2 ≤ε

[0108] Then the discretized equation is programmed and calculated, the heat flux q obtained by solving the inverse problem is brought into the forward problem, if the termination limit is not met, the optimization function is used for optimization iteration to achieve the iteration termination limit to obtain the inverse calculation of the crystallizer heat flux longitudinal distribution.

[0109] Step 7: Analyze the target distribution and visually display the target distribution.

[0110] In this step, the analysis calculation result is displayed in real time. Specifically, through the Web front-end visualization technology, the interface, model and controller are modularly processed by using the MVC programming architecture. The interface part is responsible for display and interaction with the user, etc. The controller is used to determine which model needs to be used to process the request from the view and which view needs to be returned after processing. That is, to connect the view and the model. The model holds all the data, state and program logic. The model accepts the request of the view data and returns the final processing result. The interface of the visualization human-computer interaction system is constructed by using the HTML language, the interface is designed by using the CSS, the visualization display function is realized by using the JavaScript language, and the Java language is used as the control part to contact the data and the human-computer interaction interface, and the SQL database language is used to read the temperature measurement data of the optical fiber sensor. Then the whole crystallizer heat flux inverse calculation method and visualization system based on optical fiber temperature measurement can normally run.

[0111] This embodiment realizes the calculation and display of heat flux distribution from the installation of optical fiber temperature sensor, data acquisition and preprocessing, heat flux model calculation, calculation result data processing and result visualization. Specifically, this embodiment fully utilizes the characteristics of the optical fiber temperature sensor to replace the relatively rough problem of traditional thermocouple temperature measurement with more precise and detailed layout and measurement results, and uses big data algorithm to first determine the undetermined parameters of the heat flux empirical formula as the basis for determining the initial conditions and the reliability test of the model calculation result. Then, according to the position characteristics of the optical fiber temperature sensor layout, the traditional cross-section calculation model is abandoned, and the heat flux longitudinal distribution is obtained by longitudinal fitting. Instead, the model is calculated directly through the longitudinal model according to the temperature measurement data of the longitudinal optical fiber temperature sensor as the reference point, which directly obtains the longitudinal heat flux distribution of the crystallizer, and is more accurate and efficient. Finally, the optical fiber temperature data is visualized by using the Web front-end visualization human-computer interaction interface language, and the crystallizer heat flux inverse calculation method and visualization are finally realized.

[0112] Further, as a refinement and extension of the above embodiment, in order to completely describe the specific implementation process of the embodiment, another crystallizer heat flux inverse calculation method based on optical fiber temperature measurement is provided, in which the measured temperature is denoised, including the following steps:

[0113] Step 201, using the optical fiber grating data demodulator to convert the measured temperature into digital format, and storing the converted measured temperature in the database;

[0114] Step 202, determining the smoothing coefficient, and denoising the stored measured temperature by using the exponential smoothing method.

[0115] In this step, the temperature measurement data of the optical fiber temperature sensor is collected, which is converted into digital signals by using the fiber grating data demodulator, and the real-time data is stored in a specific database.

[0116] In addition, there may be noise data in the temperature measurement data of the optical fiber temperature sensor. In order to calculate more accurately, it is necessary to remove these noise data. Based on the requirements of the data and the characteristics of the temperature measurement data of the optical fiber temperature sensor, the exponential smoothing algorithm is adopted in this embodiment. The exponential smoothing algorithm is a smoothing method based on weighted average, which reduces the influence of noise by weighted average of data. Its basic principle is to give higher weight to the latest data points, so as to pay more attention to the latest trend.

[0117] The formula of exponential smoothing can be expressed as:

[0118] T(t) = aY(t) + (1-a)T(t-1)

[0119] Where T(t) represents the smoothed value at time t; Y(t) represents the observed value or original data at time t; T(t-1) represents the smoothed value at time t-1. a is the smoothing coefficient, which is in the range of 0 to 1, representing the weight of new data.

[0120] Further, as a refinement and extension of the above embodiment, in order to complete the description of the specific implementation process of this embodiment, another method of calculating the heat flux of the crystallizer based on the optical fiber temperature is provided, in which a positive problem model of the heat transfer of the crystallizer is established according to the size of the crystallizer, including the following steps:

[0121] Step 301, determining the production line parameters, wherein the production line parameters at least include the size parameters of the continuous casting machine on the continuous casting production line;

[0122] Step 302, constructing a positive problem model of the heat transfer of the crystallizer based on the longitudinal section of the crystallizer according to the production line parameters and the arrangement position of the optical fiber temperature sensor in the crystallizer.

[0123] In this step, the model parameters are determined according to the size parameters of the continuous casting machine, and then the positive problem model is constructed, wherein the model is constructed based on the longitudinal interface because the optical fiber temperature sensor is arranged in a longitudinal row.

[0124] Further, as a refinement and extension of the above embodiment, in order to complete the description of the specific implementation process of this embodiment, another method of calculating the heat flux of the crystallizer based on the optical fiber temperature is provided, in which the measured temperature is fitted by an empirical formula, including the following steps:

[0125] An empirical formula of the longitudinal heat flux distribution of the crystallizer is obtained, and parameters in the empirical formula are determined based on fitting of the measured temperature.

[0126] In this step, first, according to the empirical formula of the longitudinal heat flux distribution of the crystallizer, a large amount of temperature data measured by the optical fiber temperature sensor is analyzed by big data fitting to determine the parameters in the empirical formula, so as to roughly calculate the heat flux distribution and provide a reference for setting the initial value of the inverse problem iteration. The roughly calculated heat flux distribution can be used for comparison with the final output target distribution to judge the accuracy of the heat flux inverse calculation.

[0127] Further, as a refinement and expansion of the above embodiment, in order to fully describe the specific implementation process of the embodiment, another optical fiber temperature-based crystallizer heat flux inverse calculation method is provided, in which the heat flux calculation result is displayed in real time, including the following steps:

[0128] Step 401, draw the calculated temperature curve and the measured temperature curve, and determine the accuracy of the forward problem model and the inverse problem model according to the comparison result between the calculated temperature and the measured temperature;

[0129] Step 402, display the heat flux distribution graph corresponding to the target distribution on the visualization interface by using the Web front-end visualization technology;

[0130] Step 403, perform data interpolation processing on the calculated temperature, and draw the temperature curve graph and the temperature cloud graph based on the interpolated calculated temperature by using the Web front-end visualization technology.

[0131] In this step, the calculated temperature and the measured temperature are drawn respectively, and the data are compared to determine the model calculation accuracy intuitively, a heat flux calculation real-time display system is constructed, the calculated heat flux distribution is displayed by using the Web front-end visualization technology, and the data are processed by interpolation to draw a vector scaling graph, and the crystallizer temperature field visualization interface display is completed.

[0132] In addition, after the iteration is terminated and the target heat flux distribution is obtained, the calculated heat flux distribution can also be analyzed and verified. First, the matching degree of the actual temperature measurement data of the optical fiber temperature sensor and the calculated data is observed, and second, the calculated heat flux curve is compared with the heat flux curve in the empirical formula.

[0133] Further, in another embodiment, the specific steps include:

[0134] Step 101: analyze and determine the customized size and arrangement position of the optical fiber temperature sensor according to the design drawing of the continuous casting machine, and perform on-site installation and testing.

[0135] In this embodiment, the fiber string customization is as shown in Figure 3 The size of the crystallizer narrow copper plate is 35*260*900 mm, and the calculated longitudinal section is 35*900 mm.

[0136] The distribution position of the fiber temperature sensor on the crystallizer and the heat flux distribution are as shown in Figure 2 The corresponding measured temperature data are as shown in Figure 4 The steel composition is as shown in Table 1, and the physical property parameters are as shown in Table 2.

[0137] Table 1 Steel composition

[0138] Ingredients C Si Mn P S Content (wt%) 0.070 0.100 1.500 0.010 0.001

[0139] Table 2 Physical property parameters

[0140]

[0141] In addition, the production speed in this embodiment is 1.30 m / min.

[0142] Step 102: arranging a database on the server side, establishing a data table in the database according to the data use, and inputting the temperature data collected in step 101 into the data table;

[0143] In this embodiment, a local fiber temperature measurement database is built on Microsoft SQL Server 2019. The table design stores the temperature measurement data of 18 temperature measurement points of the fiber over time.

[0144] Step 103: using a big data fitting method, according to the requirements of data in specific application scenarios and the temperature measurement data characteristics of the fiber temperature measurement sensor, an exponential smoothing algorithm is adopted. The exponential smoothing algorithm is a smoothing method based on weighted average, which reduces the influence of noise by weighted average of data. Its basic principle is to give higher weight to the latest data points, so as to pay more attention to the latest trend.

[0145] The formula of exponential smoothing can be expressed as:

[0146] T(t) = aY(t) + (1-a)T(t-1)

[0147] Where T(t) represents the smoothed value at time t; Y(t) represents the observed value or original data at time t; T(t-1) represents the smoothed value at time t-1. a is the smoothing coefficient, taking value in the range of 0 to 1, representing the weight of new data.

[0148] Then, the heat flux empirical equation is obtained through continuous adjustment and analysis of the results. The parameters of the empirical formula are determined to provide better initial values for iterative calculation, and the final result is:

[0149]

[0150] where h is the longitudinal distance from the meniscus, and v is the pulling speed.

[0151] Step 104: According to the arrangement characteristics of the optical fiber crystallizer, the longitudinal cross-section is selected to construct the calculation model, and the model grid division is as shown in FIG. 4, wherein the heat transfer control equation is as follows: Figure 5

[0152]

[0153] Boundary conditions in the research area:

[0154] (1) The heat flux boundary condition between the inner side of the copper plate of the narrow surface of the crystallizer and the cast blank:

[0155]

[0156] where λ is the thermal conductivity of copper; q is the heat flux between the copper plate and the cast blank;

[0157] (2) The forced convection boundary condition between the outer side of the copper plate of the narrow surface of the crystallizer and the cooling water:

[0158]

[0159] where T k is the temperature of the copper plate of the crystallizer, T w is the water temperature of the cooling water at the corresponding height; h w is the forced convection heat transfer coefficient, which can be calculated by the following formula:

[0160]

[0161] where D w is the equivalent diameter of the cooling water tank; λ w is the thermal conductivity of the cooling water; v w is the flow rate of the cooling water; μ w is the viscosity of the cooling water; C w is the specific heat capacity of the cooling water; and ρ w is the density of the cooling water.

[0162] Step 105: On the basis of the heat transfer positive problem model, the heat transfer inverse problem of the copper plate of the narrow surface of the crystallizer is solved by using the interior point method, and the objective function is:

[0163]

[0164]

[0165] T = T(t)​

[0166] T int <T k <T out

[0167] Where j is the time step; T is the temperature measurement of the fiber optic temperature sensor; T k is the cooling water temperature; T int is the inlet water temperature; T out is the outlet water temperature; R is the regularization term; after considering the accuracy and calculation efficiency, the first order regularization term is determined, and the formula is as follows:

[0168]

[0169] Where n1 and n2 represent two kinds of boundary conditions respectively.

[0170] The Jacobian matrix of the calculation iteration is as follows:

[0171]

[0172] And then the regularization term is adjusted constantly through the calculation of the Jacobian matrix to constantly optimize the objective function, and finally the optimization purpose is achieved.

[0173] Step 106: Discretize the crystallizer heat transfer positive problem control equation, and write the iteration calculation program.

[0174] Integrate the control equation:

[0175]

[0176] And then the control equation is discretized to obtain:

[0177]

[0178] The termination limit of the program iteration is:

[0179] ∑||Y i -T i || 2 ≤ε

[0180] And then the programming calculation is carried out according to the discretized equation, the heat flux q obtained by solving the inverse problem is brought into the positive problem, if the termination limit is not met, the optimization function is continuously optimized to achieve the iteration termination limit, so as to obtain the inverse calculation of the longitudinal distribution of the crystallizer heat flux.

[0181] Step 107: Compare the calculated temperature output with the measured temperature data of the fiber optic temperature sensor, as shown in Figure 6 It can be seen that the data calculation result has good accuracy. And then the longitudinal heat flux distribution diagram of the crystallizer can be obtained, as shown inFigure 7 As shown.

[0182] Step 108: According to Figure 8 The illustrated process employs the MVC programming architecture, modularizing the interface, model, and controller. The interface handles display and user interaction, while the controller determines which model to use for requests from the view and which view to navigate back to after processing. It connects the view and the model. The model holds all data, state, and program logic. It receives data requests from the view and returns the final processing result. HTML is used to create the visual interface framework, populating the corresponding box models. CSS is used to design box sizes and styles. JavaScript is used to write the logic, and Java is used to connect the functions to the database, enabling interaction with the database when functions are triggered.

[0183] When the trigger function draws the data cloud map, a bilinear interpolation algorithm is used to achieve accurate and aesthetically pleasing drawing.

[0184] The interpolation logic used is as follows:

[0185] T P =T W +T E +T S +T N

[0186] Where T P Let T be the interpolation point temperature, where T is the interpolation point temperature. w T E T S T N They are determined by the following formulas respectively:

[0187] T W =T(row,col)×(1-Dis(x))×(1-Dis(y))

[0188] T E =T(row,col+1)×(1-Dis(x))×Dis(y)

[0189] T S =T(row+1,col)×Dis(x)×(1-Dis(y))

[0190] T N =T(row+1,col+1)×Dis(x)×Dis(y)

[0191] Where Dis(x) is the ratio of the data in the data table at the x position of the total data x direction size; Dis(y) is the ratio of the data in the data table at the y position of the total data y direction size:

[0192] The calculation formula is as follows:

[0193]

[0194] Where c is the position of the interpolation point in the x direction, scaleCol represents the x direction data scale;

[0195]

[0196] Where r is the position of the interpolation point in the y direction, scaleRow represents the y direction data scale;

[0197] And for the boundary points, since it cannot meet the position requirements in the formula shown above, the following formula is used for the row boundary:

[0198] T P = T W + T E

[0199] T W = T(row-1, col) x (1-Dis(x))

[0200] T E = T(row-1, col+1) x (1-Dis(x))

[0201] The following formula is used for the column boundary:

[0202] T P = T S + T N

[0203] T W = T(row, col-1) x (1-Dis(y))

[0204] T W = T(row+1, col-1) x (1-Dis(y))

[0205] The interpolated data is then called back to the Js file through Ajax, and the drawing function library used is the D3.js function library to create an SVG canvas. The advantage of this drawing method is that the drawn graph is a vector scaling graph. The advantage of the vector scaling graph is that the display effect of the graph can be automatically adjusted with the enlargement or reduction of the graph, preventing image distortion and unsightly problems caused by enlargement or reduction. After the program runs, the Web visual human-computer interaction interface appears, which can display the image.

[0206] For the curve display of the temperature measurement data, the Echarts open source visualization chart is used, the corresponding data table in the SQLServer database is read by using the SQL language, the curve chart display format is set, and finally the curve chart display is completed.

[0207] Further, as a specific implementation of the above-mentioned crystallizer heat flux inverse calculation method based on fiber temperature measurement, the embodiment of the application provides a crystallizer heat flux inverse calculation system based on fiber temperature measurement, as shown in the system comprises an interface module, a model module and a controller module, wherein: Figure 9 The controller module is configured to control the model module to perform the steps in the above method.

[0208] The interface module is configured to visually display the target distribution obtained by the model module.

[0209] The interface module is configured to visually display the target distribution obtained by the model module.

[0210] In a specific application scenario, the controller module is configured to:

[0211] The controller module is configured to control the model module to convert the measured temperature into a digital format using a fiber grating data demodulator, and store the converted measured temperature in a database; and determine a smoothing coefficient and use an exponential smoothing method to perform noise reduction processing on the stored measured temperature.

[0212] In a specific application scenario, the controller module is configured to:

[0213] The controller module is configured to determine production line parameters, wherein the production line parameters at least include continuous casting machine size parameters on the continuous casting production line; and based on the arrangement position of the fiber temperature sensor in the crystallizer, construct a forward problem model of the crystallizer heat transfer based on the longitudinal section of the crystallizer.

[0214] In a specific application scenario, the termination condition is based on the error between the measured temperature and the calculated temperature obtained in this iteration.

[0215] In a specific application scenario, the controller module is configured to:

[0216] The controller module is configured to obtain an empirical formula of the longitudinal heat flux distribution of the crystallizer, and fit the measured temperature based on the empirical formula to determine the parameters in the empirical formula.

[0217] In a specific application scenario, the interface module is configured to:

[0218] Draw the calculated temperature curve and the measured temperature curve, and determine the accuracy of the forward problem model and the inverse problem model according to the comparison result between the calculated temperature and the measured temperature.

[0219] Web front-end visualization technology is used to display the heat flux distribution map corresponding to the target distribution in a visualization interface;

[0220] The calculated temperature is interpolated, and a temperature curve and temperature cloud map are plotted based on the interpolated calculated temperature using web front-end visualization technology.

[0221] In specific application scenarios, the forward problem model may optionally include the heat transfer control equation of the crystallizer, the heat flux boundary conditions between the inner side of the narrow copper plate of the crystallizer and the billet, and the forced convection boundary conditions between the outer side of the narrow copper plate of the crystallizer and the cooling water.

[0222] According to another aspect of this application, a medium is provided on which a program or instructions are stored, which, when executed by a processor, implement the above-described method for calculating the heat flux of a crystallizer based on fiber optic temperature measurement.

[0223] It should be noted that other corresponding descriptions of the functional modules involved in the crystallizer heat flux back calculation system based on fiber optic temperature measurement provided in this application embodiment can be found in the corresponding descriptions in the above method, and will not be repeated here.

[0224] Based on the above method, the present application also provides a storage medium storing a computer program that, when executed by a processor, implements the above-mentioned method for calculating the heat flux of a crystallizer based on fiber optic temperature measurement.

[0225] Based on this understanding, the technical solution of this application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, or portable hard drive) and includes several instructions to cause an electronic device (such as a personal computer, server, or network device) to execute the methods described in the various implementation scenarios of this application.

[0226] Based on the above, Figures 1 to 8 The method shown, and Figure 9 To achieve the above objectives, the present application also provides a device, specifically a personal computer, server, network device, etc., in the illustrated virtual system embodiment. This electronic device includes a storage medium and a processor; the storage medium stores a computer program; the processor executes the computer program to achieve the above-described objectives. Figures 1 to 8 The method for calculating the heat flux of a crystallizer based on fiber optic temperature measurement is shown.

[0227] Optionally, the electronic device can further include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, and the like. The user interface can include a display, an input unit such as a keyboard, and the like. Optionally, the user interface can further include a USB interface, a card reader interface, and the like. The network interface can optionally include a standard wired interface, a wireless interface (such as a Bluetooth interface, a WI-FI interface), and the like.

[0228] Those skilled in the art can understand that the electronic device structure provided by the embodiment does not constitute a limitation on the electronic device, and can include more or fewer components, or combine certain components, or different component arrangements.

[0229] The storage medium can further include an operating system and a network communication module. The operating system is a program for managing and saving hardware and software resources of the electronic device, and supports the running of information processing programs and other software and / or programs. The network communication module is used to realize communication between the controls in the storage medium, and communication with other hardware and software in the entity device.

[0230] From the above description of the embodiments, those skilled in the art can clearly understand that the present application can be implemented by means of software with a necessary general hardware platform, or by hardware.

[0231] Those skilled in the art can understand that the drawings are only a schematic diagram of a preferred implementation scenario, and the units or processes in the drawings are not necessarily essential for implementing the present application. Those skilled in the art can understand that the units in the system in the implementation scenario can be distributed in the system in the implementation scenario according to the description of the implementation scenario, or can be changed and located in one or more systems different from the implementation scenario. The units in the above implementation scenario can be combined as one unit, or can be further split into multiple sub-units.

[0232] The above serial numbers of the present application are only for description, and do not represent the advantages and disadvantages of the implementation scenario. The above disclosure is only several specific implementation scenarios of the present application, but the present application is not limited thereto. Any changes that can be thought of by those skilled in the art should fall within the protection scope of the present application.

Claims

1. A crystallizer heat flux inverse calculation method based on optical fiber temperature measurement, characterized by, The method comprises the following steps: Step 1: arranging the fiber temperature sensor longitudinally into a column, and arranging the fiber temperature sensor at the narrow surface copper plate part of the crystallizer on the continuous casting billet production line, and connecting the fiber temperature sensor and the fiber grating data demodulator; Step 2: collecting the measured temperature of the crystallizer in real time by using the fiber temperature sensor on the continuous casting billet production line; Step 3: performing noise reduction processing on the measured temperature, obtaining an empirical formula of the longitudinal heat flux distribution of the crystallizer, and performing empirical formula fitting on the noise reduction processed measured temperature based on the empirical formula to determine the parameters in the empirical formula; wherein the empirical formula is used to roughly calculate the heat flux distribution and set the initial value of iteration of the inverse problem model; Step 4: determining the production line parameters; and constructing a forward problem model of the heat transfer of the crystallizer based on the longitudinal section of the crystallizer according to the production line parameters and the arrangement position of the fiber temperature sensor in the crystallizer; wherein the production line parameters at least include the size parameters of the continuous casting machine on the continuous casting production line; the forward problem model is used to fit the longitudinal distribution of the heat flux of the crystallizer; Step 5: constructing an inverse problem model of the heat transfer of the crystallizer according to the known heat transfer conditions of the crystallizer and the forward problem model, wherein the inverse problem model is used to calculate the heat flux according to the measured temperature; Step 6: bringing the preset initial heat flux into the forward problem model, if the termination condition is not met, continue to iterate the inverse problem model, and bring the heat flux obtained by solving the inverse problem model into the forward problem model, rejudge whether the termination condition is met, until the termination condition is met, and the heat flux longitudinal distribution obtained by the forward problem model in this iteration is determined as the target distribution; Step 7: analyzing the target distribution and visually displaying the target distribution.

2. The method according to claim 1, wherein, The noise reduction processing on the measured temperature comprises: using the fiber grating data demodulator to convert the measured temperature into a digital format, and storing the converted measured temperature in a database; determining a smoothing coefficient, and using an exponential smoothing method to perform noise reduction processing on the stored measured temperature.

3. The method according to claim 1, wherein, The termination condition is judged based on the error between the measured temperature and the calculated temperature obtained in this iteration.

4. The method according to claim 3, wherein, The visual display of the target distribution comprises: drawing a calculated temperature curve and a measured temperature curve, and determining the accuracy of the forward problem model and the inverse problem model according to the comparison result between the calculated temperature and the measured temperature; using Web front-end visualization technology to display the heat flux distribution graph corresponding to the target distribution on a visual interface; performing data interpolation processing on the calculated temperature, and using Web front-end visualization technology to draw a temperature curve graph and a temperature cloud graph based on the interpolated calculated temperature.

5. The method of calculating the heat flux of crystallizer based on fiber temperature according to claim 1, characterized in that, The forward problem model comprises a heat transfer control equation of the crystallizer, a heat flux boundary condition between the inside of the narrow surface copper plate of the crystallizer and the billet, and a forced convection boundary condition between the outside of the narrow surface copper plate of the crystallizer and the cooling water.

6. A crystallizer heat flux inverse calculation system based on fiber temperature measurement, comprising an interface module, a model module and a controller module; The controller module is configured to control the model module to perform the method of any one of claims 1-5. The interface module is configured to visualize the target distribution obtained by the model module.

7. A storage medium having stored thereon a program or instructions, characterized in that The program or the instruction, when executed by the processor, implements the method of any one of claims 1-5.

8. An electronic device comprising a storage medium and a processor, characterized in that The storage medium stores a computer program, and the processor executes the computer program to implement the method of any one of claims 1-5.

Images