Intelligent circulating fluidized bed method desulfurization and carbon-pollutant deep purification system

Abstract

This invention relates to the field of energy conservation and environmental protection, specifically disclosing an intelligent circulating fluidized bed desulfurization and synergistic carbon-pollution deep purification system. Through a modeling method driven by knowledge and data synergy in the multi-pollutant generation-removal process, a highly reliable and interpretable model for accurately predicting the concentration of multiple pollutants is constructed. This model enables prediction of pollutant concentration changes 90 seconds in advance, with an average relative percentage error of less than 2%, overcoming the problems of control lag and difficulty in real-time accurate optimization caused by data measurement delays. An intelligent control technology is established, featuring a dual-layer structure coupling of multi-parameter synergistic optimization of hydrated lime feeder frequency, bed pressure, and water spray volume, along with model predictive control. This achieves stable ultra-low emissions throughout the entire period, while reducing outlet pollutant concentration fluctuations by more than 50%, outlet temperature fluctuations within ±2℃, bed pressure drop fluctuations within ±75Pa, hydrated lime consumption by more than 10%, water consumption by more than 5%, and CO2 emissions by more than 6%. This solves the problem of efficient and stable removal and synergistic carbon reduction of multiple pollutants in flue gas under complex operating conditions.

Description

Technical Field

[0001] This invention belongs to the field of energy conservation and environmental protection, specifically involving an intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system. Background Technology

[0002] Currently, circulating fluidized bed (CFB) flue gas desulfurization (FGD) technology has been applied on a large scale and industrially in key industries such as coal-fired power plants, industrial boilers, steel sintering machines, and sludge incineration. CFBD FGD combined with deep carbon-pollution purification is a complex industrial process characterized by multiple inputs and outputs, strong coupling, strong nonlinearity, variable parameters, multiple operating conditions, and a wide range of variable loads. Achieving intelligent control of the CFBD FGD system with deep carbon-pollution purification faces numerous challenges, primarily as follows: CFBD FGD involves a complex process of flow-mass transfer-reaction; the removal mechanisms of different pollutants vary greatly; and there is a time delay in the measurement of key operating parameters, leading to significant lag in optimization and control based on instrument data, making it difficult to support real-time and accurate optimization and control. Therefore, how to construct a predictive model for key parameters of the CFBD FGD process with efficient carbon-pollution purification and accurately quantitatively describe the operating characteristics of the CFBD FGD system is a problem that urgently needs to be solved. The operation and control of circulating fluidized bed desulfurization and synergistic carbon-pollution high-efficiency purification systems are often dynamic. Frequent load changes (20%–110%) and frequent fluctuations in coal / ore quality (fluctuations in sulfur and ash content, and fluctuations in coal / ore blending ratios) can cause changes in system operating characteristics. Current operation methods based on manual experience and traditional PID control are insufficient for real-time, precise adjustment and optimization of pollutant removal devices when operating conditions change. This leads to problems such as large fluctuations in pollutant emission concentrations, poor control performance of pollutant removal devices, high material and energy consumption of pollutant removal devices, and bed collapse. Therefore, addressing the challenges of flue gas treatment volume and pollutant concentration fluctuations caused by varying loads, parameters, and fuel / raw materials, and how to achieve operational optimization and intelligent control of "multiple devices and multiple pollutants," moving from "compliance control" to "high-reliability edge control," and significantly reducing the fluctuation range of system outlet pollutant emission concentrations while lowering operating energy and material consumption, is a pressing issue that needs to be overcome.

[0003] The adiabatic saturation temperature difference is the difference between the flue gas temperature at the desulfurization tower outlet and the adiabatic saturation temperature of the flue gas. This difference plays a crucial role in the desulfurization efficiency and stable operation of the desulfurization unit. On the one hand, a lower adiabatic saturation temperature difference results in higher humidity within the desulfurization tower, which is more conducive to the desulfurization reaction and improves desulfurization efficiency. On the other hand, an excessively low adiabatic saturation temperature difference can cause condensation in the flue gas, corroding the inner walls and the dust removal equipment. Therefore, controlling the adiabatic saturation temperature difference during operation is key to improving the performance of the flue gas desulfurization process.

[0004] The purpose of flue gas temperature control is to regulate the adiabatic saturation temperature difference by adjusting the water spray volume at the bottom of the desulfurization tower, thereby ensuring the economical and safe operation of the desulfurization system. The water spray volume can be adjusted using simple PID control. The key to flue gas temperature control is the online real-time measurement of the adiabatic saturation temperature difference. However, in actual desulfurization processes, the adiabatic saturation temperature difference cannot be measured online and must be calculated based on its definition. Calculating the adiabatic saturation temperature difference requires measuring both the flue gas temperature at the desulfurization tower outlet and the flue gas adiabatic saturation temperature. While the flue gas outlet temperature can be measured online, the key issue is the lack of equipment for directly measuring the adiabatic saturation temperature in practice. Therefore, indirect methods are generally used to obtain the adiabatic saturation temperature.

[0005] To address the aforementioned challenges, this invention proposes an intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system. Through a modeling method driven by knowledge and data synergistically covering the generation and removal processes of multiple pollutants, combined with multi-dimensional analysis of the dynamic correlation between key parameters and pollutant concentrations, a highly reliable and interpretable multi-section model for accurately predicting the concentrations of multiple pollutants such as SO2 / SO3 / CO2 / HCl is constructed. This overcomes the problems of control lag caused by data measurement delays and the difficulty in real-time accurate optimization. Furthermore, an intelligent control technology for the pollutant removal device, featuring a dual-layer structure coupling of multi-parameter synergistic optimization and model predictive control, is invented. This technology ensures stable and ultra-low emissions of multiple pollutants throughout the entire time period while reducing the fluctuation range of outlet pollutant concentration and flue gas temperature. Simultaneously, it achieves precise control of bed pressure drop and temperature, further reducing the consumption of quicklime and process water, and decreasing CO2 emissions. This solves the problem of efficient and stable removal and synergistic carbon reduction of multiple pollutants in flue gas under complex operating conditions. Summary of the Invention

[0006] This invention proposes an intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system. The system includes a quicklime preparation and conveying system, an ash circulation system, a temperature and humidity control system, a flue gas recirculation system, a pre-dust collector, a bag filter, a desulfurization tower, and an intelligent control system. The key features are: the temperature and humidity control system is equipped with multi-stage humidification and activation spray guns; the quicklime preparation and conveying system is equipped with a variable frequency feeder; the ash circulation system is equipped with an ash circulation variable frequency feeder; the flue gas recirculation system is equipped with a clean flue gas recirculation variable frequency fan; and the intelligent control system includes an online monitoring module, a key parameter prediction module, and a key parameter control module.

[0007] The online monitoring module includes parameters such as total coal feed, total ore feed, coal quality parameters, mineral composition, flue gas temperature at the sintering process outlet, wind box temperature, combustion air flow rate, main exhaust fan flow rate, flue gas oxygen content, SO2 concentration at the desulfurization tower inlet and outlet, flue gas temperature and moisture content of the inlet flue gas, feeder frequency, ash circulation feeder frequency, clean flue gas recirculation variable frequency fan frequency, and the number of multi-stage humidification and activation spray guns in operation.

[0008] The key parameter prediction module includes prediction models for the generation and removal of CO2, SO2, SO3, and HCl concentrations at the inlet and outlet of the desulfurization tower, as well as prediction models for flue gas temperature and pressure at the outlet of the desulfurization tower. The key parameter control module includes a control model for a hydrated lime frequency converter feeder, a bed pressure drop prediction and control model, and a multi-stage humidification and activation spray gun control model.

[0009] The construction of the prediction model for the generation and removal of CO2, SO2, SO3, and HCl concentrations at the inlet and outlet of the desulfurization tower mainly includes the following steps:

[0010] Step S1: Analyze the generation mechanism of CO2, SO2, SO3, and HCl acidic gases during combustion / sintering, and combine data analysis methods to screen the key parameters affecting the concentration of CO2, SO2, SO3, and HCl acidic gases at the inlet of the circulating fluidized bed desulfurization tower;

[0011] Step S2: Compensate for the delay between key variables through data processing, and construct a knowledge and data fusion driven model for the generation concentrations of CO2, SO2, SO3, HCl and HF acidic gases based on an LSTM neural network with autoregressive variables, so as to achieve accurate prediction of the concentrations of CO2, SO2, SO3 and HCl acidic gases at the inlet of the desulfurization unit in advance.

[0012] Step S3: Analyze the adsorption and removal mechanism of CO2, SO2, SO3, and HCl acidic gases in the flue gas circulating fluidized bed desulfurization tower with the absorbent, establish an adsorption and removal mechanism model for acidic gases, further introduce correction parameters into the adsorption and removal mechanism model of CO2, SO2, SO3, and HCl acidic gases, solve it through the PSO algorithm, and construct an adsorption and removal model of CO2, SO2, SO3, and HCl acidic gases based on parameter identification;

[0013] Step S4: Further, an LSTM neural network is used to compensate for the error in the parameter identification model. Combined with actual operating data, a knowledge and data fusion-driven acid gas concentration prediction model with double-layer data correction is established, which realizes the accurate prediction of the concentrations of CO2, SO2, SO3, and HCl acid gases at the outlet of the desulfurization unit.

[0014] As a preferred embodiment, the knowledge and data fusion-driven SO2 concentration prediction model for the desulfurization tower inlet is described as follows:

[0015]

[0016] in: K represents the predicted sulfur dioxide concentration at the inlet of the desulfurization tower. r K represents the SO2 emission factor caused by the operating conditions of the boiler / sintering process. r= f(T, B, L1, L2…); T is the flue gas temperature in the furnace or sintering process, B is the coal feed rate, L1 is the primary air volume, L2 is the secondary air volume, a fh A is the fly ash coefficient. ar To obtain the base ash content, S ar Received basic sulfur content, Q net，ar To receive the base low heat, ω CaO ω MgO , denoted as , where is the mass fraction of the corresponding substance in the ash, n is the hourly coal consumption, and Q is the flue gas volume at the inlet of the desulfurization tower.

[0017] As a preferred embodiment, the knowledge and data fusion-driven SO3 concentration prediction model at the desulfurization tower inlet is described as follows:

[0018]

[0019] in: The values ​​represent the predicted sulfur trioxide concentration at the inlet of the desulfurization tower, T is the flue gas temperature in the furnace or sintering process, B is the coal feed rate, L1 is the primary air volume, and L2 is the secondary air volume. is the predicted sulfur dioxide concentration at the inlet of the desulfurization tower, and Q is the flue gas volume at the inlet of the desulfurization tower.

[0020] As a preferred embodiment, the knowledge and data fusion-driven HCl concentration prediction model for the desulfurization tower inlet is described as follows:

[0021]

[0022] Where: C HCl,in，yc K represents the predicted hydrogen chloride concentration at the inlet of the desulfurization tower. r The HCl emission factor is determined by the boiler / sintering process operating conditions; T is the flue gas temperature in the furnace or sintering process; B is the coal feed rate; L1 is the primary air volume; L2 is the secondary air volume; a fh A is the fly ash coefficient. ar To obtain the base ash content, Q net，ar To receive the base low heat, ω CaO ω MgO , denoted as , where is the mass fraction of the corresponding substance in the ash, n is the hourly coal consumption, and Q is the flue gas volume at the inlet of the desulfurization tower.

[0023] As a preferred embodiment, the knowledge and data fusion-driven CO2 concentration prediction model for the desulfurization tower inlet is described as follows:

[0024]

[0025] in: Let Q be the CO2 volume at the inlet of the desulfurization tower, B be the coal feed rate, and Q be the total CO2 volume. FC is the total air volume entering the combustion process. ar Carbon content (H) of fuel / ore received basis ar For fuel / ore received basic hydrogen content, O ar V represents the basic oxygen content received by fuel / ore. CO β is the carbon volume of carbon monoxide at the inlet of the desulfurization tower, and β is the fuel characteristic coefficient. Here is the predicted sulfur dioxide concentration at the inlet of the desulfurization tower, and Q is the flue gas volume at the inlet of the desulfurization tower.

[0026] Preferably, the prediction model for SO2 concentration at the desulfurization tower outlet based on parameter identification is described as follows:

[0027]

[0028] Where: M Ca(OH)2 Z represents the feed rate of slaked lime. h For the conversion rate of slaked lime, f(t) l , t f …); t f t is the time it takes for the desulfurizing agent to complete its reaction in the absorption tower. l Residence time of desulfurizing agent in the absorption tower; V is the partial pressure of SO2 inside the desulfurization tower; T is the temperature; V is the volume of the desulfurization tower; and a is the air leakage coefficient of the desulfurization tower. Here, Q represents the predicted sulfur dioxide concentration at the inlet of the desulfurization tower, and R is a constant.

[0029] Further based on the above model, t f The complete reaction time of the desulfurizing agent in the absorption tower is t. l There are two parameters that need to be identified regarding the residence time of the desulfurizing agent in the absorption tower. The parameter identification process can be represented by an optimization problem, namely:

[0030]

[0031] In the formula, RMSE(t) l , t f The model calculates the root mean square error between the predicted and actual measured values ​​of SO2 concentration at the outlet.

[0032] The establishment of the prediction model for the outlet flue gas temperature of the desulfurization tower and the adiabatic saturation temperature difference of the desulfurization flue gas mainly includes the following steps:

[0033] Step W1: Analyze the adiabatic cooling and humidification mechanism of flue gas in the flue gas circulating fluidized bed desulfurization process, construct a model for predicting the humidity of flue gas at the desulfurization tower outlet, collect parameters such as the moisture content of flue gas at the desulfurization tower inlet, the flue gas volume at the desulfurization tower inlet, the flue gas temperature at the desulfurization tower inlet, the flue gas pressure at the desulfurization tower inlet, the water spray volume at the desulfurization tower inlet, the flue gas pressure at the desulfurization tower outlet, and the flue gas temperature at the desulfurization tower outlet, and construct a database for predicting the adiabatic saturated temperature difference of desulfurized flue gas.

[0034] Step W2: Combining data analysis methods, screen the key parameters affecting the humidity of the flue gas at the desulfurization tower outlet, further analyze the pure time delay between changes in key variables and the humidity of the flue gas at the desulfurization tower outlet, and compensate for the delay between key variables through data processing.

[0035] Step W3: By introducing correction parameters into the desulfurization tower outlet flue gas humidity mechanism model and solving it using the PSO algorithm, a parameter-identified desulfurization tower outlet flue gas humidity control mechanism model was constructed.

[0036] Step W4: Further construct a desulfurization tower outlet flue gas humidity prediction model based on an LSTM neural network with autoregressive variables, which integrates data and humidity control mechanisms.

[0037] Step W5: Further analyze the adiabatic saturation temperature regulation mechanism of desulfurized flue gas and construct a prediction model for the adiabatic saturation temperature difference of desulfurized flue gas based on the fusion of empirical knowledge and mechanism.

[0038] Step W6: Further combining the flue gas temperature at the desulfurization tower outlet, the flue gas pressure at the desulfurization tower inlet, and the flue gas humidity prediction model at the desulfurization tower outlet, a mechanism-operation data correction fusion desulfurization flue gas adiabatic saturation temperature difference prediction model is constructed, realizing accurate prediction of the adiabatic saturation temperature difference of desulfurization flue gas under variable load conditions.

[0039] Preferably, the desulfurization tower outlet flue gas humidity control mechanism model identified by the parameters can be expressed by the following formula:

[0040] H=λ*(G*(1-δ) / (Q*ρ)+H i )

[0041] Where G is the amount of water injected into the desulfurization tower, Q is the volumetric flow rate of the flue gas, ρ is the density of the flue gas, and H is the volumetric flow rate of the flue gas. i δ represents the humidity of the flue gas at the inlet of the desulfurization tower, δ represents the coefficient of residual moisture in the desulfurization ash in the total spray volume, and λ represents the correction coefficient for the humidity of the flue gas at the outlet of the desulfurization tower.

[0042] Furthermore, parameters δ and λ in the empirical model of flue gas humidity at the desulfurization tower outlet are identified. The parameter identification process can be represented by an optimization problem, namely:

[0043]

[0044] Wherein, RMSE(δ, λ) represents the root mean square error between the predicted and actual values ​​of the flue gas humidity at the desulfurization tower outlet, as determined by the empirical model.

[0045] Preferably, the adiabatic saturation temperature difference of the desulfurized flue gas can be expressed by the following formula:

[0046] ΔT s =TT s

[0047] Where, ΔT s T represents the adiabatic saturation temperature difference of the desulfurization flue gas, and T represents the outlet flue gas temperature of the desulfurization tower. s It is the adiabatic saturation temperature;

[0048] Furthermore, the prediction model for the adiabatic saturated temperature difference of desulfurized flue gas based on the fusion of empirical knowledge and mechanisms can be expressed by the following formula:

[0049]

[0050]

[0051] Where C is the specific heat of the moist flue gas, M is the molecular weight of the flue gas, and T is the outlet temperature of the desulfurization tower. s Let T be the adiabatic saturation temperature, H be the humidity of the flue gas at temperature T, and P be the temperature at temperature P. f This refers to the flue gas pressure.

[0052] The multi-stage humidification module is not only the module with the highest water consumption in the desulfurization unit, but also the most important subsystem for ensuring the desulfurization outlet flue gas temperature and the adiabatic saturation temperature difference of the desulfurization flue gas. Therefore, when constructing the intelligent control problem of the multi-stage humidification module of the flue gas desulfurization unit, the operating water consumption, the desulfurization outlet flue gas temperature, and the adiabatic saturation temperature difference of the desulfurization flue gas should be considered simultaneously. Thus, the intelligent control process can also be described by the following optimization problem:

[0053] minWater(L1,L2,L3,L4=L1p1+L2p2+L3p3+L4p4

[0054]

[0055] L1, L2, L3, and L4 represent the operating states of the humidification nozzles at levels 1 to 4. When the humidification nozzle is on, the operating state is 1; when the humidification nozzle is off, the operating state is 0. p1, p2, p3, and p4 represent the water flow rates corresponding to the humidification nozzles. This is the predicted value for the outlet flue gas temperature. This is the predicted value of the adiabatic saturation temperature difference of the desulfurization flue gas.

[0056] The hydrated lime optimization control module calculates the predicted values ​​of sulfur dioxide, sulfur trioxide, HCl, and carbon dioxide using established sulfur dioxide concentration prediction models, sulfur trioxide concentration prediction models, HCl concentration prediction models, and carbon dioxide concentration prediction models. It further combines this with real-time operating data on desulfurization tower furnace temperature, desulfurization tower furnace pressure, and hydrated lime feed rate. Through a multi-model prediction optimization control strategy, the module sends the hydrated lime feed rate command to the DCS controller via the data communication module, thereby enabling the DCS system to send the command to the field equipment.

[0057] Preferably, the predicted sulfur dioxide value, the calculated sulfur trioxide value, the calculated HCl value, the calculated carbon dioxide value, and the sulfur dioxide value at the desulfurization tower inlet are used as given instructions, and the desulfurization tower furnace temperature and desulfurization tower furnace pressure are used as feedforward instructions.

[0058] Furthermore, f(X1) is a broken-line function representing the conversion of CO2 volume at the desulfurization tower inlet into the mass of slaked lime; f(X2) is a broken-line function representing the conversion of SO3 concentration at the desulfurization tower inlet into the mass of slaked lime; f(X3) is a broken-line function representing the conversion of SO2 concentration at the desulfurization tower inlet into the mass of slaked lime; f(X4) is a broken-line function representing the conversion of HCl concentration at the desulfurization tower inlet into the mass of slaked lime; f(X5) is a broken-line function representing the conversion of desulfurization tower temperature into the mass of slaked lime; and f(X6) is a broken-line function representing the conversion of desulfurization tower furnace pressure into the mass of slaked lime.

[0059] As a preferred approach, the multi-model predictive optimization control strategy selects the method that minimizes the feed rate of the hydrated lime feeder under the given outlet SO2 concentration condition. However, in actual operation, since the frequency of the hydrated lime feeder is difficult to precisely control, the flow rate can only be adjusted by a specific frequency modulation amplitude. Frequent hydrated lime feeder frequency adjustments can significantly impact the equipment's lifespan. Furthermore, to avoid exceeding the outlet SO2 concentration limit due to untimely adjustment of the hydrated lime feeder frequency, an upper limit for outlet SO2 concentration is set to ensure that SO2 emissions meet standards.

[0060] Furthermore, the frequency switching condition of the hydrated lime feeder can be expressed by the following mathematical expression:

[0061]

[0062] S time >S time,set

[0063] In the formula, To calculate the outlet SO2 concentration while maintaining the existing operating strategy based on the SO2 removal process model, The upper limit for SO2 emission concentration at the outlet is set here to 35 mg / m³. 3 S time,setThe frequency switching time for the slaked lime feeder can be determined according to the corresponding operating procedures.

[0064] Furthermore, the specific calculation logic of the multi-model prediction optimization control strategy mainly includes the following steps: First, select the best operating strategy under the current operating condition based on the optimized operating condition library, and determine whether the current operating strategy of the hydrated lime feeder is consistent with the best strategy. If they are consistent, no adjustment is required. If they are inconsistent, first determine whether the outlet SO2 concentration under the original operating strategy exceeds the set emission limit. If it does, ignore other conditions and immediately switch the feeder frequency to ensure that the flue gas meets the emission standards. If the emission standards are met, then determine the switching time. If the switching time is greater than the specified time, then switch. If the specified switching time has not been reached since the last switch, then maintain the original operating strategy.

[0065] The optimization strategy for the bed pressure drop control module mainly includes the following steps:

[0066] Step 1: Construct a bed pressure drop prediction model based on parameter identification;

[0067] Step 2: Given a target value for bed pressure drop, discretize the key input variables such as inlet pressure, flue gas flow rate, fresh hydrated lime particle flow rate, clean flue gas circulation flow rate, and ash material circulation flow rate;

[0068] Step 3: By discretization, the optimized configuration results of the opening of the regulating valve of the clean flue gas circulation fan and the opening of the regulating valve of the ash material circulation are obtained under different working conditions. A control rule table is constructed and used as the basis for judgment to implement control actions, so as to realize the intelligent control of the clean flue gas circulation fan and the ash material circulation.

[0069] Step 4: Further introduce feedback variables into the control rules. By using the feedback correction concept in model predictive control, obtain the model error by comparing the predicted bed pressure drop value output by the bed pressure drop prediction model based on parameter identification with the actual measured bed pressure drop value. Feedback the error to correct the given reference trajectory and input it into the inference module, thereby ensuring the effectiveness of the controller.

[0070] As a preferred option, the actual pressure drop of the absorber bed is used as the feedback value, and the output command of the bed pressure drop control strategy is used as the calculation parameter for the tracking value of the material circulation regulating valve. This value is then multiplied by the corresponding regulating valve balance coefficient and used as the tracking value of the regulating valve. In order to prevent the return material trough valve from frequent small movements, a valve position holding program is set when the fluctuation amplitude is less than 1.

[0071] Furthermore, the average value of the continuous material level in different ash hoppers is taken, and the actual material level in each ash hopper is divided by the average value to obtain the balance coefficient of the return ash valve, thereby realizing the balance control of the material level in each ash hopper.

[0072] The main beneficial effects of this invention are as follows:

[0073] (1) This invention proposes an intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system. Through a modeling method driven by knowledge and data of multiple pollutant generation-removal processes, combined with multi-dimensional analysis of the dynamic correlation characteristics between key parameters and pollutant concentrations, a highly reliable and interpretable multi-section SO2 / SO3 / CO2 / HCl concentration prediction model is constructed. This model enables the prediction of pollutant concentration changes 90 seconds in advance, with an average relative percentage error of less than 2%. This overcomes the problems of control lag caused by data measurement delay and difficulty in real-time accurate optimization.

[0074] (2) A smart control technology for a pollutant removal device with a dual-layer structure of multi-parameter collaborative optimization and model predictive control was invented. While ensuring stable ultra-low emissions of multiple pollutants throughout the entire period, the fluctuation range of outlet pollutant concentration is reduced by more than 60%, the outlet temperature fluctuation is within ±2℃, the bed pressure drop fluctuation is within ±75Pa, the consumption of quicklime is reduced by more than 10%, and the process water consumption is reduced by more than 5%. To a certain extent, the CO2 emission is reduced by more than 5%, which solves the problem of efficient and stable removal and synergistic carbon reduction of multiple pollutants in flue gas under complex operating conditions.

[0075] (3) This invention overcomes the lag of the slaked lime control system, solves the problem of excessive SO2 emissions caused by the delay of the slaked lime control system when the load is raised or lowered, and also avoids the problems of increased heat loss and increased cost caused by excessive slaked lime input; at the same time, it also uses the amount of CO2 at the flue gas inlet to assist in controlling the amount of slaked lime fed, which reduces the amount of CO2 emissions to a certain extent. Attached Figure Description

[0076] Figure 1 A schematic diagram of an intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system;

[0077] Figure 2 For the production of CO2 concentration, SO2 concentration, SO3 concentration, and HCl concentration at the inlet and outlet of the desulfurization tower

[0078] Flowchart of the model construction process for removing the root cause;

[0079] Figure 3 Process for constructing a prediction model for the adiabatic saturation temperature difference of desulfurized flue gas;

[0080] Figure 4 Control strategy diagram for optimizing the control module of slaked lime;

[0081] Figure 5 Optimization strategy for bed pressure drop control module:

[0082] Figure 6The application effect of the knowledge and data fusion modeling method based on two-layer data correction;

[0083] Figure 7 This is a diagram illustrating the intelligent control effect of a multi-stage humidification module.

[0084] Figure 8 Diagram showing the effect of bed pressure drop control. Detailed Implementation

[0085] This invention proposes an intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system. The invention will be described in detail below with reference to the accompanying drawings and specific examples.

[0086] Figure 1 This is a schematic diagram of an intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system. The system includes a hydrated lime preparation and conveying system, an ash circulation system, a temperature and humidity control system, a flue gas recirculation system, a pre-dust collector, a bag filter, a desulfurization tower, and an intelligent control system. The temperature and humidity control system is equipped with multi-stage humidification and activation spray guns; the hydrated lime preparation and conveying system is equipped with a variable frequency feeder; the ash circulation system is equipped with an ash circulation variable frequency feeder; the flue gas recirculation system is equipped with a clean flue gas recirculation variable frequency fan; and the intelligent control system includes an online monitoring module, a key parameter prediction module, and a key parameter control module.

[0087] The DCS system is connected to the circulating fluidized bed desulfurization synergistic carbon-pollution deep purification system and the intelligent control system. The intelligent control system includes: a data communication module, a module for accurate prediction of the concentration of multiple pollutants such as SO2 / SO3 / CO2 / HCl, a module for calculating the amount of multiple pollutants such as SO2 / SO3 / CO2 / HCl, a module for predicting the temperature and pressure of flue gas at the desulfurization tower outlet, a module for controlling the variable frequency feeder of quicklime, a module for predicting and controlling the bed pressure drop, and a module for controlling the multi-stage humidification and activation spray gun.

[0088] Among them, the data communication module exchanges data with the DCS system; the multi-pollutant concentration accurate prediction module is connected to the data communication module and the hydrated lime frequency converter control module; the hydrated lime frequency converter control module is connected to the communication module and the SO2 / SO3 / CO2 / HCl multi-pollutant quantity calculation module; the desulfurization tower outlet flue gas temperature and flue gas pressure prediction module is connected to the data communication module and the bed pressure drop prediction control module; and the multi-stage humidification and activation spray gun input quantity prediction module is connected to the data communication module and the multi-stage humidification and activation spray gun control module.

[0089] The online monitoring module includes parameters such as total coal feed, total ore feed, coal quality parameters, mineral composition, flue gas temperature at the sintering process outlet, wind box temperature, combustion air flow rate, main exhaust fan flow rate, flue gas oxygen content, SO2 concentration at the desulfurization tower inlet and outlet, flue gas temperature and moisture content of the inlet flue gas, feeder frequency, ash circulation feeder frequency, clean flue gas recirculation variable frequency fan frequency, and the number of multi-stage humidification and activation spray guns in operation.

[0090] The data communication module exchanges data with the DCS system through the ModBus communication protocol and reads real-time data including: total coal feed rate, total ore quantity, coal quality parameters, mineral composition, flue gas temperature at the sintering process outlet, wind box temperature, combustion air flow rate, main exhaust fan flow rate, flue gas oxygen content parameters, SO2 concentration at the inlet and outlet of the desulfurization tower, flue gas temperature and moisture content of the inlet flue gas, feeder frequency, ash circulation feeder frequency, clean flue gas recirculation variable frequency fan frequency, and the number of multi-stage humidification and activation spray guns put into operation.

[0091] The real-time data read by the communication module may be distorted and incomplete due to instrument issues, data transmission problems, or purging. To ensure the model can fully extract the influence of each feature, the data is standardized / normalized and purged / processed for missing data before model construction. The specific steps are as follows:

[0092] Step 1.1: Identification and handling of data outliers, filtering out zero, negative, and null values ​​and cleaning them up.

[0093] Step 1.2: CEMS purging and maintenance data processing. Data during purging can be judged using the following formula.

[0094]

[0095] In the formula, Let i be the SO2 concentration at the inlet of the desulfurization unit measured at time i. Let be the SO2 concentration at the inlet of the desulfurization unit measured at time i+j, where j = 1, 2, 3, ..., 20; To determine the minimum change in SO2 concentration at the inlet of the purging desulfurization unit.

[0096] Step 1.3: To eliminate the impact of different feature magnitudes on the model, the data needs to be normalized, as shown in the following formula:

[0097]

[0098] In the formula, It is the normalized data of this feature at time i, x i It is the original data of this feature at time i, x min It is the minimum value of this feature, xmax It is the maximum value of this feature.

[0099] Figure 2 A flowchart is provided for constructing a prediction model for the generation and removal of CO2, SO2, SO3, and HCl concentrations at the inlet and outlet of the desulfurization tower.

[0100] The construction of the prediction model for the generation and removal of CO2, SO2, SO3, and HCl concentrations at the inlet and outlet of the desulfurization tower mainly includes the following steps:

[0101] Step S1: Analyze the formation mechanism of multiple pollutants such as CO2 / SO2 / SO3 / HCl / HF during combustion / sintering, and combine data analysis methods to screen the key parameters affecting the concentration of multiple pollutants such as CO2 / SO2 / SO3 / HCl / HF at the inlet of the circulating fluidized bed desulfurization tower;

[0102] Step S2: Compensate for the delay between key variables through data processing, and construct a knowledge and data fusion driven model for the generation concentration of multiple pollutants such as CO2 / SO2 / SO3 / HCl / HF based on an LSTM neural network with autoregressive variables, so as to achieve accurate prediction of the concentration of multiple pollutants such as CO2 / SO2 / SO3 / HCl / HF at the inlet of the desulfurization unit in advance.

[0103] Step S3: Analyze the adsorption and removal mechanism of multiple pollutants such as CO2 / SO2 / SO3 / HCl / HF in the flue gas circulating fluidized bed desulfurization tower with absorbent, establish an adsorption and removal mechanism model of acid gas, further introduce correction parameters into the adsorption and removal mechanism model of multiple pollutants such as CO2 / SO2 / SO3 / HCl / HF, solve it through PSO algorithm, and construct a parameter identification-based adsorption and removal model of multiple pollutants such as CO2 / SO2 / SO3 / HCl / HF.

[0104] Step S4: Further, an LSTM neural network is used to compensate for the error in the parameter identification model. Combined with actual operating data, a two-layer data-corrected prediction model for the removal of multiple pollutants such as CO2 / SO2 / SO3 / HCl / HF is established, which realizes the accurate prediction of the concentration of multiple pollutants such as CO2 / SO2 / SO3 / HCl / HF at the outlet of the desulfurization unit.

[0105] As a preferred embodiment, the knowledge and data fusion-driven SO2 concentration prediction model for the desulfurization tower inlet is described as follows:

[0106]

[0107] in: K represents the predicted sulfur dioxide concentration at the inlet of the desulfurization tower. r K represents the SO2 emission factor caused by the operating conditions of the boiler / sintering process. r= f(T,B,L1,L2…); T is the flue gas temperature in the furnace or sintering process, B is the coal feed rate, L1 is the primary air volume, L2 is the secondary air volume, a fh A is the fly ash coefficient. ar To obtain the base ash content, S ar Received basic sulfur content, Q net,ar To receive the base low heat, ω CaO ω MgO , denoted as , where is the mass fraction of the corresponding substance in the ash, n is the hourly coal consumption, and Q is the flue gas volume at the inlet of the desulfurization tower.

[0108] As a preferred embodiment, the knowledge and data fusion-driven SO3 concentration prediction model at the desulfurization tower inlet is described as follows:

[0109]

[0110] in: The values ​​represent the predicted sulfur trioxide concentration at the inlet of the desulfurization tower, T is the flue gas temperature in the furnace or sintering process, B is the coal feed rate, L1 is the primary air volume, and L2 is the secondary air volume. is the predicted sulfur dioxide concentration at the inlet of the desulfurization tower, and Q is the flue gas volume at the inlet of the desulfurization tower.

[0111] As a preferred embodiment, the knowledge and data fusion-driven HCl concentration prediction model for the desulfurization tower inlet is described as follows:

[0112]

[0113] Where: C HCl,in,yc K represents the predicted hydrogen chloride concentration at the inlet of the desulfurization tower. r The HCl emission factor is determined by the boiler / sintering process operating conditions; T is the flue gas temperature in the furnace or sintering process; B is the coal feed rate; L1 is the primary air volume; L2 is the secondary air volume; a fh A is the fly ash coefficient. ar To obtain the base ash content, Q net,ar To receive the base low heat, ω CaO ω MgO , denoted as , where is the mass fraction of the corresponding substance in the ash, n is the hourly coal consumption, and Q is the flue gas volume at the inlet of the desulfurization tower.

[0114] As a preferred embodiment, the knowledge and data fusion-driven CO2 concentration prediction model for the desulfurization tower inlet is described as follows:

[0115]

[0116] in: B is the CO2 volume at the inlet of the desulfurization tower, and O is the coal feed rate. FC is the total air volume entering the combustion process. ar Carbon content (H) of fuel / ore received basis ar For fuel / ore received basic hydrogen content, O ar V represents the basic oxygen content received by fuel / ore. CO β is the carbon volume of carbon monoxide at the inlet of the desulfurization tower, and β is the fuel characteristic coefficient. Here is the predicted sulfur dioxide concentration at the inlet of the desulfurization tower, and Q is the flue gas volume at the inlet of the desulfurization tower.

[0117] As a preferred embodiment, the SO2 concentration prediction model at the desulfurization tower outlet based on parameter identification is described as follows:

[0118]

[0119] Where: M Ca(OH)2 Z represents the feed rate of slaked lime. h For the conversion rate of slaked lime, f(t) l , t f …); t f t is the time it takes for the desulfurizing agent to complete its reaction in the absorption tower. l Residence time of desulfurizing agent in the absorption tower; V is the partial pressure of SO2 inside the desulfurization tower; T is the temperature; V is the volume of the desulfurization tower; and a is the air leakage coefficient of the desulfurization tower. Here, Q represents the predicted sulfur dioxide concentration at the inlet of the desulfurization tower, and R is a constant.

[0120] Furthermore, t f The complete reaction time of the desulfurizing agent in the absorption tower is t. l There are two parameters that need to be identified regarding the residence time of the desulfurizing agent in the absorption tower. The parameter identification process can be represented by an optimization problem, namely:

[0121]

[0122] In the formula, RMSE(t) l , t f The model calculates the root mean square error between the predicted and actual measured values ​​of SO2 concentration at the outlet.

[0123] Figure 3 The process for constructing a prediction model for the adiabatic saturated temperature difference of desulfurized flue gas mainly includes the following steps:

[0124] Step W1: Analyze the adiabatic cooling and humidification mechanism of flue gas in the flue gas circulating fluidized bed desulfurization process, construct a model for predicting the humidity of flue gas at the desulfurization tower outlet, collect parameters such as the moisture content of flue gas at the desulfurization tower inlet, the flue gas volume at the desulfurization tower inlet, the flue gas temperature at the desulfurization tower inlet, the flue gas pressure at the desulfurization tower inlet, the water spray volume at the desulfurization tower inlet, the flue gas pressure at the desulfurization tower outlet, and the flue gas temperature at the desulfurization tower outlet, and construct a database for predicting the adiabatic saturated temperature difference of desulfurized flue gas.

[0125] Step W2: Combining data analysis methods, screen the key parameters affecting the humidity of the flue gas at the desulfurization tower outlet, further analyze the pure time delay between changes in key variables and the humidity of the flue gas at the desulfurization tower outlet, and compensate for the delay between key variables through data processing.

[0126] Step W3: By introducing correction parameters into the desulfurization tower outlet flue gas humidity mechanism model and solving it using the PSO algorithm, a parameter-identified desulfurization tower outlet flue gas humidity control mechanism model was constructed.

[0127] Step W4: Further, based on an LSTM neural network with autoregressive variables, construct a desulfurization tower outlet flue gas humidity prediction model that integrates data and humidity control mechanisms.

[0128] Step W5: Further analyze the adiabatic saturation temperature regulation mechanism of desulfurized flue gas and construct a prediction model for the adiabatic saturation temperature difference of desulfurized flue gas based on the fusion of empirical knowledge and mechanism.

[0129] Step W6: Further combining the flue gas temperature at the desulfurization tower outlet, the flue gas pressure at the desulfurization tower inlet, and the flue gas humidity prediction model at the desulfurization tower outlet, a mechanism-operation data correction fusion desulfurization flue gas adiabatic saturation temperature difference prediction model is constructed, realizing accurate prediction of the adiabatic saturation temperature difference of desulfurization flue gas under variable load conditions.

[0130] Preferably, the desulfurization tower outlet flue gas humidity control mechanism model identified by the parameters can be expressed by the following formula:

[0131] H=λ*(G*(1-δ) / (Q*ρ)+H i )

[0132] Where G is the amount of water injected into the desulfurization tower, Q is the volumetric flow rate of the flue gas, ρ is the density of the flue gas, and H is the volumetric flow rate of the flue gas. i δ represents the humidity of the flue gas at the inlet of the desulfurization tower, δ represents the coefficient of residual moisture in the desulfurization ash in the total spray volume, and λ represents the correction coefficient for the humidity of the flue gas at the outlet of the desulfurization tower.

[0133] Furthermore, parameters δ and λ in the empirical model of flue gas humidity at the desulfurization tower outlet are identified. The parameter identification process can be represented by an optimization problem, namely:

[0134]

[0135] Wherein, RMSE(δ, λ) represents the root mean square error between the predicted and actual values ​​of the flue gas humidity at the desulfurization tower outlet, as determined by the empirical model.

[0136] Preferably, the adiabatic saturation temperature difference of the desulfurized flue gas can be expressed by the following formula:

[0137] ΔT s =TT s

[0138] Where, ΔT s T represents the adiabatic saturation temperature difference of the desulfurization flue gas, and T represents the outlet flue gas temperature of the desulfurization tower. s It is the adiabatic saturation temperature;

[0139] Furthermore, the prediction model for the adiabatic saturated temperature difference of desulfurized flue gas based on the fusion of empirical knowledge and mechanisms can be expressed by the following formula:

[0140]

[0141] Where C is the specific heat of the moist flue gas, M is the molecular weight of the flue gas, and T is the outlet temperature of the desulfurization tower. s Let T be the adiabatic saturation temperature, H be the humidity of the flue gas at temperature T, and P be the temperature at temperature P. f This refers to the flue gas pressure.

[0142] The multi-stage humidification module is not only the module with the highest water consumption in the desulfurization unit, but also the most important subsystem for ensuring the desulfurization outlet flue gas temperature and the adiabatic saturation temperature difference of the desulfurization flue gas. Therefore, when constructing the intelligent control problem of the multi-stage humidification module of the flue gas desulfurization unit, the operating water consumption, the desulfurization outlet flue gas temperature, and the adiabatic saturation temperature difference of the desulfurization flue gas should be considered simultaneously. Thus, the intelligent control process can also be described by the following optimization problem:

[0143] minWater(L1,L2,L3,L4)=L1p1+L2p2+L3p3+L4p4

[0144]

[0145] L1, L2, L3, and L4 represent the operating states of the humidification nozzles at levels 1 to 4. When the humidification nozzle is on, the operating state is 1; when the humidification nozzle is off, the operating state is 0. p1, p2, p3, and p4 represent the water flow rates corresponding to the humidification nozzles. This is the predicted value for the outlet flue gas temperature. This is the predicted value of the adiabatic saturation temperature difference of the desulfurization flue gas.

[0146] Figure 4The control strategy diagram for the lime optimization control module is as follows: Based on the established sulfur dioxide concentration prediction model, sulfur trioxide concentration prediction model, HCl concentration prediction model, and carbon dioxide concentration prediction model, the predicted values ​​of sulfur dioxide, sulfur trioxide, HCl, and carbon dioxide are calculated. Furthermore, combined with the real-time operating data of desulfurization tower furnace temperature, desulfurization tower furnace pressure, and hydrated lime feed rate, the hydrated lime feed rate instruction is sent to the DCS controller through the data communication module after adjustment by the multi-model prediction optimization control strategy. The DCS system then sends the instruction to the field equipment.

[0147] Preferably, the predicted sulfur dioxide value, the calculated sulfur trioxide value, the calculated HCl value, the calculated carbon dioxide value, and the sulfur dioxide value at the desulfurization tower inlet are used as given instructions, and the desulfurization tower furnace temperature and desulfurization tower furnace pressure are used as feedforward instructions.

[0148] Furthermore, f(X1) is a broken-line function representing the conversion of CO2 volume at the desulfurization tower inlet into the mass of slaked lime; f(X2) is a broken-line function representing the conversion of SO3 concentration at the desulfurization tower inlet into the mass of slaked lime; f(X3) is a broken-line function representing the conversion of SO2 concentration at the desulfurization tower inlet into the mass of slaked lime; f(X4) is a broken-line function representing the conversion of HCl concentration at the desulfurization tower inlet into the mass of slaked lime; f(X5) is a broken-line function representing the conversion of desulfurization tower temperature into the mass of slaked lime; and f(X6) is a broken-line function representing the conversion of desulfurization tower furnace pressure into the mass of slaked lime.

[0149] As a preferred approach, the multi-model predictive optimization control strategy selects the method that minimizes the feed rate of the hydrated lime feeder under the given outlet SO2 concentration condition. However, in actual operation, since the frequency of the hydrated lime feeder is difficult to precisely control, the flow rate can only be adjusted by a specific frequency modulation amplitude. Frequent hydrated lime feeder frequency adjustments can significantly impact the equipment's lifespan. Furthermore, to avoid exceeding the outlet SO2 concentration limit due to untimely adjustment of the hydrated lime feeder frequency, an upper limit for outlet SO2 concentration is set to ensure that SO2 emissions meet standards.

[0150] Furthermore, the frequency switching condition of the hydrated lime feeder can be expressed by the following mathematical expression:

[0151]

[0152] S time >S time,set

[0153] In the formula, To calculate the outlet SO2 concentration while maintaining the existing operating strategy based on the SO2 removal process model, The upper limit for SO2 emission concentration at the outlet is set here to 35 mg / m³. 3 S time,setThe frequency switching time for the slaked lime feeder can be determined according to the corresponding operating procedures.

[0154] Furthermore, the specific calculation logic of the multi-model prediction optimization control strategy mainly includes the following steps: First, select the best operating strategy under the current operating condition based on the optimized operating condition library, and determine whether the current operating strategy of the hydrated lime feeder is consistent with the best strategy. If they are consistent, no adjustment is required. If they are inconsistent, first determine whether the outlet SO2 concentration under the original operating strategy exceeds the set emission limit. If it does, ignore other conditions and immediately switch the feeder frequency to ensure that the flue gas meets the emission standards. If the emission standards are met, then determine the switching time. If the switching time is greater than the specified time, then switch. If the specified switching time has not been reached since the last switch, then maintain the original operating strategy.

[0155] Figure 5 The optimization strategy for the bed pressure drop control module mainly includes the following steps:

[0156] Step X1: Construct a bed pressure drop prediction model based on parameter identification;

[0157] Step X2: Given a target value for bed pressure drop, discretize the key input variables such as inlet pressure, flue gas flow rate, fresh hydrated lime particle flow rate, clean flue gas circulation flow rate, and ash material circulation flow rate;

[0158] Step X3: By discretizing, the optimized configuration results of the opening of the regulating valve of the clean flue gas circulation fan and the opening of the regulating valve of the ash material circulation are obtained under different working conditions. A control rule table is constructed and used as the basis for judgment to implement control actions, so as to realize the intelligent control of the clean flue gas circulation fan and the ash material circulation.

[0159] Step X4: Further introduce feedback variables into the control rules. By using the feedback correction concept in model predictive control, obtain the model error by comparing the predicted bed pressure drop value output by the bed pressure drop prediction model based on parameter identification with the actual measured bed pressure drop value. Feedback the error to correct the given reference trajectory and input it into the inference module, thereby ensuring the effectiveness of the controller.

[0160] As a preferred option, the actual pressure drop of the absorber bed is used as the feedback value, and the output command of the bed pressure drop control strategy is used as the calculation parameter for the tracking value of the material circulation regulating valve. This value is then multiplied by the corresponding regulating valve balance coefficient and used as the tracking value of the regulating valve. In order to prevent the return material trough valve from frequent small movements, a valve position holding program is set when the fluctuation amplitude is less than 1.

[0161] Furthermore, the average value of the continuous material level in different ash hoppers is taken, and the actual material level in each ash hopper is divided by the average value to obtain the balance coefficient of the return ash valve, thereby realizing the balance control of the material level in each ash hopper.

[0162] To verify the accuracy and effectiveness of the knowledge and data fusion modeling method based on two-layer data correction, this invention was applied in an engineering study of a flue gas circulating fluidized bed desulfurization project in China. Operating conditions were selected, covering stable, sudden load, low load, medium load, and high load. Data was collected at 5-second intervals, resulting in 72 consecutive hours of data for verification. The results are as follows: Figure 6 As shown, the model predictions of SO2 emission concentrations are consistent with the actual measured values. Based on parameter identification, a recurrent neural network is used to construct a data correction model to further compensate for model errors, achieving an RMSE of 0.51 mg / m³. 3 This indicates that the SO2 emission concentration model constructed based on the knowledge and data fusion modeling method with two-layer data correction has high prediction accuracy.

[0163] Based on the established sulfur dioxide concentration prediction models, sulfur trioxide concentration prediction models, HCl concentration prediction models, and carbon dioxide concentration prediction models, the predicted values ​​of sulfur dioxide, sulfur trioxide, HCl, and carbon dioxide are calculated. Furthermore, combined with the real-time operating data of desulfurization tower furnace temperature, desulfurization tower furnace pressure, and hydrated lime feed rate, the hydrated lime feed rate instruction is sent to the DCS controller through the data communication module after adjustment by the multi-model prediction optimization control strategy. The DCS system then sends the instruction to the field equipment.

[0164] Based on the accurate prediction model of sulfur dioxide emission concentration using knowledge and data fusion with two-layer data correction, optimized control of sulfur dioxide and slaked lime was carried out, and the results are as follows: Figure 7 As shown, the average sulfur dioxide emission concentration before control according to this invention patent is 20 mg / m³. 3 The optimized average emission concentration is 30 mg / m³. 3 The fluctuation of SO2 concentration at the outlet is reduced by more than 40%, achieving high-precision "edge-cutting" control of SO2 emission concentration. Consequently, the amount of quicklime feed is reduced by 10%, carbon dioxide emissions are reduced by 9%, and the desulfurization efficiency in the desulfurization tower reaches 97.5%.

[0165] Based on the accurate prediction model of the adiabatic saturation temperature difference of desulfurized flue gas, intelligent control of multi-stage humidification modules was carried out, and the results are as follows: Figure 7As shown, before the application of this patent, the flue gas temperature at the desulfurization tower outlet was between 95 and 110℃, with a fluctuation of approximately ±7.5℃, indicating a large fluctuation range. After the application of this patent, the flue gas temperature at the desulfurization tower outlet is stably controlled at 105℃, with a fluctuation of approximately ±2℃. Compared to before the application of the patent, the fluctuation of the flue gas temperature at the desulfurization tower outlet is reduced by more than 70%, achieving high-precision "edge-cut" control of the flue gas temperature at the desulfurization tower outlet. Consequently, the water consumption of the multi-stage humidification module is reduced by more than 5%, further achieving appropriate control of the adiabatic saturation temperature difference. This improves desulfurization efficiency while avoiding excessively low adiabatic saturation temperature differences that could cause condensation in the flue gas, corrosion of the inner wall, and damage to the separation and dust removal equipment.

[0166] Increasing the opening of the material circulation regulating valve will increase the bed pressure drop, and vice versa. The function of the bed pressure drop control loop is to control the bed pressure drop by adjusting the opening of the material circulation regulating valve. After the application of this patented technology, such as... Figure 8 As shown, the bed pressure drop under varying loads is continuously stabilized and controlled between 1500Pa and 1650Pa, reducing the fluctuation range by more than 30% compared to before the application of the results.

[0167] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system, the system comprising a quicklime preparation and conveying system, an ash circulation system, a temperature and humidity control system, a flue gas recirculation system, a pre-dust collector, a bag filter, a desulfurization tower, and an intelligent control system, characterized in that, The temperature and humidity control system is equipped with multi-stage humidification and activation spray guns; the quicklime preparation and conveying system is equipped with a variable frequency feeder; the ash circulation system is equipped with an ash circulation variable frequency feeder; the flue gas recirculation system is equipped with a clean flue gas recirculation variable frequency fan; and the intelligent control system includes an online monitoring module, a key parameter prediction module, and a key parameter control module. The online monitoring module includes parameters such as total coal feed, total ore feed, coal quality parameters, mineral composition, flue gas temperature at the sintering process outlet, wind box temperature, combustion air flow rate, main exhaust fan flow rate, flue gas oxygen content, SO2 concentration at the desulfurization tower inlet and outlet, flue gas temperature and moisture content of the inlet flue gas, feeder frequency, ash circulation feeder frequency, clean flue gas recirculation frequency converter fan frequency, and the number of multi-stage humidification and activation spray guns put into operation. The key parameter prediction module includes prediction models for the generation and removal of CO2, SO2, SO3, and HCl concentrations at the inlet and outlet of the desulfurization tower, as well as prediction models for flue gas temperature and pressure at the outlet of the desulfurization tower. The key parameter control module includes a control model for a hydrated lime frequency converter feeder, a bed pressure drop prediction and control model, and a multi-stage humidification and activation spray gun control model. The establishment of the prediction model for the outlet flue gas temperature of the desulfurization tower and the adiabatic saturation temperature difference of the desulfurization flue gas mainly includes the following steps: Step W1: Analyze the adiabatic cooling and humidification mechanism of flue gas in the flue gas circulating fluidized bed desulfurization process, construct a model for predicting the humidity of flue gas at the desulfurization tower outlet, collect parameters such as the moisture content of flue gas at the desulfurization tower inlet, the flue gas volume at the desulfurization tower inlet, the flue gas temperature at the desulfurization tower inlet, the flue gas pressure at the desulfurization tower inlet, the water spray volume at the desulfurization tower inlet, the flue gas pressure at the desulfurization tower outlet, and the flue gas temperature at the desulfurization tower outlet, and construct a database for predicting the adiabatic saturated temperature difference of desulfurized flue gas. Step W2: Combining data analysis methods, screen the key parameters affecting the humidity of the flue gas at the desulfurization tower outlet, further analyze the pure time delay between changes in key variables and the humidity of the flue gas at the desulfurization tower outlet, and compensate for the delay between key variables through data processing. Step W3: By introducing correction parameters into the desulfurization tower outlet flue gas humidity mechanism model and solving it using the PSO algorithm, a parameter-identified desulfurization tower outlet flue gas humidity control mechanism model was constructed. Step W4: Further construct a desulfurization tower outlet flue gas humidity prediction model based on an LSTM neural network with autoregressive variables, which integrates data and humidity control mechanisms. Step W5: Further analyze the adiabatic saturation temperature regulation mechanism of desulfurized flue gas and construct a prediction model for the adiabatic saturation temperature difference of desulfurized flue gas based on the fusion of empirical knowledge and mechanism. Step W6: Further combining the flue gas temperature at the desulfurization tower outlet, the flue gas pressure at the desulfurization tower inlet, and the flue gas humidity prediction model at the desulfurization tower outlet, a mechanism-operation data correction fusion desulfurization flue gas adiabatic saturation temperature difference prediction model is constructed, realizing accurate prediction of the adiabatic saturation temperature difference of desulfurization flue gas under variable load conditions.

2. The intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system according to claim 1, characterized in that: The construction of the prediction model for the generation and removal of CO2, SO2, SO3, and HCl concentrations at the inlet and outlet of the desulfurization tower mainly includes the following steps: Step S1: Analyze the generation mechanism of CO2, SO2, SO3, and HCl acidic gases during combustion / sintering, and combine data analysis methods to screen the key parameters affecting the concentration of CO2, SO2, SO3, and HCl acidic gases at the inlet of the circulating fluidized bed desulfurization tower; Step S2: Compensate for the delay between key variables through data processing, and construct a knowledge and data fusion driven model for the generation concentrations of CO2, SO2, SO3, HCl and HF acidic gases based on an LSTM neural network with autoregressive variables, so as to achieve accurate prediction of the concentrations of CO2, SO2, SO3 and HCl acidic gases at the inlet of the desulfurization unit in advance. Step S3: Analyze the adsorption and removal mechanism of CO2, SO2, SO3, and HCl acidic gases in the flue gas circulating fluidized bed desulfurization tower with the absorbent, establish an adsorption and removal mechanism model for acidic gases, further introduce correction parameters into the adsorption and removal mechanism model of CO2, SO2, SO3, and HCl acidic gases, solve it through the PSO algorithm, and construct an adsorption and removal model of CO2, SO2, SO3, and HCl acidic gases based on parameter identification; Step S4: Further, LSTM neural network is used to compensate for the error of parameter identification model. Combined with actual operation data, a knowledge and data fusion driven acid gas concentration prediction model with double-layer data correction is established, which realizes accurate prediction of CO2, SO2, SO3 and HCl acid gas concentrations at the outlet of desulfurization unit. The knowledge and data fusion-driven SO2 concentration prediction model at the desulfurization tower inlet is described as follows: in: K represents the predicted sulfur dioxide concentration at the inlet of the desulfurization tower. r K represents the SO2 emission factor caused by the operating conditions of the boiler / sintering process. r =f(T,B,L1,L2); T is the flue gas temperature in the furnace or sintering process, B is the coal feed rate, L1 is the primary air volume, L2 is the secondary air volume, and a fh A is the fly ash coefficient. ar To obtain the base ash content, S ar Received basic sulfur content, Q net,ar To receive the base low heat, ω CaO ω MgO , denoted as , where is the mass fraction of the corresponding substance in the ash, n is the hourly coal consumption, and Q is the flue gas volume at the inlet of the desulfurization tower.

3. The intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system according to claim 2, characterized in that: The knowledge and data fusion-driven model for predicting SO3 concentration at the inlet of a desulfurization tower is described as follows: in: The values ​​represent the predicted sulfur trioxide concentration at the inlet of the desulfurization tower, T is the flue gas temperature in the furnace or sintering process, B is the coal feed rate, L1 is the primary air volume, and L2 is the secondary air volume. Here is the predicted sulfur dioxide concentration at the inlet of the desulfurization tower, and Q is the flue gas volume at the inlet of the desulfurization tower. The knowledge and data fusion-driven model for predicting HCl concentration at the inlet of the desulfurization tower is described as follows: Where: C HCl,in,yc K represents the predicted hydrogen chloride concentration at the inlet of the desulfurization tower. r The HCl emission factor is determined by the boiler / sintering process operating conditions; T is the flue gas temperature in the furnace or sintering process; B is the coal feed rate; L1 is the primary air volume; L2 is the secondary air volume; a fh A is the fly ash coefficient. ar To obtain the base ash content, Q net,ar To receive the base low heat, ω CaO ω MgO , denoted as the mass fraction of the corresponding substance in the ash, n is the hourly coal consumption, and Q is the flue gas volume at the inlet of the desulfurization tower. The knowledge and data fusion-driven model for predicting CO2 concentration at the inlet of a desulfurization tower is described as follows: in: Let Q be the CO2 volume at the inlet of the desulfurization tower, B be the coal feed rate, and Q be the total CO2 volume. F C is the total air volume entering the combustion process. ar Carbon content (H) of fuel / ore received basis ar For fuel / ore received basic hydrogen content, O ar V represents the basic oxygen content received by fuel / ore. CO β is the carbon volume of carbon monoxide at the inlet of the desulfurization tower, and β is the fuel characteristic coefficient. Here is the predicted sulfur dioxide concentration at the inlet of the desulfurization tower, and Q is the flue gas volume at the inlet of the desulfurization tower. The prediction model for SO2 concentration at the outlet of the desulfurization tower based on parameter identification is described below: Where: M Ca(OH)2 Z represents the feed rate of slaked lime. h Z represents the conversion rate of slaked lime. h =f(t) l , t f );t f t is the time it takes for the desulfurizing agent to complete its reaction within the absorption tower. l Residence time of desulfurizing agent in the absorption tower; V is the partial pressure of SO2 inside the desulfurization tower; T is the temperature; V is the volume of the desulfurization tower; and a is the air leakage coefficient of the desulfurization tower. Here, Q represents the predicted sulfur dioxide concentration at the inlet of the desulfurization tower, and R is a constant. Further based on the above model, t f The complete reaction time of the desulfurizing agent in the absorption tower is t. l There are two parameters that need to be identified regarding the residence time of the desulfurizing agent in the absorption tower. The parameter identification process can be represented by an optimization problem, namely: In the formula, RMSE(t) l , t f The model calculates the root mean square error between the predicted and actual measured values ​​of SO2 concentration at the outlet.

4. The intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system according to claim 1, characterized in that, The desulfurization tower outlet flue gas humidity control mechanism model identified by the parameters is expressed by the following formula: H=λ*(G*(1-δ) / (Q*ρ)+H i ) Where G is the amount of water injected into the desulfurization tower, Q is the volumetric flow rate of the flue gas, ρ is the density of the flue gas, and H is the volumetric flow rate of the flue gas. i δ is the humidity of the flue gas at the inlet of the desulfurization tower, δ is the coefficient of the residual moisture in the desulfurization ash in the total spray volume, and λ is the humidity correction coefficient of the flue gas at the outlet of the desulfurization tower. The parameters δ and λ in the empirical model of flue gas humidity at the desulfurization tower outlet are identified. The parameter identification process is represented by an optimization problem, namely: Wherein, RMSE(δ, λ) represents the root mean square error between the predicted and actual values ​​of the flue gas humidity at the desulfurization tower outlet, as predicted by the empirical model. The adiabatic saturation temperature difference of the desulfurized flue gas is expressed by the following formula: ΔT s =T-T s Where, ΔT s T represents the adiabatic saturation temperature difference of the desulfurization flue gas, and T represents the outlet flue gas temperature of the desulfurization tower. s It is the adiabatic saturation temperature; The prediction model for the adiabatic saturated temperature difference of desulfurized flue gas based on the fusion of empirical knowledge and mechanisms is expressed by the following formula: Where C is the specific heat of the moist flue gas, M is the molecular weight of the flue gas, and T is the outlet temperature of the desulfurization tower. s Let T be the adiabatic saturation temperature, H be the humidity of the flue gas at temperature T, and P be the temperature at temperature P. f This refers to the flue gas pressure.

5. The intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system according to claim 1, characterized in that: The intelligent control and optimization description of the multi-stage humidification module in the flue gas desulfurization unit is as follows: min Water(L1,L2,L3,L4)=L1p1+L2p2+L3p3+L4p4 L1, L2, L3, and L4 represent the operating states of the humidification nozzles from level 1 to level 4. When the humidification nozzle is on, the operating state is 1; when the humidification nozzle is off, the operating state is 0. p1, p2, p3, and p4 represent the water flow rates corresponding to the humidification nozzles. This is the predicted value for the outlet flue gas temperature. This is the predicted value of the adiabatic saturation temperature difference of the desulfurization flue gas.

6. The intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system according to claim 1, characterized in that: The hydrated lime optimization control module calculates the predicted values ​​of sulfur dioxide, sulfur trioxide, HCl, and carbon dioxide using established sulfur dioxide concentration prediction models, sulfur trioxide concentration prediction models, HCl concentration prediction models, and carbon dioxide concentration prediction models. It further combines the real-time operating data of desulfurization tower furnace temperature, desulfurization tower furnace pressure, and hydrated lime feed rate with the multi-model prediction optimization control strategy, and sends the hydrated lime feed rate command to the DCS controller through the data communication module. The DCS system then sends the command to the field equipment. The predicted value of sulfur dioxide, the calculated value of sulfur trioxide, the calculated value of HCl, the calculated value of carbon dioxide, and the sulfur dioxide value at the inlet of the desulfurization tower are used as given instructions, and the furnace temperature and furnace pressure of the desulfurization tower are used as feedforward instructions. f(X1) is a broken-line function representing the conversion of CO2 volume at the desulfurization tower inlet into slaked lime mass; f(X2) is a broken-line function representing the conversion of SO3 concentration at the desulfurization tower inlet into slaked lime mass; f(X3) is a broken-line function representing the conversion of SO2 concentration at the desulfurization tower inlet into slaked lime mass; f(X4) is a broken-line function representing the conversion of HCl concentration at the desulfurization tower inlet into slaked lime mass; f(X5) is a broken-line function representing the conversion of desulfurization tower temperature into slaked lime mass; f(X6) is a broken-line function representing the conversion of desulfurization tower furnace pressure into slaked lime mass. The multi-model prediction optimization control strategy selects the method with the minimum feed rate of the hydrated lime feeder under the condition of setting the outlet SO2 concentration. At the same time, in actual operation, since it is difficult to achieve precise control of the frequency of the hydrated lime feeder, the flow rate of hydrated lime can only be adjusted by a specific frequency adjustment range. Frequent hydrated lime feeder frequency will greatly affect the service life of the equipment. In addition, in order to avoid the problem of exceeding the outlet SO2 concentration due to the untimely adjustment of the hydrated lime feeder frequency, an upper limit for the outlet SO2 concentration is set to ensure that SO2 emissions meet the standards. The frequency switching condition of the hydrated lime feeder is expressed by the following mathematical expression: S time >S time,set In the formula, To calculate the outlet SO2 concentration while maintaining the existing operating strategy based on the SO2 removal process model, The upper limit for SO2 emission concentration at the outlet is set here to 35 mg / m³. 3 ; S time,set The frequency switching time for the slaked lime feeder shall be determined according to the relevant operating procedures. The specific calculation logic of the multi-model prediction optimization control strategy mainly includes the following steps: First, select the best operating strategy under the current operating condition based on the optimized operating condition library, and determine whether the current operating strategy of the hydrated lime feeder is consistent with the best strategy. If they are consistent, no adjustment is required. If they are inconsistent, first determine whether the outlet SO2 concentration under the original operating strategy exceeds the set emission limit. If it exceeds the limit, ignore other conditions and immediately switch the feeder frequency to prioritize ensuring that the flue gas meets the emission standards. If the emission standards are met, determine the switching time. If the switching time is greater than the specified time, switch. If the specified switching time has not been reached since the last switch, maintain the original operating strategy.

7. The intelligent circulating fluidized bed desulfurization and carbon-pollution deep purification system according to claim 1, characterized in that: The optimization strategy for the bed pressure drop control module mainly includes the following steps: Step 1: Construct a bed pressure drop prediction model based on parameter identification; Step 2: Given a target value for bed pressure drop, discretize the key input variables such as inlet pressure, flue gas flow rate, fresh hydrated lime particle flow rate, clean flue gas circulation flow rate, and ash material circulation flow rate; Step 3: By discretization, the optimized configuration results of the opening of the regulating valve of the clean flue gas circulation fan and the opening of the regulating valve of the ash material circulation are obtained under different working conditions. A control rule table is constructed and used as the basis for judgment to implement control actions, so as to realize the intelligent control of the clean flue gas circulation fan and the ash material circulation. Step 4: Further introduce feedback variables into the control rules. By using the feedback correction concept in model predictive control, obtain the model error by comparing the predicted bed pressure drop value output by the bed pressure drop prediction model based on parameter identification with the actual measured bed pressure drop value. Feed back the error to correct the given reference trajectory and input it into the inference module, thereby ensuring the effectiveness of the controller. The actual pressure drop of the absorber bed is used as the feedback value, and the output command of the bed pressure drop control strategy is used as the calculation parameter for the tracking value of the material circulation regulating valve. The output command value of the bed pressure drop control strategy is multiplied by the corresponding regulating valve balance coefficient and then used as the tracking value of the material circulation regulating valve. In order to prevent frequent small movements of the return material trough valve, a valve position holding program is set when the fluctuation amplitude is less than 1. The continuous material level of different ash hoppers is taken as the average value, and the actual material level of each ash hopper is divided by the average value as the balance coefficient of the return ash valve to achieve the balance control of the material level of each ash hopper.

Images