A method and device for desulfurization and denitrification of sintering flue gas

Abstract

The present invention discloses a method and device for desulfurization and denitrification of sintering flue gas. The present invention relates to the technical field of desulfurization and denitrification, and solves the technical problems of insufficient control accuracy, difficulty in real-time and efficient adjustment of desulfurization and denitrification process parameters, resulting in low desulfurization and denitrification efficiency, high ammonia escape rate and increased operating costs. The present invention deploys distributed sensors to collect key parameters in real time, combines mathematical models and PID control algorithms, and can dynamically adjust the desulfurizer injection amount, ammonia escape rate and catalyst operating temperature according to real-time data, thereby achieving precise control of the desulfurization and denitrification process, improving desulfurization and denitrification efficiency, reducing ammonia escape rate, and reducing resource waste. According to different regional NO x The concentration difference and reaction rate constant are used to fine-tune the amount of ammonia sprayed, further improving the denitrification effect and ensuring that the flue gas emissions meet the standards.

Description

Technical Field

[0001] The present invention relates to the technical field of desulfurization and denitrification, and in particular to a method and a device for desulfurization and denitrification of sintering flue gas. Background Art

[0002] The sintering industry produces a large amount of sintering flue gas containing pollutants such as SO2 and NOx during production. Direct discharge of these gases can cause serious environmental pollution. Currently, sintering flue gas desulfurization and denitrification technologies are widely used.

[0003] However, existing methods and devices lack control precision when dealing with complex and changing flue gas conditions, making it difficult to adjust desulfurization and denitrification process parameters in real time and efficiently. This results in low desulfurization and denitrification efficiency, high ammonia escape rates, and increased operating costs. Furthermore, their ability to predict and address potential equipment failures is weak, making it impossible to prevent them before they occur, reducing production continuity and stability. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention provides a method and device for desulfurization and denitrification of sintering flue gas, which solves the problems of insufficient control accuracy, difficulty in real-time and efficient adjustment of desulfurization and denitrification process parameters, resulting in low desulfurization and denitrification efficiency, high ammonia escape rate and increased operating costs.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a method for desulfurization and denitrification of sintering flue gas, the method specifically comprising the following steps:

[0006] Obtain distributed sensor data and preprocess it to obtain preprocessed data;

[0007] According to the chemical reaction principle of flue gas desulfurization and denitrification, a corresponding mathematical model is established, and the pre-processed data is substituted into the predicted value to obtain the predicted value, and the deviation analysis signal is generated by comparing it with the actual value;

[0008] The deviation analysis signal is adjusted according to the PID control algorithm, and the proportional term, integral term, and differential term are calculated and combined to obtain the output and generate the control information;

[0009] Obtain NOx concentrations at installation points in different regions and compare them with the average concentration to identify abnormal installation points. Calculate the adjustment ratio corresponding to their NOx concentrations and then use this ratio as a standard to adjust the ammonia injection rate and generate adjustment information.

[0010] At the same time, the reaction rates of different areas are monitored and compared with the preset values ​​to screen out the areas to be adjusted. The corresponding rate ratio is calculated based on the reaction rate of the area to be adjusted, and the ammonia injection amount is adjusted secondary based on it to generate ammonia injection amount adjustment information.

[0011] As a further solution of the present invention, the specific method of obtaining the preprocessed data is:

[0012] Obtain distributed sensor data, which is deployed at key locations of sintering flue gas emission pipelines and related equipment. The distributed sensor data includes flue gas flow, SO2 concentration, NOx concentration, temperature, and pressure;

[0013] Preprocessing operations include filtering, denoising, and normalization.

[0014] As a further solution of the present invention, the specific method of establishing the corresponding mathematical model based on the chemical reaction principle of flue gas desulfurization and denitrification is:

[0015] According to the chemical reactions in the desulfurization and denitrification process, a mathematical relationship based on the law of conservation of mass is established, and related equations are established based on the principles of heat and mass transfer. According to the data characteristics and theoretical analysis results, a nonlinear regression model is selected as the standard;

[0016] The linear relationship between the desulfurizer injection amount, ammonia escape rate and catalyst operating temperature and the flue gas flow rate, SO2 concentration, NOx concentration, temperature and pressure parameters is analyzed and expressed by the linear formula y=β0+β1x1+β2x2+…+β n x n , where y represents the desulfurization agent injection amount, ammonia escape rate or catalyst operating temperature dependent variable, x1, x2, ..., x n represents the flue gas flow rate, SO2 concentration, NOx concentration, temperature, and pressure independent variables, β0, β1, ..., β n is the parameter to be estimated and the mathematical model is obtained.

[0017] As a further solution of the present invention, the specific method of generating the deviation analysis signal by comparing with the actual value is:

[0018] The pre-processed data is substituted into the mathematical model, and the real-time measured desulfurization efficiency or outlet SO2 concentration is used as the feedback signal to obtain the corresponding predicted value. At the same time, the actual value and the predicted value are subjected to deviation analysis. If there is a deviation between the two, a deviation analysis signal is generated. Otherwise, if there is no deviation between the two, a normal monitoring signal is generated.

[0019] As a further solution of the present invention, the specific method of generating the control information is:

[0020] Calculate the proportional term, P = K p e(t),k p is the proportional coefficient, e(t) is the deviation between the predicted value and the actual value;

[0021] Calculate the integral term, Where Δt is the sampling time interval, k iis the integration coefficient;

[0022] Calculate the differential term, K d is the differential coefficient;

[0023] The proportional term, integral term, and differential term obtained from the above analysis are combined to obtain the controller output u(t), where And based on the obtained output u(t), the corresponding control information is generated.

[0024] As a further solution of the present invention, the specific method of generating the adjustment information is:

[0025] Obtain the installation point number of the NOx concentration sensor in different areas, denoted as a, where a = 1, 2, ..., b, where b represents the number of installation points. At the same time, obtain the NOx concentration data corresponding to each installation point, calculate the average concentration, compare the concentrations at each point, and screen out abnormal installation points, denoted as h, where h = 1, 2, ..., u, where u represents the number of abnormal installation points. At the same time, obtain the ratio of the NOx concentration at the abnormal installation point to the average concentration, and record it as the adjustment ratio;

[0026] Obtain the ammonia injection amount of all normal installation points, calculate its average as the standard value, and then adjust the corresponding ammonia injection amount according to the adjustment ratio of abnormal installation points to generate adjustment information.

[0027] As a further solution of the present invention, the specific method of screening to obtain the region to be adjusted is:

[0028] Obtain all monitoring areas and the corresponding reaction efficiency of the monitoring areas at the same time, and compare the obtained reaction efficiency with the preset value. If the reaction efficiency of the monitoring area is greater than the preset value, no processing will be done. On the contrary, if the reaction efficiency of the monitoring area is lower than the preset value, the corresponding monitoring area will be marked as the area to be adjusted, and a secondary adjustment analysis will be performed on the area to be adjusted.

[0029] As a further solution of the present invention, the specific method of generating the ammonia injection amount adjustment information is:

[0030] Obtain the area to be regulated and determine its reaction rate constant k through experiments. At the same time, obtain the reaction rate constant of the monitoring area and calculate the average value as the standard rate. Based on the standard rate, denitrification efficiency and the total amount of NOx in the flue gas, calculate the benchmark ammonia injection amount.

[0031] Compare the reaction rate constant of the area to be adjusted with the standard rate to obtain the rate ratio, and adjust the ammonia injection amount of the area to be adjusted according to the formula = reference ammonia injection amount × (1 + rate ratio) to generate ammonia injection amount adjustment information.

[0032] A sintering flue gas desulfurization and denitrification device, comprising a desulfurization reactor module, a distributed sensor module, a central control module and a control information output module;

[0033] Desulfurization reactor module, which is used to store the sintering flue gas generated by the sintering machine, obtain the corresponding flue gas data, and transmit it to the central control module;

[0034] Distributed sensor module, which is used to obtain sensor data installed in different areas of the desulfurization reactor. The sensor data includes SO2 concentration, NOx concentration, temperature, and pressure, and transmits the obtained sensor data to the central control module;

[0035] The central control module is used to analyze the acquired sensor data. By analyzing historical data and processing it using a machine learning algorithm, a mathematical model is generated between the flue gas flow rate, SO2 concentration, NOx concentration, temperature, pressure parameters and the desulfurizer injection amount, ammonia escape rate and catalyst operating temperature. The obtained pre-processed data is substituted into the mathematical model to obtain a predicted value, which is compared with the actual value to generate a deviation analysis signal;

[0036] Then, the deviation analysis signal is processed by the PID control algorithm, and the proportional term, integral term, and differential term are calculated respectively. The combined output is obtained to generate the control information, which is then transmitted to the control information output module.

[0037] It is also used to compare the NOx concentration in different areas with the average concentration to screen out abnormal installation points, and calculate the corresponding adjustment ratio. At the same time, it generates adjustment information based on the adjustment ratio and transmits it to the control information output module. Then, it compares and analyzes the reaction rates in different areas to screen out the areas to be adjusted. At the same time, it adjusts the ammonia injection amount according to the reaction rate constant, generates ammonia injection amount adjustment information, and transmits it to the control information output module.

[0038] The control information output module is used to display the acquired control information, adjustment information and ammonia injection amount adjustment information to the corresponding operator.

[0039] The present invention provides a method and apparatus for desulfurization and denitrification of sintering flue gas. Compared with the prior art, it has the following advantages:

[0040] The present invention deploys distributed sensors to collect key parameters in real time, and combines mathematical models and PID control algorithms to dynamically adjust the desulfurizer injection amount, ammonia escape rate and catalyst operating temperature according to real-time data, thereby achieving precise control of the desulfurization and denitrification process, improving the desulfurization and denitrification efficiency, reducing the ammonia escape rate, and reducing resource waste. According to the NOx concentration differences and reaction rate constants in different regions, the ammonia injection amount is finely adjusted to further improve the denitrification effect and ensure that the flue gas emissions meet the standards.

[0041] The present invention combines historical data with machine learning models to accurately predict potential faults such as catalyst aging and bag filter blockage, and sets reasonable thresholds to issue early warnings, allowing staff to take measures in advance, reduce the risk of equipment failure, reduce maintenance costs, and ensure production continuity and stability. At the same time, it integrates functions such as data collection, analysis, control, and fault prediction, achieving collaborative optimization between various links and improving the overall operating efficiency of the sintering flue gas desulfurization and denitrification system. Compared with traditional distributed systems, it has a higher level of intelligence and better environmental and economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a diagram of the steps and methods of the present invention;

[0043] Figure 2 This is a module diagram of the device of the present invention. DETAILED DESCRIPTION

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the 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.

[0045] Example 1

[0046] See also Figure 1 The present application provides a method for desulfurization and denitrification of sintering flue gas, which specifically includes the following steps:

[0047] Distributed sensors are deployed at key locations of sintering flue gas emission pipelines and related equipment, and are used to collect key parameters such as flue gas flow, SO2 concentration, NOx concentration, temperature, and pressure in real time.

[0048] Step 1: Preprocess the sensor data to obtain preprocessed data, and the preprocessing operations include filtering, denoising, and normalization. The preprocessed data is transmitted to the central control system through a wired or wireless communication network.

[0049] Step 2: Based on the chemical reaction principles of flue gas desulfurization and denitrification and actual operating experience, a mathematical model is established between the flue gas flow rate, SO2 concentration, NOx concentration, temperature, pressure parameters and the desulfurizer injection amount, ammonia escape rate and catalyst operating temperature. A large amount of historical data is analyzed and a machine learning algorithm is applied to optimize the model parameters. The specific method of establishing the mathematical model is as follows:

[0050] Collect a large amount of on-site operation data, including flue gas flow, SO2 concentration, NOx concentration, temperature, pressure parameters under different operating conditions, as well as the corresponding desulfurizer injection amount, ammonia escape rate of the SCR denitrification system and catalyst operating temperature data. The data collected here are pre-processed data;

[0051] According to the chemical reactions in the desulfurization and denitrification process, a mathematical relationship based on the law of conservation of mass is established. For example, in the desulfurization process, based on the reaction equation of sulfur dioxide and desulfurizer, the theoretical relationship between the amount of desulfurizer injection and the sulfur dioxide concentration can be derived, and the influence of flue gas flow on the reaction is taken into account. At the same time, the influence of temperature and pressure on the desulfurization and denitrification reaction is considered. Based on the principles of heat and mass transfer, relevant equations are established. For example, changes in temperature will affect the reaction rate and chemical equilibrium, and changes in pressure will affect the solubility of the gas and the reaction kinetics. According to the data characteristics and theoretical analysis results, a suitable mathematical model form is selected. Common models include linear regression models, nonlinear regression models (such as polynomial regression, exponential regression, etc.), artificial neural network models, support vector machine models, etc., and the regression line model is selected in this application;

[0052] The relationship between the desulfurizer injection amount, ammonia escape rate and catalyst operating temperature and the flue gas flow rate, SO2 concentration, NOx concentration, temperature and pressure parameters is analyzed. Here, the linear relationship is analyzed as an example and expressed by the linear relationship formula: y=β0+β1x1+β2x2+…+β n x n , where y represents the desulfurization agent injection amount, ammonia escape rate or catalyst operating temperature dependent variable, x1, x2, ..., x n represents the flue gas flow rate, SO2 concentration, NOx concentration, temperature, and pressure independent variables, β0, β1, ..., β n is the parameter to be estimated, and the parameter here is estimated by the least square method. The established model is verified and optimized using the test set to obtain the mathematical model;

[0053] Step 3: Substitute the obtained pre-processed data into the mathematical model, use the real-time measured desulfurization efficiency or outlet SO2 concentration as the feedback signal, obtain the corresponding predicted value, and perform deviation analysis on the actual value and the predicted value. If there is a deviation between the two, a deviation analysis signal is generated. Otherwise, if there is no deviation between the two, a normal monitoring signal is generated. At the same time, the obtained deviation analysis signal is adjusted, specifically through the PID control algorithm. The specific adjustment method is as follows:

[0054] Calculate the proportional term, integral term, and differential term, and the specific calculation formula is as follows:

[0055] Proportional term: P = K p e(t),k pis the proportional coefficient, which is used to amplify or reduce the impact of the deviation. e(t) is the deviation between the predicted value and the actual value, which is obtained by subtracting the actual value from the predicted value.

[0056] Integral item: Where Δt is the sampling time interval, k i is the integral coefficient, which is used to eliminate the steady-state error of the system;

[0057] Differential term: K d is the differential coefficient, which is used to predict the changing trend of the deviation;

[0058] The proportional term, integral term, and differential term obtained from the above analysis are combined to obtain the controller output u(t), where Based on the obtained output u(t), corresponding control information is generated, and the control information here specifically refers to the pump speed or valve opening parameter.

[0059] Step 4: Then obtain the installation points of NOx concentration sensors in different areas and label them as a, where a=1, 2, ..., b, where b represents the number of installation points. At the same time, obtain NOx concentration data corresponding to different installation points. Then compare the obtained NOx concentration with the average concentration, where the average concentration is expressed as the concentration calculated based on different areas. Screen out abnormal installation points and label them as h, where h=1, 2, ..., u, where u represents the number of abnormal installation points. At the same time, obtain the ratio of the NOx concentration at the abnormal installation point to the average concentration and record it as the adjustment ratio;

[0060] Next, the ammonia injection rates corresponding to all normal installation points are obtained and their corresponding average ammonia injection rates are calculated. This average is used as the standard value. The ammonia injection rates of abnormal installation points are then adjusted based on the adjustment ratio corresponding to the abnormal installation points. For example, if the NOx concentration in a certain area is 1.2 times the average concentration, the ammonia injection rate in that area can be set to 1.2 times the average ammonia injection rate. Adjustment information is generated, and efficiency monitoring and analysis are performed on the monitoring area corresponding to the adjustment information.

[0061] Obtain all monitoring areas, and the monitoring areas here represent the areas after the ammonia injection amount is adjusted. At the same time, obtain the reaction efficiency corresponding to the monitoring areas, and compare the obtained reaction efficiency with the preset value. The specific value of the preset value is determined by the operator according to the actual situation. If the reaction efficiency of the monitoring area is greater than the preset value, no processing is performed. On the contrary, if the reaction efficiency of the monitoring area is lower than the preset value, the corresponding monitoring area is marked as the area to be adjusted, and a secondary adjustment analysis is performed on the area to be adjusted;

[0062] All the regions to be adjusted are obtained, and the reaction rate constant k corresponding to the regions to be adjusted is obtained through experimental analysis. The reaction rate constant is obtained as follows:

[0063] Multiple sampling points were set up in different areas of the reactor to collect flue gas samples to analyze the changes in NOx and ammonia concentrations. At the same time, variables in the experiment, such as temperature, flue gas flow rate, catalyst type and loading, were controlled to ensure that other conditions, except for regional factors, remained consistent.

[0064] At different reaction time points, flue gas samples were collected from each sampling point, the NOx and ammonia concentrations were analyzed, and the data were recorded. Multiple groups of experiments were conducted by changing the initial NOx concentration, ammonia concentration and other conditions;

[0065] According to the experimental data, the reaction kinetics equation The reaction rate constants k in different regions were calculated by fitting the data using methods such as nonlinear regression;

[0066] Then, all monitoring areas are obtained, and the monitoring areas here are represented as areas where the reaction rate is greater than the preset value. At the same time, the reaction rate constants corresponding to the monitoring areas are obtained, and the average of the reaction rate constants is calculated and recorded as the standard rate. Then, based on the standard rate, the denitrification efficiency and the total amount of NOx in the flue gas are designed, and a preliminary baseline ammonia injection amount is calculated. The reaction rate of the area to be adjusted is calculated by ratioing the standard rate to obtain the rate ratio, specifically the rate ratio = reaction rate / standard rate. At the same time, the ammonia injection amount of the area to be adjusted is adjusted according to the obtained rate ratio, specifically according to the formula = baseline ammonia injection amount × (1 + rate ratio). For example, if the reaction rate constant of a certain area is 80% of the average rate constant, the ammonia injection amount of the area can be increased to 1.2-1.5 times the baseline ammonia injection amount. For areas where the reaction rate constant is greater than the average rate constant, the ammonia injection amount is reduced accordingly, such as adjusting the ammonia injection amount to 0.8-0.9 times the baseline ammonia injection amount, and generating ammonia injection amount adjustment information.

[0067] Example 2

[0068] As the second embodiment of the present invention, it is implemented on the basis of the first embodiment, and differs from the first embodiment in the following aspects:

[0069] This example combines historical data with a machine learning model to predict potential faults such as catalyst aging and bag filter blockage. The specific processing method is as follows:

[0070] Collect historical operating data related to catalysts and bag filters, including but not limited to flue gas flow, temperature, pressure, composition (such as SO2 and NOx concentrations), ammonia injection volume, absorption liquid circulation volume, equipment operating hours, maintenance records, etc. This data can be obtained from distributed sensor networks, continuous emission monitoring systems, and equipment operation logs;

[0071] Clean the collected data to remove duplicate, erroneous, or incomplete data records. For missing values, choose an appropriate filling method based on the data characteristics, such as mean filling, median filling, or interpolation filling based on similar working conditions.

[0072] Extract representative features from the raw data based on the characteristics of the problem and domain knowledge. For example, for catalyst aging prediction, features such as catalyst usage time, cumulative flue gas volume treated, residence time at different temperature ranges, and ammonia escape rate can be extracted.

[0073] For bag filter blockage prediction, features such as flue gas flow rate change rate, bag pressure difference, cleaning frequency, and dust concentration can be extracted.

[0074] At the same time, some categorical features are encoded so that they can be processed by machine learning models;

[0075] Select an appropriate machine learning model based on the data characteristics and the nature of the prediction task. Common models include decision trees, random forests, support vector machines, and artificial neural networks (such as multi-layer perceptrons, recurrent neural networks and their variants, LSTMs, and GRUs). For data with time series characteristics, recurrent neural networks and their variants are generally better able to capture temporal dependencies in the data and are suitable for predicting potential faults that develop over time, such as catalyst aging and bag filter blockage.

[0076] The preprocessed data is divided into a training set and a test set, usually in a ratio of 7:3 or 8:2. The selected machine learning model is trained using the training set. By adjusting the model parameters (such as the weights and biases of the neural network), the model is able to minimize the loss function (such as mean square error, cross entropy, etc.) and thus learn the underlying patterns in the data. During the training process, some optimization algorithms such as stochastic gradient descent, Adagrad, Adadelta, etc. can be used to accelerate the convergence of the model. In order to prevent the model from overfitting, regularization techniques (such as L1 and L2 regularization) and Dropout can be used.

[0077] Use the test set to evaluate the trained model using metrics such as root mean square error (RMSE), mean absolute error (MAE), accuracy, recall, and F1 value. For fault prediction, since there are usually fewer fault samples and this is an imbalanced data problem, metrics such as F1 value that comprehensively consider both accuracy and recall are more important. Use these metrics to judge the performance of the model. If the model performance does not meet the requirements, optimize the model.

[0078] During actual operation, the distributed sensor network collects real-time equipment operating data, which is then processed and feature extracted using data preprocessing methods. This real-time data is fed into a trained machine learning model to predict potential faults such as catalyst aging and bag filter blockage.

[0079] Reasonable thresholds are set for prediction results based on actual production needs and the safe operating range of the equipment. For example, when the predicted degree of catalyst aging exceeds a certain threshold, it indicates that the catalyst may be about to fail and needs to be inspected and replaced in a timely manner. When the predicted probability of bag filter blockage exceeds a certain threshold, an early warning signal is issued to prompt staff to clean the dust or check the bag condition in a timely manner.

[0080] Example 3

[0081] As the third embodiment of the present invention, the focus is on combining the implementation processes of the first and second embodiments.

[0082] Example 4

[0083] See also Figure 2 The present application provides a sintering flue gas desulfurization and denitrification device, which includes a desulfurization reactor module, a distributed sensor module, a central control module and a control information output module, and is combined with Figure 2 It can be known that the functional modules are electrically connected in a unidirectional manner.

[0084] Desulfurization reactor module, which is used to store the sintering flue gas generated by the sintering machine, obtain the corresponding flue gas data, and transmit it to the central control module;

[0085] Distributed sensor module, which is used to obtain sensor data installed in different areas of the desulfurization reactor. The sensor data includes SO2 concentration, NOx concentration, temperature, and pressure, and transmits the obtained sensor data to the central control module;

[0086] The central control module is used to analyze the acquired sensor data. By analyzing the historical data and processing it using a machine learning algorithm, a mathematical model is generated between the flue gas flow rate, SO2 concentration, NOx concentration, temperature, pressure and other parameters and the desulfurizer injection amount, ammonia escape rate and catalyst operating temperature. The processing method here is similar to the processing process of step 2 in Example 1. The obtained pre-processed data is substituted into the mathematical model to obtain a predicted value, which is compared with the actual value to generate a deviation analysis signal.

[0087] Then, the deviation analysis signal is processed by the PID control algorithm, and the proportional term, integral term, and differential term are calculated respectively. The combined output is obtained to generate the control information, which is then transmitted to the control information output module.

[0088] It is also used to compare the NOx concentration in different areas with the average concentration to screen out abnormal installation points, calculate the corresponding adjustment ratio, and generate adjustment information based on the adjustment ratio and transmit it to the control information output module. The specific processing method is the same as the processing process of step three in Example 1. Then, the reaction rates in different areas are compared and analyzed to screen out the areas to be adjusted. At the same time, the ammonia injection amount is adjusted according to the reaction rate constant to generate ammonia injection amount adjustment information and transmit it to the control information output module. The specific processing method is the same as the processing process of step four in Example 1.

[0089] The control information output module is used to display the acquired control information, adjustment information and ammonia injection amount adjustment information to the corresponding operator.

[0090] Some of the data in the above formulas are calculated based on their numerical values ​​and are not substituted into parameter units for calculation. At the same time, the contents not described in detail in this specification belong to the existing technology known to those skilled in the art.

[0091] 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. A method for desulfurization and denitrification of sintering flue gas, characterized in that: The method specifically comprises the following steps: Obtain distributed sensor data and preprocess it to obtain preprocessed data; According to the chemical reaction principle of flue gas desulfurization and denitrification, a corresponding mathematical model is established, and the pre-processed data is substituted into the predicted value to obtain the predicted value, and the deviation analysis signal is generated by comparing it with the actual value; The deviation analysis signal is adjusted according to the PID control algorithm, and the proportional term, integral term, and differential term are calculated and combined to obtain the output and generate the control information; Get the NO of installation points in different areas x The concentration is compared with the average concentration to screen out abnormal installation points and calculate their NO x The concentration corresponds to the adjustment ratio, and then the ammonia injection amount is adjusted based on it to generate adjustment information. The specific method is as follows: Get different area NO x The installation point of the concentration sensor is marked as a, and a=1, 2, ..., b, where b represents the number of installation points. At the same time, the NO corresponding to different installation points is obtained. x Concentration data, and calculate the average concentration, compare the concentration of each point, filter out the abnormal installation points and label them as h, and h = 1, 2, ..., u, where u represents the number of abnormal installation points, and obtain the abnormal installation point NO x The ratio of the concentration to the average concentration is recorded as the adjustment ratio; Obtain the ammonia injection amount of all normal installation points, calculate its average as the standard value, and then adjust the corresponding ammonia injection amount based on the adjustment ratio of the abnormal installation points to generate adjustment information; At the same time, the reaction rates of different areas are monitored and compared with the preset values ​​to select the areas to be adjusted. The corresponding rate ratio is calculated based on the reaction rate of the area to be adjusted, and the ammonia injection amount is adjusted secondary based on it to generate ammonia injection amount adjustment information. The specific method is as follows: Obtain the area to be adjusted, determine its reaction rate constant k through experiments, and at the same time, obtain the reaction rate constant of the monitoring area, calculate the average value as the standard rate, and calculate the reaction rate according to the standard rate, denitrification efficiency and NO in flue gas. x Total amount, calculate the benchmark ammonia injection amount; Compare the reaction rate constant of the area to be adjusted with the standard rate to obtain the rate ratio, and adjust the ammonia injection amount of the area to be adjusted according to the formula = reference ammonia injection amount × (1 + rate ratio) to generate ammonia injection amount adjustment information.

2. The method for desulfurization and denitrification of sintering flue gas according to claim 1, characterized in that: The specific method of obtaining the preprocessed data is: Obtain distributed sensor data, and distributed sensors are deployed in key locations of sintering flue gas emission pipelines and related equipment. At the same time, distributed sensor data includes flue gas flow, SO2 concentration, NO x concentration, temperature, pressure; Preprocessing operations include filtering, denoising, and normalization.

3. The method for desulfurization and denitrification of sintering flue gas according to claim 1, characterized in that: The specific method of establishing the corresponding mathematical model based on the chemical reaction principle of flue gas desulfurization and denitrification is as follows: According to the chemical reactions in the desulfurization and denitrification process, a mathematical relationship based on the law of conservation of mass is established, and related equations are established based on the principles of heat and mass transfer. According to the data characteristics and theoretical analysis results, a nonlinear regression model is selected as the standard; Analyze the desulfurization agent injection amount, ammonia escape rate and catalyst operating temperature and flue gas flow, SO2 concentration, NO x The linear relationship between parameters such as concentration, temperature, and pressure is expressed by the linear formula y=β0+β1x1+β2x2+…+β n x n , where y represents the dependent variable such as desulfurizer injection amount, ammonia escape rate or catalyst operating temperature, x1, x2, ..., x n Indicates flue gas flow, SO2 concentration, NO x Independent variables such as concentration, temperature, and pressure, β0, β1, ..., β n is the parameter to be estimated and the mathematical model is obtained.

4. The method for desulfurization and denitrification of sintering flue gas according to claim 1, characterized in that: The specific method of generating the deviation analysis signal by comparing with the actual value is: The pre-processed data is substituted into the mathematical model, and the real-time measured desulfurization efficiency or outlet SO2 concentration is used as the feedback signal to obtain the corresponding predicted value. At the same time, the actual value and the predicted value are subjected to deviation analysis. If there is a deviation between the two, a deviation analysis signal is generated. Otherwise, if there is no deviation between the two, a normal monitoring signal is generated.

5. The method for desulfurization and denitrification of sintering flue gas according to claim 1, characterized in that: The specific method of generating the control information is: Calculate the proportional term, P = K p e(t),k p is the proportional coefficient, e(t) is the deviation between the predicted value and the actual value; Calculate the integral term, Where Δt is the sampling time interval, k i is the integration coefficient; Calculate the differential term, K d is the differential coefficient; The proportional term, integral term, and differential term obtained from the above analysis are combined to obtain the controller output u(t), where And based on the obtained output u(t), the corresponding control information is generated.

6. The method for desulfurization and denitrification of sintering flue gas according to claim 1, characterized in that: The specific method of screening to obtain the area to be adjusted is: Obtain all monitoring areas and the corresponding reaction efficiency of the monitoring areas at the same time, and compare the obtained reaction efficiency with the preset value. If the reaction efficiency of the monitoring area is greater than the preset value, no processing will be done. On the contrary, if the reaction efficiency of the monitoring area is lower than the preset value, the corresponding monitoring area will be marked as the area to be adjusted, and a secondary adjustment analysis will be performed on the area to be adjusted.

7. A sintering flue gas desulfurization and denitrification device, used to implement a sintering flue gas desulfurization and denitrification method according to any one of claims 1 to 6, characterized in that: The device includes a desulfurization reactor module, a distributed sensor module, a central control module and a control information output module; Desulfurization reactor module, which is used to store the sintering flue gas generated by the sintering machine, obtain the corresponding flue gas data, and transmit it to the central control module; Distributed sensor module, which is used to obtain sensor data installed in different areas of the desulfurization reactor, and the sensor data includes SO2 concentration, NO x concentration, temperature, and pressure, while transmitting the acquired sensor data to the central control module; Central control module, which is used to analyze the acquired sensor data, analyze historical data, and use machine learning algorithms to generate flue gas flow, SO2 concentration, NO x A mathematical model is established between parameters such as concentration, temperature, and pressure and the amount of desulfurizer injected, ammonia escape rate, and catalyst operating temperature. The pre-processed data is substituted into the mathematical model to obtain a predicted value, which is then compared with the actual value to generate a deviation analysis signal. Then, the deviation analysis signal is processed by the PID control algorithm, and the proportional term, integral term, and differential term are calculated respectively. The combined output is obtained to generate the control information, which is then transmitted to the control information output module. Also used to set different area NO x The concentration is compared with the average concentration to screen out abnormal installation points, and the corresponding adjustment ratio is calculated. At the same time, adjustment information is generated based on the adjustment ratio and transmitted to the control information output module. Then, the reaction rates of different areas are compared and analyzed to screen out the areas to be adjusted. At the same time, the ammonia injection amount is adjusted according to the reaction rate constant, and the ammonia injection amount adjustment information is generated and transmitted to the control information output module. The control information output module is used to display the acquired control information, adjustment information and ammonia injection amount adjustment information to the corresponding operator.

Images