Electronic display glass kiln auxiliary control method and system based on artificial intelligence

Abstract

The invention discloses an electronic display glass kiln auxiliary control method and system based on artificial intelligence, relates to the technical field of kiln pressure predictive control, and solves the problems that in glass kiln control, a time period suitable for kiln pressure control is difficult to set according to kiln combustion characteristics and environmental factors, and the kiln pressure control is difficult to set. And the prediction of the relevance between the flue gas emission and the atmospheric pressure is neglected, so that the precise control of the kiln pressure is difficult to realize. According to the invention, time periods are divided based on environmental data, smoke data of a next time period are obtained through a smoke prediction module, and atmospheric pressure of the next time period is obtained through an atmospheric pressure prediction model; the smoke outlet area is determined based on the smoke data and the atmospheric pressure of the next time period, and a rotary gate plate in the smoke outlet is adjusted based on the smoke outlet area; according to the method, through a data-driven prediction model and a dynamic regulation and control mechanism, glass kiln pressure discharge control is upgraded from experience-driven control to intelligent prediction control, and the problem of response lag of kiln pressure control is solved.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent manufacturing and relates to kiln pressure prediction and control technology, specifically to an electronic display glass kiln auxiliary control method and system based on artificial intelligence. Background Art

[0002] Electronic display glass is a crucial foundational material for the modern electronic information industry, widely used in electronic products such as smartphones, tablets, and televisions. The quality and production efficiency of electronic display glass are becoming increasingly important. As the core equipment in electronic display glass production, the operating status of the kiln directly affects product quality, energy consumption, and production costs.

[0003] At present, most electronic display glass furnace pressure control methods find it difficult to set the time period that adapts to the kiln pressure control according to the kiln combustion characteristics and environmental factors. They can only make passive fuzzy adjustments based on operating experience, and cannot conduct targeted analysis of the temperature and humidity stable periods and the periods of violent fluctuations, resulting in poor and unstable kiln pressure control accuracy. At the same time, they ignore the analysis and prediction of the correlation between flue gas emissions and atmospheric pressure, and cannot accurately control the kiln pressure. In addition, traditional kiln pressure regulation modes mostly rely on passive responses of real-time feedback, and there is a response lag problem.

[0004] Therefore, the present invention discloses an artificial intelligence-based electronic display glass furnace auxiliary control method and system to solve the above technical problems. Summary of the Invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes an electronic display glass kiln auxiliary control method and system based on artificial intelligence, which is used to solve the technical problems that in the control of the glass kiln, it is difficult to set the time period that adapts to the kiln pressure control according to the kiln combustion characteristics and environmental factors, and the correlation between flue gas emissions and atmospheric pressure is not predicted and analyzed, making it difficult to achieve precise control of the kiln pressure. The present invention divides the time period based on environmental data, obtains the flue gas data of the next time period through the flue gas prediction module, and obtains the atmospheric pressure of the next time period through the atmospheric pressure prediction model; determines the smoke outlet area based on the smoke data and atmospheric pressure of the next time period, and adjusts the rotating gate inside the smoke outlet based on the smoke outlet area when the next time period is reached to solve the above problems.

[0006] To achieve the above objectives, the first aspect of the present invention provides an artificial intelligence-based electronic display glass furnace auxiliary control method, comprising:

[0007] Collect environmental data at each time point on the smoke exhaust vent location, as well as the combustion duration, material type, and amount of the glass furnace; divide the time period based on the environmental data; the environmental data includes temperature, humidity, and altitude;

[0008] A flue gas prediction module is established. The combustion duration of the next time period, the type and amount of materials already used in this combustion are input into the flue gas prediction module to obtain the flue gas data for the next time period. The flue gas data includes the flue gas flow rate, flue gas pressure, and flue gas density at the smoke outlet. The flue gas prediction module is trained through an artificial intelligence model.

[0009] Establish an atmospheric pressure prediction model, input the environmental data of the next time period into the atmospheric pressure prediction model to obtain the atmospheric pressure of the next time period; wherein the atmospheric pressure prediction model is obtained through artificial intelligence model training;

[0010] The smoke outlet area is determined based on the smoke data and atmospheric pressure of the next time period, and the rotary shutter inside the smoke outlet is adjusted based on the smoke outlet area when the next time period arrives.

[0011] Preferably, the collection of environmental data of the smoke exhaust port location at each time point, as well as the burning time of the glass furnace, the type of material and the amount of material added, includes:

[0012] The temperature and humidity at the smoke exhaust outlet at each time point are obtained through the weather forecast platform, and the altitude of the smoke exhaust outlet is extracted from the database. The time difference between each time period and the start time of combustion is obtained through the timer, and the time difference is marked as the combustion duration; the type of material is obtained through the ingredient information transmitted from the raw material workshop, and the amount of each type of material manually designed in each time period is extracted from the database; wherein the types of materials include fuel, clarifier, colorant and decolorizer.

[0013] Preferably, dividing the time periods based on environmental data includes:

[0014] At the target time point of the previous day, the temperature and humidity of the smoke exhaust outlet location at each time point today are extracted. Consecutive time points with the same temperature and humidity are connected and integrated into the first period to be allocated. Time points not integrated into the first period to be allocated are integrated into several second periods to be allocated. The target time point is manually set.

[0015] Obtain the current allocated time range of the glass furnace, and divide the first time periods to be allocated into several time periods according to the middle value FZ of the allocated time range;

[0016] Extract the second time period to be allocated in sequence, obtain the variance WDC of each temperature and the variance SDC of each humidity in the second time period to be allocated, and obtain the dynamic duration DC based on the formula DC = FZ×α / exp(β×WDC / BWD+(1-β)×SDC / BSD); when the dynamic duration DC is greater than the maximum value FD of the allocation duration range, use the maximum value FD to update the dynamic duration DC; when the dynamic duration DC is less than the minimum value FX of the allocation duration range, use the minimum value FX to update the dynamic duration DC; divide the second time period to be allocated into several time periods according to the dynamic duration DC; wherein BWD is the manually set temperature standard deviation, and BSD is the manually set humidity standard deviation; α is the amplitude adjustment coefficient of the 1 / exp() function set according to the air pressure fluctuation amplitude of the day, and the value range of α is (0,1]; β is the proportional adjustment coefficient manually set according to the importance of temperature and humidity to atmospheric pressure changes, and the value range of β is (0,1].

[0017] Preferably, the step of obtaining the allocated time range of the current glass furnace includes:

[0018] Extract the time from the start of combustion in the current glass furnace to the appearance of smoke at the smoke outlet from the database, and subtract the material heating time from the time to obtain the smoke exhaust time for each combustion; wherein the material heating time is manually obtained based on the initial temperature of the glass furnace;

[0019] Obtain the variance of several smoke discharge durations, and determine whether the variance is less than a variance threshold. If so, extract the maximum and minimum values ​​from the several smoke discharge durations, and mark the duration range between the maximum and minimum values ​​as the allocated duration range. If not, remove the smoke discharge duration that has the largest difference from the mode of the smoke discharge durations from the several smoke discharge durations, and re-determine the variance until the variance is less than the variance threshold. Then, retain the remaining smoke discharge durations, and mark the duration range between the maximum and minimum values ​​among the remaining smoke discharge durations as the allocated duration range. The variance threshold is obtained empirically.

[0020] Preferably, the step of dividing the first time periods to be allocated into a plurality of time periods according to the middle value FZ of the allocation duration range includes:

[0021] Extract the first time period to be allocated in sequence, obtain the duration YC of the first time period to be allocated, and use the formula A planned number of times GC of the first time period to be allocated is obtained, and the first time period to be allocated is evenly divided into a plurality of time periods based on the planned number of times GC.

[0022] Preferably, the flue gas prediction module is obtained through artificial intelligence model training, including:

[0023] Extracting the combustion duration, the type and amount of materials already put into the combustion in several time periods from the historical flue gas database, and the corresponding flue gas data for the next time period; wherein the historical flue gas database includes the combustion duration, the type and amount of materials already put into the combustion in several time periods, and the flue gas data for the next time period manually set based on the combustion duration, the type and amount of materials already put into the combustion in several time periods;

[0024] The combustion duration, material type and amount in several time periods, as well as the corresponding flue gas data of the next time period are integrated into several groups of training data and verification data; the training data is used to train the artificial intelligence model, and the test data is used to test the trained artificial intelligence model, and the artificial intelligence model is adjusted according to the test results; finally, the input is the combustion duration of the next time period, the type and amount of materials already put into this combustion, and the output is the flue gas data of the next time period, and the flue gas prediction module is obtained; among them, the artificial intelligence model includes the BP neural network model and the RBF neural network model.

[0025] Preferably, the atmospheric pressure prediction model is obtained through artificial intelligence model training, including:

[0026] Extracting environmental data and atmospheric pressure for each time period from a historical atmospheric pressure database; wherein the historical atmospheric pressure database includes environmental data and monitored atmospheric pressure;

[0027] Integrate environmental data and atmospheric pressure into several groups of training data and test data; use the training data to train the artificial intelligence model, use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally, obtain an atmospheric pressure prediction model whose input is the environmental data of the next time period and output is the atmospheric pressure of the next time period; wherein, the artificial intelligence model includes a BP neural network model and an RBF neural network model.

[0028] Preferably, determining the smoke outlet area based on smoke data and atmospheric pressure in the next time period includes:

[0029] Extract the flue gas flow rate YQ, flue gas pressure YP, flue gas density YD and atmospheric pressure DP from the environmental data of the next time period;

[0030] The chimney outlet area YS for the next time period is determined based on the flue gas flow rate YQ, flue gas pressure YP, flue gas density YD and atmospheric pressure DP. The chimney outlet area YS satisfies the following formula:

[0031]

[0032] Preferably, the step of adjusting the rotary shutter inside the smoke outlet based on the area of ​​the smoke outlet when the next time period arrives includes:

[0033] When the time reaches the next time period, the chimney outlet area YS of the next time period is extracted, and the rotary gate in the smoke outlet is rotated according to the chimney outlet area YS.

[0034] The second aspect of the present invention provides an artificial intelligence-based electronic display glass furnace auxiliary control system, comprising: a model prediction module, and a data collection module and an auxiliary control module connected thereto;

[0035] The data collection module is used to collect environmental data of the smoke exhaust outlet at each time point, as well as the combustion time of the glass furnace, the type of material and the amount of material put in; and divide the time period based on the environmental data; wherein the environmental data includes temperature, humidity and altitude;

[0036] The model prediction module is used to establish a flue gas prediction module, input the combustion duration of the next time period, the type and amount of materials already put into this combustion into the flue gas prediction module to obtain the flue gas data of the next time period; establish an atmospheric pressure prediction model, input the environmental data of the next time period into the atmospheric pressure prediction model to obtain the atmospheric pressure of the next time period; wherein the flue gas data includes the flue gas flow rate, flue gas pressure and flue gas density at the smoke outlet; the flue gas prediction module is obtained through artificial intelligence model training; the atmospheric pressure prediction model is obtained through artificial intelligence model training;

[0037] The auxiliary control module is used to determine the smoke outlet area based on the smoke data and atmospheric pressure of the next time period, and adjust the rotating shutter inside the smoke outlet based on the smoke outlet area when the next time period arrives.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] 1. The present invention collects environmental data of the smoke exhaust port at each time point, as well as the combustion time, material type and amount of the glass kiln; divides time periods based on the environmental data; establishes a smoke prediction module, inputs the combustion time of the next time period, the material type and amount already put into this combustion into the smoke prediction module to obtain the smoke data of the next time period; establishes an atmospheric pressure prediction model, inputs the environmental data of the next time period into the atmospheric pressure prediction model to obtain the atmospheric pressure of the next time period; determines the smoke outlet area based on the smoke data and atmospheric pressure of the next time period, and adjusts the rotary gate inside the smoke outlet based on the smoke outlet area when the next time period is reached. This solves the technical problems in the auxiliary control of the glass kiln that it is difficult to set a control time period that adapts to the kiln exhaust pressure according to the kiln combustion characteristics and environmental factors, and ignores the analysis and prediction of the correlation between smoke emissions and atmospheric pressure, making it difficult to achieve precise kiln pressure control. The present invention upgrades the glass kiln exhaust pressure control from "experience-driven" to "intelligent prediction" through a data-driven prediction model and a dynamic control mechanism, achieving comprehensive optimization in terms of efficiency, environmental protection, safety, cost, etc., and has significant industrial application value.

[0040] 2. The present invention significantly improves the intelligence level and system stability of the dynamic control of the chimney outlet area by introducing a predictive adjustment mechanism based on the flue gas data and atmospheric pressure of the next time period. Traditional kiln pressure adjustment modes mostly rely on passive responses of real-time feedback, which have problems such as response lag and insufficient adjustment accuracy, and easily lead to pressure fluctuations inside the kiln, thereby affecting combustion efficiency and equipment operation safety. The present invention constructs a multi-dimensional data correlation model of flue gas flow, flue gas pressure, flue gas density and atmospheric pressure, and combines mathematical formulas to realize forward-looking calculation of the chimney outlet area, upgrading the adjustment strategy from "passive response" to "active prediction", fundamentally solving the limitations of traditional control methods; the core value of the present invention lies in the use of data-driven predictive models to identify the dynamic change trends of flue gas parameters and atmospheric pressure in advance, thereby completing the precise adjustment of the chimney outlet area before the flue gas state fluctuates significantly, effectively avoiding the risk of kiln pressure imbalance caused by adjustment delays. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0042] Figure 1 Schematic diagram of the operating steps of the present invention;

[0043] Figure 2A schematic diagram of the operation steps for dividing time periods of the present invention;

[0044] Figure 3 Schematic diagram of the system module of the present invention. DETAILED DESCRIPTION

[0045] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0046] See also Figure 1 The first embodiment of the present invention provides an artificial intelligence-based electronic display glass furnace auxiliary control method, comprising:

[0047] Collect environmental data at each time point on the smoke exhaust vent location, as well as the combustion duration, material type, and amount of the glass furnace; divide the time period based on the environmental data; the environmental data includes temperature, humidity, and altitude;

[0048] A flue gas prediction module is established. The combustion duration of the next time period, the type and amount of materials already used in this combustion are input into the flue gas prediction module to obtain the flue gas data for the next time period. The flue gas data includes the flue gas flow rate, flue gas pressure, and flue gas density at the smoke outlet. The flue gas prediction module is trained through an artificial intelligence model.

[0049] Establish an atmospheric pressure prediction model, input the environmental data of the next time period into the atmospheric pressure prediction model to obtain the atmospheric pressure of the next time period; wherein the atmospheric pressure prediction model is obtained through artificial intelligence model training;

[0050] The smoke outlet area is determined based on the smoke data and atmospheric pressure of the next time period, and the rotary shutter inside the smoke outlet is adjusted based on the smoke outlet area when the next time period arrives.

[0051] It should be noted that the daily output of electronic display glass kilns is usually only a few tons or tens of tons, and its ability to resist the adverse effects of external meteorological changes is relatively weak. When the external atmospheric pressure is relatively stable, stable control of the kiln pressure can be achieved by adjusting the kiln pressure control gate opening or the induced draft fan frequency. However, under relatively severe meteorological conditions, the flue gas emission resistance fluctuates frequently due to changes in atmospheric pressure. For electronic display glass kilns with smaller tonnage, it is difficult to achieve stable control of the kiln pressure by the above means; therefore, when meteorological conditions change, the present invention compensates by adjusting the opening of the rotating gate inside the smoke exhaust port to change the chimney outlet area. Without changing the kiln pressure gate opening and the induced draft fan frequency, the exhaust flue gas flow rate is guaranteed to be stable, thereby achieving stable control of the kiln pressure.

[0052] This application collects environmental data on the location of the smoke exhaust outlet at various time points, as well as the combustion duration, material type, and amount of the glass furnace, including:

[0053] The temperature and humidity at the smoke exhaust outlet at each time point are obtained through the weather forecast platform, and the altitude of the smoke exhaust outlet is extracted from the database. The time difference between each time period and the start time of combustion is obtained through the timer, and the time difference is marked as the combustion duration; the type of material is obtained through the ingredient information transmitted from the raw material workshop, and the amount of each type of material manually designed in each time period is extracted from the database; among them, the types of materials include fuel, clarifier, colorant and decolorizer.

[0054] See also Figure 2 In this application, time periods are divided based on environmental data, including:

[0055] At the target time point of the previous day, the temperature and humidity of the smoke exhaust outlet location at each time point today are extracted. Consecutive time points with the same temperature and humidity are connected and integrated into the first period to be allocated. Time points not integrated into the first period to be allocated are integrated into several second periods to be allocated. The target time point is manually set.

[0056] Obtain the current allocated time range of the glass furnace, and divide the first time periods to be allocated into several time periods according to the middle value FZ of the allocated time range;

[0057] The second time period to be allocated is extracted in sequence, and the variance WDC of each temperature and the variance SDC of each humidity in the second time period to be allocated are obtained. The dynamic duration DC is obtained based on the formula DC = FZ×α / exp(β×WDC / BWD+(1-β)×SDC / BSD); when the dynamic duration DC is greater than the maximum value FD of the allocation duration range, the dynamic duration DC is updated using the maximum value FD; when the dynamic duration DC is less than the minimum value FX of the allocation duration range, the dynamic duration DC is updated using the minimum value FX; the second time period to be allocated is divided into several time periods according to the dynamic duration DC; where BWD is the manually set standard deviation of temperature, and BSD is the manually set standard deviation of humidity; α is the amplitude adjustment coefficient of the 1 / exp() function set according to the amplitude of the air pressure fluctuation on that day, and the value range of α is (0,1]; β is the proportional adjustment coefficient manually set according to the importance of temperature and humidity to the atmospheric pressure change, and the value range of β is (0,1].

[0058] Notably, this invention achieves intelligent and precise control of the glass furnace's flue gas emissions by constructing a smoke exhaust time division method dynamically adjusted based on environmental data, ensuring accurate control of furnace pressure. The comprehensive benefits are reflected in the deep integration of multi-dimensional technical optimization and practical application value. This method breaks through the limitations of the traditional fixed time division model. By introducing a dynamic response mechanism for real-time environmental parameters, it couples and analyzes key meteorological factors such as temperature and humidity with the furnace's operating characteristics, forming a time division system with adaptive capabilities.

[0059] Specifically, the system first acquires historical environmental data at manually set target time points. Using continuous temperature and humidity feature recognition technology, stable periods are divided into the first allocation period, while periods of significant fluctuation are extracted as the second allocation period. This hierarchical processing strategy not only ensures the stability of the basic period but also provides targeted targets for dynamic adjustment. During the period division stage, the present invention innovatively uses the median value FZ of the allocation duration range as a benchmark, combining the temperature standard deviation BWD and the humidity standard deviation BSD to construct a dynamic duration calculation model. By introducing the exponential function adjustment formula DC = FZ × α / exp(β × WDC / BWD + (1-β) × SDC / BSD), a quantitative assessment of the intensity of temperature and humidity fluctuations is achieved. In this formula, the α coefficient is dynamically adjusted based on the amplitude of the daily air pressure fluctuation, while the β coefficient is manually set to reflect the relative influence of temperature and humidity on atmospheric pressure changes. This parameter design allows the system to maintain the universality of the basic algorithm while also allowing for personalized configuration for specific working conditions. When the dynamic duration DC exceeds the preset range of [FX, FD], the system ensures the engineering feasibility of the division result through the maximum value truncation and minimum value replacement mechanism. This fault-tolerant design effectively avoids the impact of extreme environmental data on system stability. The implementation of the present invention not only significantly improves the scientific nature of the smoke exhaust period division, so that the period length is accurately matched with the environmental change trend, but also reduces the risk of equipment loss due to sudden changes in temperature and humidity through a dynamic adjustment mechanism, while optimizing the continuity and uniformity of flue gas emissions, providing reliable technical support for energy conservation and consumption reduction and safe production of glass kilns. In addition, the solution of the present invention also has good scalability, and its core algorithm framework can be migrated to other industrial scenarios that require dynamic response to environmental parameters, such as operation period optimization of metallurgical furnaces, chemical reactors and other equipment, and has broad application prospects and significant economic benefits.

[0060] It should be noted that the target time point can be set to 23:59:00 of the previous day.

[0061] It should be noted that the value of α is proportional to the amplitude of the air pressure fluctuation on that day. α is used to adjust the degree of influence of the temperature variance WDC and the humidity variance SDC on the dynamic duration DC. When other conditions remain unchanged, the larger α is, the greater the impact on the dynamic duration DC is, and the smaller α is, the smaller the impact on the dynamic duration DC is.

[0062] It should be noted that the value of β is proportional to the importance of temperature and inversely proportional to the importance of humidity.

[0063] In this application, the allocated time range of the current glass furnace is obtained, including:

[0064] Extract the time from the start of combustion in the current glass furnace to the appearance of smoke at the smoke outlet from the database, and subtract the material heating time from the time to obtain the smoke exhaust time for each combustion; wherein the material heating time is manually obtained based on the initial temperature of the glass furnace;

[0065] Obtain the variance of several smoke discharge durations and determine whether the variance is less than a variance threshold. If so, extract the maximum and minimum values ​​from the several smoke discharge durations, and mark the duration range between the maximum and minimum values ​​as the allocated duration range. If not, remove the smoke discharge duration with the largest difference from the mode of the smoke discharge durations from the several smoke discharge durations, and re-determine the variance until the variance is less than the variance threshold. Then, retain the remaining smoke discharge durations, and mark the duration range between the maximum and minimum values ​​among the remaining smoke discharge durations as the allocated duration range. The variance threshold is obtained empirically.

[0066] It should be noted that electronic display glass kilns are generally small kilns, and the smoke exhaust time is generally more than ten seconds; for example: for an electronic display glass kiln with a daily melting capacity of 50 tons, the smoke exhaust time is 15 seconds after passing through the treatment system.

[0067] In this application, the first time periods to be allocated are divided into several time periods according to the middle value FZ of the allocation duration range, including:

[0068] Extract the first time period to be allocated in sequence, and obtain the duration YC of the first time period to be allocated, based on the formula The planned number of times GC of the first time period to be allocated is obtained, and the first time period to be allocated is evenly divided into a plurality of time periods based on the planned number of times GC.

[0069] It should be noted that the specific steps of dividing the second time period to be allocated into several time periods according to the dynamic duration DC are as follows:

[0070] Extract the second time period to be allocated in sequence, and obtain the duration RC of the second time period to be allocated, based on the formula The planned number HC of the second time period to be allocated is obtained, and the second time period to be allocated is evenly divided into a plurality of time periods based on the planned number HC.

[0071] The smoke prediction module in this application is obtained through artificial intelligence model training, including:

[0072] Extracting the combustion duration, the type and amount of materials already put into the combustion in several time periods from the historical flue gas database, and the corresponding flue gas data for the next time period; wherein the historical flue gas database includes the combustion duration, the type and amount of materials already put into the combustion in several time periods, and the flue gas data for the next time period manually set based on the combustion duration, the type and amount of materials already put into the combustion in several time periods;

[0073] The combustion duration, material type and amount in several time periods, as well as the corresponding flue gas data of the next time period are integrated into several groups of training data and verification data; the training data is used to train the artificial intelligence model, and the test data is used to test the trained artificial intelligence model, and the artificial intelligence model is adjusted according to the test results; finally, the input is the combustion duration of the next time period, the type and amount of materials already put into this combustion, and the output is the flue gas data of the next time period, and the flue gas prediction module is obtained; among them, the artificial intelligence model includes the BP neural network model and the RBF neural network model.

[0074] It should be noted that the flue gas data includes the flue gas flow rate, flue gas pressure and flue gas density at the smoke outlet. The flue gas data will be affected by the combustion time, material type and amount added. The flue gas flow rate is proportional to the combustion time and amount added, the flue gas pressure is proportional to the combustion time and amount added, and the flue gas density is related to the material type. If the material type of equal volume contains more high calorific value materials, such as natural gas, more CO2 will be generated after combustion, and the flue gas density will be higher; if the material type of equal volume contains more low calorific value materials, such as hydrogen, more water vapor may be produced, and the flue gas density will be lower.

[0075] Specifically, the trained artificial intelligence model is tested using test data, and the specific steps for adjusting the artificial intelligence model based on the test results are as follows:

[0076] The combustion time, the type and amount of materials already put into the combustion in the test data are input into the trained artificial intelligence model to obtain the corresponding flue gas data of the next time period, and the corresponding flue gas data are compared with the corresponding flue gas data in the test data. When the difference between each data in the flue gas data is within the corresponding threshold, which is obtained based on experience, there is no need to adjust the parameters, and the next set of test data is tested; if the difference between each data in the flue gas data is not within the corresponding threshold, the corresponding parameters are adjusted until the difference between each data in the flue gas data of the corresponding test data is within the corresponding threshold, and then the next set of test data is tested. When the number of test data with the difference between each data within the threshold in the output flue gas data obtained from all test data accounts for 90% or more of the total test data, the input is the combustion time of the next time period, the type and amount of materials already put into this combustion, and the output is the flue gas data of the next time period. The flue gas prediction module is obtained.

[0077] The atmospheric pressure prediction model in this application is obtained through artificial intelligence model training, including:

[0078] Extracting environmental data and atmospheric pressure for each time period from a historical atmospheric pressure database; wherein the historical atmospheric pressure database includes environmental data and monitored atmospheric pressure;

[0079] Integrate environmental data and atmospheric pressure into several groups of training data and test data; use the training data to train the artificial intelligence model, use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally, obtain an atmospheric pressure prediction model whose input is the environmental data of the next time period and output is the atmospheric pressure of the next time period; wherein, the artificial intelligence model includes a BP neural network model and an RBF neural network model.

[0080] It should be noted that changes in atmospheric pressure are related to changes in temperature, humidity, and altitude. Since the altitude of the smoke exhaust outlet is fixed, and the higher the altitude, the lower the atmospheric pressure. Therefore, changes in atmospheric pressure are mainly analyzed by temperature and humidity, and air pressure is inversely proportional to temperature and humidity. The principle is as follows:

[0081] 1. The increase in temperature increases the distance between air molecules, reduces the density, and causes the air pressure to drop;

[0082] 2. The molar mass of water vapor is less than the average molar mass of dry air. Therefore, for the same volume, an increase in humidity will cause the air density to decrease, and thus the air pressure to drop.

[0083] Specifically, the trained artificial intelligence model is tested using test data, and the specific steps for adjusting the artificial intelligence model based on the test results are as follows:

[0084] The environmental data in the test data is input into the trained artificial intelligence model to obtain the corresponding atmospheric pressure, and the corresponding atmospheric pressure is compared with the corresponding monitored atmospheric pressure in the test data. When the difference between the two is within the threshold, and the threshold is obtained based on experience, there is no need to adjust the parameters, and the next set of test data is tested; if it is not within the threshold, the corresponding parameters are adjusted until the difference between the atmospheric pressure of the corresponding test data and the corresponding monitored atmospheric pressure in the test data is within the threshold, and then the next set of test data is tested. When the number of test data whose output atmospheric pressure difference obtained from all test data is within the threshold accounts for 90% or more of the total test data, an atmospheric pressure prediction model is obtained with the input being the environmental data of the next time period and the output being the atmospheric pressure of the next time period.

[0085] In this application, the smoke outlet area is determined based on the smoke data and atmospheric pressure of the next time period, including:

[0086] Extract the flue gas flow rate YQ, flue gas pressure YP, flue gas density YD and atmospheric pressure DP from the environmental data of the next time period;

[0087] The chimney outlet area YS for the next time period is determined based on the flue gas flow rate YQ, flue gas pressure YP, flue gas density YD and atmospheric pressure DP. The chimney outlet area YS satisfies the following formula:

[0088]

[0089] It is worth noting that the present invention significantly improves the intelligence level and system stability of the dynamic control of the chimney outlet area by introducing a predictive adjustment mechanism based on the flue gas data and atmospheric pressure in the next time period. The traditional kiln pressure adjustment mode mostly relies on passive response of real-time feedback, which has a response lag problem and easily leads to pressure fluctuations inside the kiln, thereby affecting the combustion efficiency and equipment operation safety. The present invention constructs a multi-dimensional data correlation model of flue gas flow, flue gas pressure, flue gas density and atmospheric pressure, and combines the mathematical formula By realizing forward-looking calculation of the chimney outlet area, the regulation strategy is upgraded from "passive response" to "active prediction", which fundamentally solves the limitations of traditional control methods. The core value of this invention lies in the use of a data-driven prediction model to identify the dynamic change trends of flue gas parameters and atmospheric pressure in advance, thereby completing the precise adjustment of the chimney outlet area before the flue gas state fluctuates significantly, effectively avoiding the risk of kiln pressure imbalance caused by adjustment delays.

[0090] Furthermore, the present invention realizes the dynamic optimization configuration of the chimney outlet area by quantitatively analyzing the relationship between the flue gas flow rate and the pressure difference. For example, when the flue gas flow rate YQ increases, the denominator in the formula is adjusted due to the change in the difference between the flue gas pressure YP and the atmospheric pressure DP, ensuring that the calculation result of YS can match the fluid mechanics characteristics under the current working conditions, thereby maintaining the stability and efficiency of flue gas emissions. At the same time, the introduction of flue gas density YD enhances the adaptability of the model to different working conditions, especially in high temperature, high humidity or complex gas composition environments, which can more accurately reflect the actual flow resistance and avoid calculation deviations caused by parameter simplification. This mathematical modeling method based on physical laws not only improves the scientific nature of the control logic, but also provides reliable technical support for the refined management of industrial kilns.

[0091] It is worth noting that the implementation of the present invention is of great significance to improving the stable operation of the kiln system; by adjusting the chimney outlet area in advance, the fluctuation amplitude of the negative pressure inside the kiln can be effectively reduced, and the risk of fuel waste, excessive emissions or equipment damage caused by unstable pressure can be reduced, while extending the service life of key components. In practical applications, the present invention can also be integrated with the intelligent monitoring system to form a closed-loop control chain of "prediction-regulation-feedback" to further enhance the adaptive ability of the system. This active adjustment strategy is not only applicable to traditional industrial kilns, but can also be extended to high-temperature processing scenarios such as waste incineration and metallurgical smelting, providing innovative solutions for industrial energy conservation, emission reduction and safe operation. In summary, the present invention has created a new paradigm for kiln pressure control through the deep integration of data prediction and fluid mechanics principles, which has significant engineering value and promotion prospects.

[0092] In this application, when the next time period arrives, the rotary gate inside the smoke outlet is adjusted based on the area of ​​the smoke outlet, including:

[0093] When the time reaches the next time period, the chimney outlet area YS of the next time period is extracted, and the rotary gate in the smoke outlet is rotated according to the chimney outlet area YS.

[0094] It should be noted that if the chimney outlet area DYS in the current time period corresponds to a rotary gate opening DM, and the chimney outlet area YS in the next time period corresponds to a rotary gate opening M; then, the rotary gate opening ΔM that needs to be adjusted to ensure kiln pressure stability is ΔM = M-DM.

[0095] In another embodiment, the auxiliary control can also control the servo motor through the kiln pressure auxiliary control cabinet to drive the rotating flue to turn along the slide rail according to the wind direction feedback from the meteorological instrument, so that the direction of the auxiliary smoke exhaust port is consistent with the wind direction, thereby reducing the resistance to smoke emission.

[0096] See also Figure 3 , the second aspect of the present invention provides an electronic display glass furnace auxiliary control system based on artificial intelligence, including: a model prediction module, and a data collection module, an auxiliary control module and a database connected thereto;

[0097] Data collection module: used to collect environmental data of the smoke exhaust outlet at various time points, as well as the combustion duration, material type, and amount of the glass furnace; divide the time period based on the environmental data; the environmental data includes temperature, humidity, and altitude;

[0098] Model prediction module: used to establish a flue gas prediction module, input the combustion duration of the next time period, the type and amount of materials already put into this combustion into the flue gas prediction module to obtain the flue gas data for the next time period; establish an atmospheric pressure prediction model, input the environmental data of the next time period into the atmospheric pressure prediction model to obtain the atmospheric pressure for the next time period; the flue gas data includes the flue gas flow rate, flue gas pressure and flue gas density at the smoke outlet; the flue gas prediction module is obtained through artificial intelligence model training; the atmospheric pressure prediction model is also obtained through artificial intelligence model training;

[0099] Auxiliary control module: used to determine the smoke outlet area based on the smoke data and atmospheric pressure of the next time period, and adjust the rotating gate inside the smoke outlet based on the smoke outlet area when the next time period arrives;

[0100] Database: used to store data.

[0101] Some of the data in the above formula are calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0102] Working principle of the present invention:

[0103] The present invention first collects the environmental data of the smoke exhaust port at each time point, as well as the combustion time, material type and amount of the glass kiln; divides the time period based on the environmental data; then establishes a smoke prediction module, inputs the combustion time of the next time period, the material type and amount already put into this combustion into the smoke prediction module to obtain the smoke data of the next time period; establishes an atmospheric pressure prediction model, inputs the environmental data of the next time period into the atmospheric pressure prediction model to obtain the atmospheric pressure of the next time period; finally, determines the smoke outlet area based on the smoke data and atmospheric pressure of the next time period, and adjusts the rotating gate inside the smoke exhaust port based on the smoke outlet area when the next time period is reached.

[0104] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An artificial intelligence-based electronic display glass furnace auxiliary control method, characterized in that: include: Collect environmental data at each time point on the smoke exhaust vent location, as well as the combustion duration, material type, and amount of the glass furnace; divide the time period based on the environmental data; the environmental data includes temperature, humidity, and altitude; A flue gas prediction module is established. The combustion duration of the next time period, the type and amount of materials already used in this combustion are input into the flue gas prediction module to obtain the flue gas data for the next time period. The flue gas data includes the flue gas flow rate, flue gas pressure, and flue gas density at the smoke outlet. The flue gas prediction module is trained through an artificial intelligence model. Establish an atmospheric pressure prediction model, input the environmental data of the next time period into the atmospheric pressure prediction model to obtain the atmospheric pressure of the next time period; wherein the atmospheric pressure prediction model is obtained through artificial intelligence model training; The smoke outlet area is determined based on the smoke data and atmospheric pressure of the next time period, and the rotary shutter inside the smoke outlet is adjusted based on the smoke outlet area when the next time period arrives.

2. The artificial intelligence-based electronic display glass furnace auxiliary control method according to claim 1 is characterized in that: The environmental data collected at each time point of the smoke exhaust port location, as well as the burning time of the glass furnace, material type and amount of material added, include: The temperature and humidity at the smoke exhaust outlet at each time point are obtained through the weather forecast platform, and the altitude of the smoke exhaust outlet is extracted from the database. The time difference between each time period and the start time of combustion is obtained through the timer, and the time difference is marked as the combustion duration; the type of material is obtained through the ingredient information transmitted from the raw material workshop, and the amount of each type of material manually designed in each time period is extracted from the database; wherein the types of materials include fuel, clarifier, colorant and decolorizer.

3. The artificial intelligence-based electronic display glass furnace auxiliary control method according to claim 1, characterized in that: The dividing of time periods based on environmental data includes: At the target time point of the previous day, the temperature and humidity of the smoke exhaust outlet location at each time point today are extracted. Consecutive time points with the same temperature and humidity are connected and integrated into the first period to be allocated. Time points not integrated into the first period to be allocated are integrated into several second periods to be allocated. The target time point is manually set. Obtain the current allocated time range of the glass furnace, and divide the first time periods to be allocated into several time periods according to the middle value FZ of the allocated time range; Extract the second time period to be allocated in sequence, obtain the variance WDC of each temperature and the variance SDC of each humidity in the second time period to be allocated, and obtain the dynamic duration DC based on the formula DC = FZ×α / exp(β×WDC / BWD+(1-β)×SDC / BSD); when the dynamic duration DC is greater than the maximum value FD of the allocation duration range, use the maximum value FD to update the dynamic duration DC; when the dynamic duration DC is less than the minimum value FX of the allocation duration range, use the minimum value FX to update the dynamic duration DC; divide the second time period to be allocated into several time periods according to the dynamic duration DC; wherein BWD is the manually set temperature standard deviation, and BSD is the manually set humidity standard deviation; α is the amplitude adjustment coefficient of the 1 / exp() function set according to the air pressure fluctuation amplitude of the day, and the value range of α is (0,1]; β is the proportional adjustment coefficient manually set according to the importance of temperature and humidity to atmospheric pressure changes, and the value range of β is (0,1].

4. The artificial intelligence-based electronic display glass furnace auxiliary control method according to claim 3 is characterized in that: The method of obtaining the allocated time range of the current glass furnace includes: Extract the time from the start of combustion in the current glass furnace to the appearance of smoke at the smoke outlet from the database, and subtract the material heating time from the time to obtain the smoke exhaust time for each combustion; wherein the material heating time is manually obtained based on the initial temperature of the glass furnace; Obtain the variance of several smoke discharge durations, and determine whether the variance is less than a variance threshold. If so, extract the maximum and minimum values ​​from the several smoke discharge durations, and mark the duration range between the maximum and minimum values ​​as the allocated duration range. If not, remove the smoke discharge duration that has the largest difference from the mode of the smoke discharge durations from the several smoke discharge durations, and re-determine the variance until the variance is less than the variance threshold. Then, retain the remaining smoke discharge durations, and mark the duration range between the maximum and minimum values ​​among the remaining smoke discharge durations as the allocated duration range. The variance threshold is obtained empirically.

5. The artificial intelligence-based electronic display glass furnace auxiliary control method according to claim 1, characterized in that: The first time periods to be allocated are divided into a plurality of time periods according to the middle value FZ of the allocation duration range, including: Extract the first time period to be allocated in sequence, obtain the duration YC of the first time period to be allocated, and use the formula A planned number of times GC of the first time period to be allocated is obtained, and the first time period to be allocated is evenly divided into a plurality of time periods based on the planned number of times GC.

6. The artificial intelligence-based electronic display glass furnace auxiliary control method according to claim 1, characterized in that: The flue gas prediction module is obtained through artificial intelligence model training, including: Extracting the combustion duration, the type and amount of materials already put into the combustion in several time periods from the historical flue gas database, and the corresponding flue gas data for the next time period; wherein the historical flue gas database includes the combustion duration, the type and amount of materials already put into the combustion in several time periods, and the flue gas data for the next time period manually set based on the combustion duration, the type and amount of materials already put into the combustion in several time periods; The combustion duration, material type and amount in several time periods, as well as the corresponding flue gas data of the next time period are integrated into several groups of training data and verification data; the training data is used to train the artificial intelligence model, and the test data is used to test the trained artificial intelligence model, and the artificial intelligence model is adjusted according to the test results; finally, the input is the combustion duration of the next time period, the type and amount of materials already put into this combustion, and the output is the flue gas data of the next time period, and the flue gas prediction module is obtained; among them, the artificial intelligence model includes the BP neural network model and the RBF neural network model.

7. The artificial intelligence-based electronic display glass furnace auxiliary control method according to claim 1, characterized in that: The atmospheric pressure prediction model is obtained through artificial intelligence model training, including: Extracting environmental data and atmospheric pressure for each time period from a historical atmospheric pressure database; wherein the historical atmospheric pressure database includes environmental data and monitored atmospheric pressure; Integrate environmental data and atmospheric pressure into several groups of training data and test data; use the training data to train the artificial intelligence model, use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally, obtain an atmospheric pressure prediction model whose input is the environmental data of the next time period and output is the atmospheric pressure of the next time period; wherein, the artificial intelligence model includes a BP neural network model and an RBF neural network model.

8. The artificial intelligence-based electronic display glass furnace auxiliary control method according to claim 1, characterized in that: The determining of the smoke outlet area based on the smoke data and the atmospheric pressure in the next time period includes: Extract the flue gas flow rate YQ, flue gas pressure YP, flue gas density YD and atmospheric pressure DP from the environmental data of the next time period; The chimney outlet area YS for the next time period is determined based on the flue gas flow rate YQ, flue gas pressure YP, flue gas density YD and atmospheric pressure DP. The chimney outlet area YS satisfies the following formula:

9. The artificial intelligence-based electronic display glass furnace auxiliary control method according to claim 1, characterized in that: The method of adjusting the rotating shutter inside the smoke outlet based on the area of ​​the smoke outlet when the next time period arrives includes: When the time reaches the next time period, the chimney outlet area YS of the next time period is extracted, and the rotary gate in the smoke outlet is rotated according to the chimney outlet area YS.

10. An artificial intelligence-based electronic display glass furnace auxiliary control system, used to run the artificial intelligence-based electronic display glass furnace auxiliary control method according to any one of claims 1 to 9, characterized in that: include: Model prediction module, and the data collection module and auxiliary control module connected thereto; The data collection module is used to collect environmental data of the smoke exhaust outlet at each time point, as well as the combustion time of the glass furnace, the type of material and the amount of material put in; and divide the time period based on the environmental data; wherein the environmental data includes temperature, humidity and altitude; The model prediction module is used to establish a flue gas prediction module, input the combustion duration of the next time period, the type and amount of materials already put into this combustion into the flue gas prediction module to obtain the flue gas data of the next time period; establish an atmospheric pressure prediction model, input the environmental data of the next time period into the atmospheric pressure prediction model to obtain the atmospheric pressure of the next time period; wherein the flue gas data includes the flue gas flow rate, flue gas pressure and flue gas density at the smoke outlet; the flue gas prediction module is obtained through artificial intelligence model training; the atmospheric pressure prediction model is obtained through artificial intelligence model training; The auxiliary control module is used to determine the smoke outlet area based on the smoke data and atmospheric pressure of the next time period, and adjust the rotating shutter inside the smoke outlet based on the smoke outlet area when the next time period arrives.

Images