A temperature trend prediction method for a magnesium-carbon brick drying kiln

Abstract

The application relates to the technical field of data processing, in particular to a temperature trend prediction method for a magnesium-carbon brick drying kiln, which comprises the following steps: acquiring the temperature and humidity of each temperature zone in the drying kiln at each moment; for each temperature zone, evaluating the correlation damage of temperature and humidity in the local range at each moment and the fluctuation of the temperature zone itself to obtain the decoupling eigenvalue of each temperature zone at each moment; evaluating the difference of temperature change between adjacent temperature zones to obtain the spatial disturbance value of each temperature zone at each moment; determining the adjustment coefficient of each temperature zone at each moment to adjust the horizontal smoothing coefficient of the prediction model, and using the adjusted horizontal smoothing coefficient to perform real-time prediction on the temperature trend of each temperature zone in the drying kiln. The application significantly improves the accuracy and anti-interference ability of temperature prediction of different temperature zones in the drying kiln.

Description

Technical Field

[0001] This application relates to the field of data processing technology, specifically to a method for predicting the temperature trend of a magnesia-carbon brick drying kiln. Background Technology

[0002] Magnesia-carbon bricks are key refractory materials in steelmaking. During production, the pressed magnesia-carbon brick blanks need to undergo heat treatment in a drying kiln to cure the phenolic resin and remove moisture from the brick blanks. Therefore, the control of the drying temperature directly affects the product performance and service life of magnesia-carbon bricks. As the core equipment in magnesia-carbon brick production, the drying kiln typically adopts a tunnel-type structure, with its interior divided into different temperature zones. The brick blanks are loaded in layers onto kiln cars and pass through each temperature zone sequentially via tracks to complete the drying process.

[0003] When predicting the temperature trend of a magnesia-carbon brick drying kiln, the significant differences in the moisture content of the brick blanks entering the kiln and the large amount of water evaporation during drying will cause varying degrees of temperature fluctuations. Secondly, the temperature zones inside the drying kiln are interconnected, and the temperature between adjacent temperature zones will undergo air convection heat exchange, resulting in a high degree of non-stationarity and spatial coupling interference in the temperature evolution within the drying kiln. Traditional quadratic exponential smoothing algorithms generally use a fixed horizontal smoothing coefficient, which leads to a serious trend capture lag when facing drastic changes in the local thermodynamic field, and an oversensitive prediction oscillation when facing random disturbances in adjacent areas, affecting the accuracy of temperature trend prediction. Summary of the Invention

[0004] To address the aforementioned technical problems, a method for predicting the temperature trend of a magnesia-carbon brick drying kiln is provided to solve the existing issues.

[0005] The solution to the technical problem presented in this application is a method for predicting the temperature trend of a magnesia-carbon brick drying kiln, comprising the following steps: Obtain the temperature and humidity of each temperature zone inside the drying cellar at various times; For each temperature zone, the degree of disruption of the correlation between temperature and humidity in the local area at each time point and its own fluctuation are evaluated to characterize the dynamic changes in the temperature-humidity coupling relationship and obtain the decoupling characteristic value of each temperature zone at each time point. The differences in temperature changes between adjacent temperature zones are evaluated to characterize the temperature conduction disturbance between adjacent temperature zones, and the spatial disturbance value of each temperature zone at each time is obtained. Based on the decoupling eigenvalues ​​and spatial perturbation values, the adjustment coefficients for each temperature zone at each time point are determined to adjust the horizontal smoothing coefficient of the prediction model. Using the adjusted horizontal smoothing coefficients, the temperature trend of each temperature zone in the drying kiln is predicted in real time.

[0006] Preferably, obtaining the decoupling characteristic values ​​of each temperature zone at each time step includes: For each time point in each temperature zone, a preset time window is constructed; for each time window, the differences in the changing trends of temperature and humidity are analyzed, and the trend difference is calculated; the discrete change characteristics of temperature and humidity within the time window are analyzed, and the local fluctuation is calculated. The decoupling eigenvalues ​​are positively correlated with the trend difference and local volatility.

[0007] Preferably, the calculation process of the trend difference is as follows: negatively map the humidity, perform curve fitting on all negative mapping results within the time window, perform curve fitting on all temperatures within the time window, and calculate the difference between the two fitted curves as the trend difference.

[0008] Preferably, the calculation process for the local volatility is as follows: The dispersion of all temperatures and the dispersion of all humidity within the time window are calculated separately and used as the local dispersion of temperature and humidity at each time point, respectively. The local fluctuation is the sum of the local dispersion of temperature and the local dispersion of humidity.

[0009] Preferably, obtaining the spatial perturbation value of each temperature zone at each time step includes: For each temperature zone, the temperature zones adjacent to it are defined as neighboring temperature zones; Analyze the differences in temperature trends between each temperature zone and neighboring temperature zones within a time window, and calculate the temperature deviation. Analyze the differences in temperature fluctuation between each temperature zone and its neighboring temperature zones within a time window, calculate the fluctuation difference, and select the minimum value of the fluctuation difference between each temperature zone and all its neighboring temperature zones at each time point. The spatial disturbance value is positively correlated with the temperature deviation, but negatively correlated with the minimum value.

[0010] Preferably, the temperature deviation is calculated as follows: the differences between all temperatures in each temperature zone within each time window and all temperatures in all neighboring temperature zones within the same time window are positively integrated to form the temperature deviation.

[0011] Preferably, the formula for calculating the fluctuation difference is: calculate the difference between the local temperature dispersion of each temperature zone and its neighboring temperature zones at each time point, and use it as the fluctuation difference.

[0012] Preferably, the adjustment coefficient is positively correlated with the decoupling characteristic value and negatively correlated with the spatial disturbance value.

[0013] Preferred, the first Each temperature zone The adjusted horizontal smoothing coefficient at each moment The calculation formula is: ,in, This is the preset minimum horizontal smoothing coefficient. The preset maximum horizontal smoothing coefficient, For the first Each temperature zone Adjustment factor for time.

[0014] Preferably, the real-time prediction of the temperature trend of each temperature zone in the drying cellar includes: using a quadratic exponential smoothing algorithm to predict the temperature of each temperature zone in real time based on the adjusted horizontal smoothing coefficient.

[0015] This application has at least the following beneficial effects: This application obtains the decoupled feature values ​​of each temperature zone at each time step. Its beneficial effect lies in identifying local unsteady-state conditions caused by the explosive evaporation of moisture in brick blanks by quantifying the degree of disruption of local temperature and humidity correlations and the intensity of local data fluctuations. It keenly captures the instantaneous trends caused by physical phase transitions, solving the problem that traditional algorithms cannot perceive the inherent logical disruption of temperature and humidity, and providing a basis for improving the model's response sensitivity under severe perturbations. Furthermore, it obtains the spatial perturbation values ​​of each temperature zone at each time step. Its beneficial effect lies in quantifying the degree of temperature interference from neighboring temperature zones by analyzing the differences in temperature changes between adjacent temperature zones and the comparative relationship between their fluctuation intensities. This effectively identifies spurious fluctuations caused by the diffusion of temperature fluctuations from neighboring regions, avoiding overreaction of the prediction model to external random noise. This method ensures the smoothness and robustness of temperature prediction results. An adjustment coefficient for each temperature zone at each time point is determined to adjust the horizontal smoothing coefficient of the prediction model. Using the adjusted horizontal smoothing coefficient, the temperature trend of each temperature zone within the drying kiln is predicted in real time. The beneficial effect lies in the dynamic adjustment of the horizontal smoothing coefficient through the adjustment coefficient. When a temperature zone is highly affected by its own moisture evaporation, the horizontal smoothing coefficient is increased to allow the prediction to quickly track actual temperature changes. When a temperature zone is significantly affected by conduction interference from neighboring temperature zones, the horizontal smoothing coefficient is decreased, making the prediction rely more on historical trends and filtering out instantaneous interference. Through this adaptive mechanism, the accuracy and anti-interference ability of temperature prediction for different temperature zones within the drying kiln are significantly improved, effectively avoiding quality problems such as over-firing or insufficient drying of magnesia-carbon bricks due to temperature runaway. Attached Figure Description

[0016] The following section provides a more detailed explanation of a method for predicting the temperature trend of a magnesium-carbon brick drying kiln according to this application, with reference to the accompanying drawings.

[0017] Figure 1 A flowchart illustrating the steps of a method for predicting the temperature trend of a magnesia-carbon brick drying kiln provided in this application embodiment; Figure 2 A flowchart illustrating the steps of the method for obtaining decoupled feature values ​​provided in this application embodiment. Detailed Implementation

[0018] The following, in conjunction with the accompanying drawings and embodiments, provides a more detailed description of the temperature trend prediction method for a magnesia-carbon brick drying kiln proposed in this application.

[0019] Please see Figure 1 The document illustrates a flowchart of a method for predicting the temperature trend of a magnesia-carbon brick drying kiln according to an embodiment of this application. The method includes the following steps: Step 1: Obtain the temperature and humidity of each temperature zone in the drying cellar at each time.

[0020] Magnesia-carbon bricks are non-burning carbon composite refractory materials made from high-melting-point alkaline oxide magnesium oxide and high-melting-point carbon materials that are difficult to wet by slag, with the addition of various non-oxide additives and bonded together with carbonaceous binders. The heat treatment, or drying process, of magnesia-carbon bricks mainly refers to the removal of moisture from the product. However, the heat treatment process is essentially a curing process of phenolic resin. Therefore, the drying temperature has a direct impact on the performance of magnesia-carbon bricks, requiring precise prediction and control of the temperature during the drying process.

[0021] Based on the above analysis, the magnesia-carbon brick drying kiln is typically divided into multiple temperature zones along its tunnel length, including a preheating zone, a constant temperature zone, and a high temperature zone, according to the drying process requirements. The preheating zone is used to slowly raise the temperature of the brick blanks to near the resin softening point, preventing thermal shock cracking. The constant temperature zone is used to remove a large amount of moisture from the brick blanks and is the core area of ​​the drying process. The high temperature zone is used to fully cure the phenolic resin, ensuring the final strength of the brick blanks. Therefore, S-type thermocouples and humidity sensors are deployed at the top of each temperature zone to collect the temperature and humidity of each zone in real time during the operation of the drying kiln. In this embodiment, the data acquisition frequency is 1Hz. As for other implementation methods, the implementer can set the frequency according to the actual situation.

[0022] After filling missing values ​​in the collected data, the data is normalized. In this embodiment, the K-nearest neighbor algorithm is used for filling missing values, and the maximum-minimum normalization method is used for normalization. The K-nearest neighbor algorithm and the maximum-minimum normalization method are well-known techniques and will not be described in detail here.

[0023] Thus, the temperature and humidity of each temperature zone inside the drying cellar at each time point are obtained.

[0024] Step 2: For each temperature zone, assess the extent to which the correlation between temperature and humidity in the local area is disrupted and its own fluctuations at each time point to characterize the dynamic changes in the temperature-humidity coupling relationship and obtain the decoupling characteristic value of each temperature zone at each time point.

[0025] Furthermore, during the operation of the magnesia-carbon brick drying kiln, the pressed magnesia-carbon brick blanks are typically placed in layers on specific kiln cars within the kiln. These kiln cars are then moved and dried in different temperature zones by hydraulic or mechanical devices, moving along tracks laid at the bottom of the drying kiln. Because the moisture content of the magnesia-carbon brick blanks entering the drying kiln varies, during the drying process, the brick blanks with higher moisture content experience rapid moisture evaporation, absorbing a large amount of heat. Since the rates of moisture evaporation and heat conduction differ significantly in their physical mechanisms, this disrupts the correlation between temperature and humidity, weakening the originally stable negative correlation and causing significant fluctuations in temperature and humidity. As drying continues, the moisture content of the brick blanks gradually decreases, moisture evaporation stabilizes, and the changes in temperature and humidity also tend to stabilize, with the correlation gradually recovering and remaining at a high level. This dynamic fluctuation in the correlation means that traditional quadratic exponential smoothing algorithms cannot accurately capture sudden trend shifts, resulting in significant prediction lags or oscillations.

[0026] Secondly, when using the quadratic exponential smoothing algorithm to predict temperature, its horizontal smoothing coefficient is used to control the sensitivity to the current data and determines the model's trust in the current data. The larger the value, the faster it can track the actual temperature changes. Therefore, when the correlation between temperature and humidity is detected to be disrupted, the horizontal smoothing coefficient should be increased to capture the true instantaneous temperature trend and effectively avoid prediction lag errors caused by over-reliance on historical stable data.

[0027] Based on this analysis, for each temperature zone, decoupling characteristic values ​​are calculated by analyzing the correlation between the changing trends of temperature and humidity within a local area, as well as the fluctuations in temperature and humidity. The flowchart of the method for obtaining decoupling characteristic values ​​provided in this application embodiment is shown below. Figure 2 As shown, specifically: For each time point within each temperature zone, a preset time window is constructed; In this embodiment, 300 consecutive moments preceding each moment are grouped into a time window. As for other implementation methods, the implementer can set it according to the actual situation.

[0028] Negative mapping is performed on humidity, curve fitting is performed on all negative mapping results within the time window, curve fitting is performed on all temperatures within the time window, and the difference between the two fitted curves is calculated as the trend difference quantity. In this embodiment, the negative mapping process is as follows: the humidity is the result after normalization, so the difference between the value 1 and the normalized humidity is taken as the result of the negative mapping; secondly, the least squares method is used for curve fitting; the trend difference is calculated as follows: the Fraser distance between the two fitted curves is taken as the trend difference. The least squares method and the calculation of the Fraser distance are well-known techniques and will not be described in detail here.

[0029] It should be noted that, since temperature and humidity exhibit a significant negative correlation during normal, stable, and dry periods, a negative mapping of humidity is used to transform the inverse evolution trend of temperature and humidity into a geometric correspondence in the same direction, in order to determine when the correlation between temperature and humidity is disrupted.

[0030] The dispersion of all temperatures within the time window is calculated as the local dispersion of the temperature at each time point. The dispersion of all humidity levels within the time window is calculated as the local dispersion of humidity at each time point. For each temperature zone, the sum of the local dispersion of temperature and the local dispersion of humidity is taken as the local fluctuation at each time point. In this embodiment, the degree of dispersion is measured by calculating the coefficient of variation of all humidity and all temperature within the time window. The calculation of the coefficient of variation is a well-known technique and will not be described in detail here.

[0031] The decoupling characteristic value, trend difference, and local volatility of each temperature zone at each time point are all positively correlated. It should be noted that a positive correlation means that the dependent variable increases as the independent variable increases, and decreases as the independent variable decreases.

[0032] In this embodiment, the product of the trend difference and the local volatility is used as the decoupling characteristic value.

[0033] It should be noted that the greater the trend difference, the more inconsistent the negative mapping results between temperature and humidity are, reflecting the more significant the weakening of the correlation between temperature and humidity; the greater the local fluctuation, the more drastic the fluctuation of temperature or humidity itself is, and the more frequent and unstable the heat exchange in this area is. The larger the decoupling characteristic value is, the more unstable the disturbance environment is, reflecting the enhanced local thermodynamic field disturbance caused by the difference in the moisture content of the brick blank. The horizontal smoothing coefficient should be increased in the future.

[0034] Thus, the decoupling characteristic values ​​of each temperature zone at each time point are obtained.

[0035] Step 3: Evaluate the differences in temperature changes between adjacent temperature zones to characterize the temperature conduction interference between adjacent temperature zones and obtain the spatial perturbation value of each temperature zone at each time.

[0036] Furthermore, due to the long tunnel-like spatial layout inside the drying kiln, adjacent temperature zones are dynamically linked through air convection and heat exchange. If the moisture content of the bricks in one temperature zone is significantly higher, the humid and cold air generated by its intense moisture evaporation will interfere with adjacent temperature zones through negative pressure flow and heat diffusion within the tunnel, causing non-local instantaneous temperature and humidity fluctuations in those zones. Secondly, for a given temperature zone, if its temperature fluctuation is less than that of adjacent zones, it indicates that the temperature fluctuation in that zone is caused by instantaneous interference from strong evaporation in adjacent areas conducted through air convection; conversely, it indicates that the temperature fluctuation in that zone originates from the explosive evaporation of moisture from its own bricks. Therefore, when predicting temperature, if the temperature fluctuation is caused by instantaneous external interference, it is necessary to reduce the horizontal smoothing coefficient to enhance the reliance on historical data and avoid the predicted value being skewed by instantaneous fluctuations.

[0037] Based on the above analysis, by analyzing the differences in temperature fluctuations and trends within a local area between each temperature zone and its adjacent temperature zones, as well as the differences in temperature fluctuations, the spatial disturbance value is calculated, specifically as follows: For each temperature zone, the temperature zones adjacent to it are defined as neighboring temperature zones; The differences between all temperatures of each temperature zone within each time window and all temperatures of all its neighboring temperature zones within the same time window are calculated and positively integrated to form the temperature deviation. In this embodiment, all temperatures within a time window for each temperature zone are combined to form a temperature sequence. The Euclidean distance between each temperature zone and its neighboring temperature zones at each time point is calculated. If each temperature zone has two neighboring temperature zones, the sum of the Euclidean distances between each temperature zone and all its neighboring temperature zones at each time point is used as the temperature deviation. If each temperature zone has only one neighboring temperature zone, the Euclidean distance between each temperature zone and that neighboring temperature zone is used as the temperature deviation. The calculation of the Euclidean distance is a well-known technique and will not be elaborated here. As other implementation methods, implementers may use other methods of the prior art, such as DTW distance, etc. This embodiment does not impose any special restrictions on this.

[0038] Calculate the difference in local temperature dispersion between each temperature zone and its neighboring temperature zones at each time point, and use it as the fluctuation difference quantity; In this embodiment, the difference between the local temperature dispersion of each temperature zone and its neighboring temperature zones at each time point is calculated as the fluctuation difference.

[0039] Select the minimum value of the fluctuation difference between each temperature zone and all its neighboring temperature zones at each time point; For each temperature zone, the spatial disturbance value at each time point is positively correlated with the temperature deviation, and negatively correlated with the minimum value. It should be noted that a negative correlation means that the dependent variable decreases as the independent variable increases, and increases as the independent variable decreases.

[0040] In this embodiment, the minimum value is negatively mapped, and its product with the temperature deviation is taken as the value. The negative mapping process involves using an exponential function for negative mapping, assuming the minimum value is denoted as... ,but The result is used as the result of the negative mapping, where, This represents an exponential function with the natural constant as its base.

[0041] It should be noted that the larger the temperature deviation, the more significantly the temperature evolution trajectory of this temperature zone differs from that of adjacent temperature zones, and the more uneven the spatial distribution of the temperature field. The smaller the fluctuation difference, the smaller the minimum value. When the minimum value is less than 0, it indicates that the temperature fluctuation of this temperature zone is less than that of adjacent temperature zones, and the temperature fluctuation of this temperature zone is mainly affected by the fluctuation of adjacent regions. When the minimum value is greater than 0, it indicates that the temperature fluctuation of this temperature zone mainly originates from its own evaporation. If the minimum value is equal to 0, it indicates that the temperature fluctuations of the two temperature zones are comparable and in a relatively balanced state. The larger the obtained spatial disturbance value, the more the temperature fluctuation of this temperature zone is mainly caused by external conduction. In this case, the reliance on historical data should be increased during prediction, and the horizontal smoothing coefficient should be reduced in the future.

[0042] Thus, the spatial perturbation values ​​for each temperature zone at each time point are obtained.

[0043] Step 4: Based on the decoupling eigenvalues ​​and spatial perturbation values, determine the adjustment coefficients for each temperature zone at each time point to adjust the horizontal smoothing coefficients of the prediction model. Using the adjusted horizontal smoothing coefficients, the temperature trend of each temperature zone in the drying kiln is predicted in real time.

[0044] Furthermore, based on the decoupling eigenvalues ​​and spatial perturbation values, the adjustment coefficients are determined as follows: The adjustment coefficient of each temperature zone at each time point is positively correlated with the decoupling characteristic value, but negatively correlated with the spatial disturbance value. In this embodiment, the normalized result of the ratio of the decoupled eigenvalue and the spatial perturbation value is used as the adjustment coefficient of each temperature zone at each time. The maximum-minimum normalization method is used to normalize the ratio of each temperature zone at each time and all previous times. The maximum-minimum method is a well-known technique and will not be described in detail here.

[0045] It should be noted that, in order to avoid the denominator being 0 when calculating the ratio, a parameter adjustment factor is added to the denominator. In this embodiment, the parameter adjustment factor is set to 0.1. As for other implementation methods, the implementer can set it according to the actual situation. Secondly, the larger the adjustment coefficient, the weaker the temperature and humidity correlation in the temperature zone, the more violent the fluctuation, and the less the conduction interference from adjacent areas. At this time, the temperature change is mainly driven by real factors such as its own moisture evaporation. When making predictions, the trust in the current data should be enhanced to quickly track the real temperature trend.

[0046] Furthermore, based on the adjustment coefficient, the horizontal smoothing coefficient of the quadratic exponential smoothing algorithm is adjusted, specifically as follows: The formula for calculating the adjusted horizontal smoothing coefficient for each temperature zone at each time point is as follows: in, For the first Each temperature zone The corresponding adjusted horizontal smoothing coefficient at each moment. This is the preset minimum horizontal smoothing coefficient. The preset maximum horizontal smoothing coefficient, For the first Each temperature zone Adjustment factor for time; In this embodiment, the preset minimum horizontal smoothing coefficient is 0.1 and the preset maximum horizontal smoothing coefficient is 0.9. As for other implementation methods, the implementer can set them according to the actual situation.

[0047] Based on the adjusted horizontal smoothing coefficient, the temperature of each temperature zone is predicted in real time using a quadratic exponential smoothing algorithm. Based on the temperature prediction results, the temperature inside the drying kiln is regulated to avoid quality defects in magnesia-carbon bricks caused by temperature fluctuations. In this embodiment, the trend smoothing coefficient of the quadratic exponential smoothing algorithm is used to control the response speed to trend changes. When the value is close to 1, the trend changes quickly, which is suitable for capturing sudden trends. When the value is close to 0, the trend changes slowly, which is suitable for describing stable growth scenarios. In this embodiment, the trend smoothing coefficient is set to 0.2. As other implementation methods, implementers can set it according to the actual situation. The quadratic exponential smoothing algorithm is a well-known technology and will not be described in detail here.

[0048] It should be noted that by using the predicted temperature values ​​at continuous times, a temperature trend curve for a single temperature zone is constructed, and the trend curve is fed back to the temperature and humidity control module of the drying kiln in real time to dynamically adjust the supplementary heating and combustion power and the dehumidification fan frequency of the corresponding temperature zone, thereby achieving precise thermal compensation and feedforward control of the drying process of magnesia-carbon bricks.

Claims

1. A method for predicting the temperature trend of a magnesia-carbon brick drying kiln, characterized in that, The method includes the following steps: Obtain the temperature and humidity of each temperature zone inside the drying cellar at various times; For each temperature zone, the degree of disruption of the correlation between temperature and humidity in the local area at each time point and its own fluctuation are evaluated to characterize the dynamic changes in the temperature-humidity coupling relationship and obtain the decoupling characteristic value of each temperature zone at each time point. The differences in temperature changes between adjacent temperature zones are evaluated to characterize the temperature conduction disturbance between adjacent temperature zones, and the spatial disturbance value of each temperature zone at each time is obtained. Based on the decoupling eigenvalues ​​and spatial perturbation values, the adjustment coefficients for each temperature zone at each time point are determined to adjust the horizontal smoothing coefficient of the prediction model. Using the adjusted horizontal smoothing coefficients, the temperature trend of each temperature zone in the drying kiln is predicted in real time.

2. The method for predicting the temperature trend of a magnesia-carbon brick drying kiln as described in claim 1, characterized in that, The process of obtaining the decoupling characteristic values ​​of each temperature zone at each time step includes: For each time point in each temperature zone, a preset time window is constructed; for each time window, the differences in the changing trends of temperature and humidity are analyzed, and the trend difference is calculated; the discrete change characteristics of temperature and humidity within the time window are analyzed, and the local fluctuation is calculated. The decoupling eigenvalues ​​are positively correlated with the trend difference and local volatility.

3. The method for predicting the temperature trend of a magnesia-carbon brick drying kiln as described in claim 2, characterized in that, The calculation process for the trend difference is as follows: negatively map the humidity, perform curve fitting on all negative mapping results within the time window, perform curve fitting on all temperatures within the time window, and calculate the difference between the two fitted curves as the trend difference.

4. The method for predicting the temperature trend of a magnesia-carbon brick drying kiln as described in claim 2, characterized in that, The calculation process for the local volatility is as follows: The dispersion of all temperatures and the dispersion of all humidity within the time window are calculated separately and used as the local dispersion of temperature and humidity at each time point, respectively. The local fluctuation is the sum of the local dispersion of temperature and the local dispersion of humidity.

5. The method for predicting the temperature trend of a magnesia-carbon brick drying kiln as described in claim 4, characterized in that, The process of obtaining the spatial perturbation values ​​of each temperature zone at each time point includes: For each temperature zone, the temperature zones adjacent to it are defined as neighboring temperature zones; Analyze the differences in temperature trends between each temperature zone and neighboring temperature zones within a time window, and calculate the temperature deviation. Analyze the differences in temperature fluctuation between each temperature zone and its neighboring temperature zones within a time window, calculate the fluctuation difference, and select the minimum value of the fluctuation difference between each temperature zone and all its neighboring temperature zones at each time point. The spatial disturbance value is positively correlated with the temperature deviation, but negatively correlated with the minimum value.

6. The method for predicting the temperature trend of a magnesia-carbon brick drying kiln as described in claim 5, characterized in that, The calculation process for the temperature deviation is as follows: the differences between all temperatures in each temperature zone within each time window and all temperatures in all neighboring temperature zones within the same time window are calculated and positively integrated to form the temperature deviation.

7. The method for predicting the temperature trend of a magnesia-carbon brick drying kiln as described in claim 5, characterized in that, The formula for calculating the fluctuation difference is as follows: calculate the difference between the local temperature dispersion of each temperature zone and its neighboring temperature zones at each time point, and use it as the fluctuation difference.

8. The method for predicting the temperature trend of a magnesia-carbon brick drying kiln as described in claim 1, characterized in that, The adjustment coefficient is positively correlated with the decoupling characteristic value, but negatively correlated with the spatial disturbance value.

9. The method for predicting the temperature trend of a magnesia-carbon brick drying kiln as described in claim 1, characterized in that, No. Each temperature zone The adjusted horizontal smoothing coefficient at each moment The calculation formula is: ,in, This is the preset minimum horizontal smoothing coefficient. The preset maximum horizontal smoothing coefficient, For the first Each temperature zone Adjustment factor for time.

10. The method for predicting the temperature trend of a magnesia-carbon brick drying kiln as described in claim 1, characterized in that, The real-time prediction of the temperature trend of each temperature zone in the drying cellar includes: using a quadratic exponential smoothing algorithm to predict the temperature of each temperature zone in real time based on the adjusted horizontal smoothing coefficient.

Images