Biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method

Abstract

The invention relates to the technical field of energy conservation and environmental protection, and discloses a biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method.The biomass oxygen-enriched combustion optimization and pollutant intelligent control system comprises a boiler temperature and pressure prediction system, a fuel prediction regulation and control system, a primary and secondary air prediction regulation and control system, a flue gas amount prediction system, a flue gas recirculation regulation and control system and a desulfurization and denitrification intelligent control system; the fuel prediction regulation and control system comprises a fuel identification module, a fuel heat value prediction module and a fuel demand prediction module; the primary and secondary air predictive regulation and control system comprises a primary air calculation module and a secondary air calculation module; the flue gas quantity prediction system comprises an induced draft fan control module and a flue gas prediction module; the flue gas recirculation regulation and control system comprises a recirculation quantity prediction module and a supplement quantity prediction module; the desulfurization and denitrification intelligent control system comprises a pollutant prediction module, a desulfurization agent prediction module and a denitrification agent prediction module. According to the invention, dynamic optimization and accurate control of the biomass boiler under the oxygen-enriched combustion condition can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of energy conservation and environmental protection, and in particular to a biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method. Background Art

[0002] Biomass oxygen-enriched combustion technology is a method of burning by using oxygen-rich gas (oxygen concentration is higher than the standard content of about 21% in the air) to replace ordinary air. This technology is mainly used in the combustion process of biomass fuels. It increases the flame temperature and accelerates the combustion process by increasing the oxygen concentration, thereby significantly improving the combustion efficiency. In addition, this efficient combustion method allows the carbon and sulfur in the fuel to be oxidized more thoroughly, effectively reducing the emissions of carbon monoxide, sulfides and particulate matter. At the same time, biomass oxygen-enriched combustion is one of the important paths to achieve low-cost carbon capture and an important technical method for green carbon production in biomass power plants.

[0003] However, in actual application, biomass boilers that implement oxygen-enriched combustion technology face many challenges. First, compared with traditional coal-fired boilers, unlike traditional coal-fired boilers with stable operating conditions and a single fuel source, the operating conditions of biomass boilers are more complex and changeable. Biomass fuels come from a wide range of sources and have large differences in composition, which leads to unstable boiler operation and a high concentration of carbon dioxide produced after combustion. Second, it is difficult to accurately control key parameters such as the amount of fuel input, the combustion temperature and pressure in the boiler, the ratio of oxygen to air, and the amount of desulfurizer and denitrifier used. At present, there is a lack of efficient and advanced control technologies and methods. These problems not only have an adverse effect on the combustion effect, but also increase the workload of operators, increase operating costs, and bring potential safety hazards.

[0004] In summary, although biomass oxygen-enriched combustion technology has shown great potential in terms of environmental protection and economic benefits, in order to give full play to its advantages, the above-mentioned technical difficulties need to be overcome. Summary of the invention

[0005] The present invention aims to provide a biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method, which can solve the problems of combustion system and pollutant intelligent control under variable fuel, multiple operating conditions and large circulation rate faced by biomass boilers under oxygen-enriched combustion conditions, and realize dynamic optimization and precise control of biomass boilers under oxygen-enriched combustion conditions.

[0006] To achieve the above objectives, the present invention provides the following basic solutions.

[0007] Solution 1

[0008] Biomass oxygen-enriched combustion optimization and pollutant intelligent control system, including boiler temperature and pressure prediction system, fuel prediction and control system, primary and secondary air prediction and control system, flue gas volume prediction system, flue gas recirculation control system and desulfurization and denitrification intelligent control system;

[0009] The boiler temperature and pressure prediction system is used to receive load instructions and predict the boiler operating pressure and boiler operating temperature corresponding to the load; the fuel prediction and control system includes a fuel identification module for confirming the type of fuel, a fuel calorific value prediction module for predicting the calorific value of the fuel, and a fuel demand prediction module for predicting the amount of fuel entering the furnace;

[0010] The primary and secondary air prediction and control system includes a primary air calculation module and a secondary air calculation module; the primary air calculation module takes the amount of fuel fed into the furnace, the operating temperature of the boiler and the historical data of the boiler bed pressure as input to calculate the primary air volume and adjust the opening or frequency of the primary fan in real time; the secondary air calculation module takes the amount of fuel fed into the furnace and the operating temperature of the boiler as input to calculate the secondary air volume and adjust the opening or frequency of the secondary fan in real time;

[0011] The flue gas volume prediction system comprises an induced draft fan control module for controlling the opening or frequency of the flue gas induced draft fan and a flue gas prediction module for predicting the flue gas volume; the flue gas prediction module takes the boiler operating pressure and the opening or frequency of the flue gas induced draft fan as input and takes the predicted flue gas volume as output;

[0012] The flue gas recirculation control system comprises a recirculation volume prediction module and a supplementary volume prediction module; the recirculation volume prediction module takes the flue gas volume as input, and takes the predicted carbon capture flue gas volume and recirculation air volume as output; the supplementary volume prediction module takes the recirculation air volume, primary air volume and secondary air volume as input, and takes the air supplementary volume and pure oxygen supplementary volume as output, and respectively controls the air blower and the oxygen generator in real time according to the output data;

[0013] The desulfurization and denitrification intelligent control system includes a pollutant prediction module, a desulfurizer prediction module and a denitrifier prediction module; the pollutant prediction module takes the fuel type and the flue gas volume as input, and the sulfur compound output and the nitrogen oxide output as output; the desulfurizer prediction module takes the sulfur compound output and the desulfurizer type as input, and the desulfurizer spray amount as output; the denitrifier prediction module takes the nitrogen oxide output and the denitrifier type as input, and the denitrifier spray amount as output.

[0014] Solution 2

[0015] The method for optimizing biomass oxygen-enriched combustion and intelligently controlling pollutants uses the biomass oxygen-enriched combustion optimization and intelligently controlling pollutants control system as described in Scheme 1 to optimize biomass oxygen-enriched combustion and intelligently control pollutants; the method comprises the following steps:

[0016] The boiler temperature and pressure prediction system receives the load instruction and predicts the boiler operating pressure and boiler operating temperature corresponding to the load; the fuel identification module of the fuel prediction and control system confirms the fuel type; the fuel calorific value prediction module of the fuel prediction and control system predicts the fuel calorific value; the fuel demand prediction module of the fuel prediction and control system predicts the amount of fuel entering the furnace;

[0017] The primary air calculation module of the primary and secondary air prediction and control system takes the amount of fuel fed into the furnace, the operating temperature of the boiler and the historical data of the boiler bed pressure as input to calculate the primary air volume and adjust the opening or frequency of the primary fan in real time; The secondary air calculation module of the primary and secondary air prediction and control system takes the amount of fuel fed into the furnace and the operating temperature of the boiler as input to calculate the secondary air volume and adjust the opening or frequency of the secondary fan in real time;

[0018] The induced draft fan control module of the flue gas volume prediction system controls the opening degree or frequency of the induced draft fan; the flue gas prediction module of the flue gas volume prediction system takes the boiler operating pressure and the opening degree or frequency of the induced draft fan as input, predicts the flue gas volume and outputs it;

[0019] The recirculation volume prediction module of the flue gas recirculation control system takes the flue gas volume as input, predicts the carbon capture flue gas volume and the recirculation air volume, and outputs the predictions; the supplementary volume prediction module of the flue gas recirculation control system takes the recirculation air volume, primary air volume, and secondary air volume as input, predicts the air supplementary volume and the pure oxygen supplementary volume, and outputs the predictions, and controls the air fan and the oxygen generator in real time according to the output data;

[0020] The pollutant prediction module of the desulfurization and denitrification intelligent control system takes the fuel type and flue gas volume as input, predicts the sulfur compound output and nitrogen oxide output and outputs; the desulfurizer prediction module of the flue gas recirculation control system takes the sulfur compound output and the desulfurizer type as input, predicts the desulfurizer spray amount and outputs; the denitrifier prediction module of the flue gas recirculation control system takes the nitrogen oxide output and the denitrifier type as input, predicts the denitrifier spray amount and outputs.

[0021] The working principle and advantages of the present invention are:

[0022] The biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method of the present invention can solve the combustion system and pollutant intelligent control problems faced by biomass boilers under oxygen-enriched combustion conditions, such as variable fuels, multiple operating conditions, and large circulation rates, and realize dynamic optimization and precise control of biomass boilers under oxygen-enriched combustion conditions. The key point is: this solution provides a targeted and highly intelligent control solution for biomass power plants, which can realize dynamic adjustment and control of the combustion conditions of biomass boilers under oxygen-enriched combustion conditions and different fuel conditions. This solution integrates a variety of intelligent algorithms, which can realize real-time oxygen replenishment, dynamic adjustment of primary and secondary air, and optimization of the combustion process; each module can form a pre-prediction and post-feedback mechanism, and form an integrated combustion control and pollutant control system coupling model, thereby realizing dynamic intelligent regulation of multiple pollutants in biomass boilers, effectively reducing pollutant emissions, and greatly improving the pollution control level of traditional biomass boilers.

[0023] In particular, first, this solution designs a boiler temperature and pressure prediction system and a fuel prediction and control system, which can cope with dual-variable operating conditions such as load conversion and fuel conversion that are unique to biomass power plants, and solves the combustion system and pollutant intelligent control problems faced by biomass boilers under oxygen-enriched combustion conditions under variable fuels, multiple operating conditions, and large circulation rates. Second, each module in this solution is constructed as a multi-input-single-output model, which can achieve a high degree of prediction accuracy and thus realize precise control of biomass boilers. Third, this solution designs a flue gas recirculation control system and a desulfurization and denitrification intelligent control system, which can be coupled with environmental protection devices to achieve full control of pollutants and help achieve low-cost carbon capture. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a schematic diagram of the overall system structure of an embodiment of the biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method of the present invention;

[0025] Figure 2 It is a schematic diagram of a boiler temperature and pressure prediction module of an embodiment of a biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method of the present invention;

[0026] Figure 3 It is a schematic diagram of the structure of the fuel prediction and control system of the embodiment of the biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method of the present invention;

[0027] Figure 4 It is a structural schematic diagram of the primary and secondary air prediction and control system of the biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method embodiment of the present invention;

[0028] Figure 5 It is a schematic diagram of the structure of a flue gas volume prediction system according to an embodiment of the biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method of the present invention;

[0029] Figure 6 It is a schematic diagram of the structure of the flue gas recirculation control system of the embodiment of the biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method of the present invention;

[0030] Figure 7 It is a schematic diagram of the structure of the desulfurization and denitrification intelligent control system of the biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method embodiment of the present invention. DETAILED DESCRIPTION

[0031] The following is a further detailed description through specific implementation methods:

[0032] The embodiment is basically as shown in the attached Figure 1 Shown: Biomass oxygen-enriched combustion optimization and pollutant intelligent control system, including boiler temperature and pressure prediction system, fuel prediction and control system, primary and secondary air prediction and control system, flue gas volume prediction system, flue gas recirculation control system and desulfurization and denitrification intelligent control system.

[0033] The boiler temperature and pressure prediction system is used to receive load instructions and predict the boiler operating pressure and boiler operating temperature corresponding to the load.

[0034] Specifically, the temperature and pressure prediction system consists of boiler temperature and pressure prediction modules, such as Figure 2 As shown in Figure 2, this module adopts a recurrent neural network structure. During the training process of the recurrent neural network, the relationship between parameters such as boiler load, operating temperature and pressure and future temperature and pressure is learned, such as Figure 2 As shown. This module uses a large amount of historical operation data of biomass boilers as training samples, performs network training and adjusts the network weights based on the verification function (MSE function) until the model error is minimized. The historical operation data includes parameters such as boiler load, boiler operating temperature and pressure of the biomass boiler in the past year, and standardization and normalization methods are used to process the historical operation data. In addition, in order to cope with changes in the operating status of the boiler, this module is also equipped with an online learning mechanism, that is, the network parameters are regularly updated according to the latest operating data to ensure that the prediction model always maintains a high accuracy and adaptability.

[0035] The fuel prediction and control system includes a fuel identification module for confirming the type of fuel, a fuel calorific value prediction module for predicting the calorific value of the fuel, and a fuel demand prediction module for predicting the amount of fuel entering the furnace. Figure 3 shown.

[0036] The fuel prediction and control system is preset with a fuel database; the fuel database is set with a fuel calorific value comparison table, as shown in Table 1, which lists the unit calorific values ​​of biomass fuel base materials. Specifically, the fuel calorific value comparison table includes the unit calorific values ​​of different biomass fuel base materials under different humidity conditions (such as 20% humidity, 50% humidity, 80% humidity) and different structural conditions (such as block structure, powder structure).

[0037] Table 1 Fuel calorific value comparison table (example)

[0038]

[0039]

[0040] When confirming the type of fuel, the fuel recognition module first uses a high-resolution camera to collect fuel images for biomass boilers, and identifies the type of fuel based on an image recognition model; the image recognition model includes an input layer, a first CNN layer, a second CNN layer, a fully connected layer, and an output layer. After the fuel image is collected, an image preprocessing operation is also performed, including: image scaling, scaling all images to a uniform size; normalization processing, scaling the pixel values ​​of all images to the range of [0,1], or standardization processing, converting all images to a range with a mean of 0 and a standard deviation of 1; transformation processing, expanding the image content by methods such as rotation, flipping, and color transformation. The image recognition model is pre-trained using a fuel substrate image set, and the network weights are adjusted based on a verification function (MSE function) until the model error is minimized (i.e., the MSE loss is reduced to less than 0.001). The fuel substrate image set includes: biomass fuel substrate images under different humidity conditions (such as 20% humidity, 50% humidity, 80% humidity) and different structural conditions (such as block structure and powder structure).

[0041] Optionally, after the fuel identification module outputs the fuel type, professionals may be introduced to review the output result to further ensure the accuracy of the identification.

[0042] The fuel types include the types and proportions of biomass fuel substrates; the fuel calorific value prediction module calls the fuel database based on the fuel types to calculate the fuel calorific value. Specifically, the fuel calorific value prediction module also uses unique hot encoding to encode the type of biomass fuel substrate and its corresponding unit calorific value based on the fuel database, which can eliminate the order relationship between categories to facilitate adaptation to machine learning algorithms, simplify feature engineering, avoid information loss, and improve model performance. The fuel calorific value prediction module is provided with a random forest regression model for calculating the fuel calorific value.

[0043] The fuel demand prediction module adopts a support vector regression model structure using a linear kernel function, forms a basic model by training historical data of fuel heat, boiler load, boiler operating temperature and pressure, and fuel input, and updates it through self-learning according to real-time data. Specifically, this module uses a large amount of historical operating data of biomass boilers as training samples, performs network training, corrects model parameters, and uses the determination coefficient (R 2 ) to evaluate the training effect until the coefficient of determination (R 2 ) reaches about 0.95, the training is terminated. The historical operation data includes parameters such as boiler load, boiler operating temperature and pressure, and fuel input amount of the biomass boiler in the past year, and the historical operation data are processed using standardization and normalization methods.

[0044] The primary and secondary air prediction and control system includes a primary air calculation module, a secondary air calculation module, a primary fan control module and a secondary fan control module. Figure 4 shown.

[0045] The primary air calculation module uses the historical data of the amount of fuel fed into the furnace, the operating temperature of the boiler, and the bed pressure of the boiler as input to calculate the primary air volume and adjust the opening or frequency of the primary fan in real time; the secondary air calculation module uses the historical data of the amount of fuel fed into the furnace and the operating temperature of the boiler as input to calculate the secondary air volume and adjust the opening or frequency of the secondary fan in real time. Specifically, both the primary air calculation module and the secondary air calculation module use the least squares method to calculate the air volume.

[0046] In the primary air calculation module, the historical data is first standardized and a multivariate linear regression model is established. The basic formula of the model is: A1 = k1·F+k2·T+k3·P+b; where k1, k2, k3 are model coefficients, b is a constant, A1 is the primary air volume, F, T, P are the fuel input, boiler operating temperature and boiler bed pressure respectively. Based on the historical data, the least squares method is used to estimate the values ​​of k1, k2, k3 and b to obtain a complete multivariate linear regression model for real-time calculation and regulation of the primary air volume.

[0047] Furthermore, the primary wind calculation module regulates the opening or frequency of the primary fan through the primary fan control module.

[0048] In the secondary air calculation module, the historical data is first standardized, and the secondary air calculation model is constructed according to the combustion principle. The basic formula of the model is A2 = k·F·(1-e -λT )-A1. In the formula, k and λ are model coefficients, A1 is the primary air volume, F and T are the fuel input and boiler operating temperature respectively. Based on historical data, the least squares method is used to estimate the values ​​of k and λ, and a complete secondary air calculation model is obtained to perform real-time calculation and regulation of secondary air volume.

[0049] Furthermore, the secondary air calculation module regulates the opening or frequency of the secondary fan through the secondary fan control module.

[0050] Both the primary fan control module and the secondary fan control module are regulated based on the PID model.

[0051] The smoke volume prediction system includes an induced draft fan control module for controlling the opening degree or frequency of the smoke induced draft fan and a smoke prediction module for predicting the smoke volume. Figure 5 shown.

[0052] The flue gas prediction module takes the boiler operating pressure and the opening or frequency of the flue gas induced draft fan as input and takes the predicted flue gas volume as output.

[0053] In the flue gas prediction module, a neural network model is constructed (including two LSTM layers and one fully connected layer, each LSTM layer contains 128 neurons). The neural network model is pre-trained using a data set consisting of historical data on boiler operating pressure, flue gas induced draft fan opening or frequency, and flue gas volume, and the network weights are adjusted based on the validation function (MSE function) until the model error is minimized (i.e., the MSE loss is reduced to below 0.001).

[0054] The induced draft fan control module controls the opening degree / frequency of the flue gas induced draft fan in real time based on the PID model according to the boiler operating pressure output by the boiler pressure stabilization prediction module.

[0055] The flue gas recirculation control system includes a recirculation amount prediction module and a supplementary amount prediction module, such as Figure 6 shown.

[0056] The recirculation volume prediction module takes the flue gas volume as input and the predicted carbon capture flue gas volume and recirculation air volume as output; the supplementary volume prediction module takes the recirculation air volume, primary air volume and secondary air volume as input, and takes the air supplementary volume and pure oxygen supplementary volume as output, and respectively regulates the air fan and oxygen generator in real time according to the output data.

[0057] The recirculation volume prediction module is provided with a CNN-LSTM network structure (comprising a CNN layer, an LSTM layer and a fully connected layer, wherein the LSTM layer comprises 128 neurons). The network structure is pre-trained with a data set consisting of historical data of total flue gas volume, recirculation air volume, and carbon capture flue gas volume, and the network weights are adjusted based on a validation function (MSE function) until the model error is minimized (i.e., the MSE loss is reduced to below 0.001). The network structure is capable of learning the spatiotemporal features in historical data, wherein the CNN layer is responsible for extracting local features, and the LSTM layer processes time series data, and then jointly constructs a dynamic prediction model with real-time update function.

[0058] The replenishment prediction module is provided with a GRU network structure (including two GRU layers and one fully connected layer, each GRU layer including 128 neurons). The network structure is pre-trained using a data set consisting of historical data of total smoke volume, recirculation air volume, air replenishment volume and oxygen volume, and the network weights are adjusted based on the validation function (MSE function) until the model error is minimized (i.e., the MSE loss is reduced to below 0.001). The GRU network structure has excellent long-term dependency capture capabilities and low computational complexity, and can be adapted to the dynamic prediction tasks in this application scenario.

[0059] The flue gas recirculation control system also includes a PID control module; the PID control module adjusts the opening or frequency of the air fan in real time according to the output of the supplementary amount prediction module to control the air supplementary amount, thereby optimizing the combustion process.

[0060] The desulfurization and denitrification intelligent control system includes a pollutant prediction module, a desulfurization agent prediction module and a denitrification agent prediction module. Figure 7 shown.

[0061] The pollutant prediction module takes the fuel type and flue gas volume as input, and the sulfur compound output and nitrogen oxide output as output; the desulfurizer prediction module takes the sulfur compound output and the desulfurizer type as input, and the desulfurizer spray amount as output; the denitrifier prediction module takes the nitrogen oxide output and the denitrifier type as input, and the denitrifier spray amount as output.

[0062] Specifically, the pollutant prediction module is provided with a CNN-LSTM network structure. The network structure is pre-trained using a data set consisting of historical data on fuel type, total flue gas volume, flue gas composition, sulfide output, and nitrogen oxide output, and the network weights are adjusted based on a validation function (MSE function) until the model error is minimized (i.e., the MSE loss is reduced to below 0.001).

[0063] A DBN-Stacking integrated model is respectively provided in the desulfurizer prediction module and the denitrifier prediction module; the DBN-Stacking integrated model includes an input layer, a CNN layer, an RNN layer, a DBN layer, a simplified CNN layer and an output layer. Among them, the DBN-Stacking integrated model in the desulfurizer prediction module is pre-trained with a data set consisting of historical data on sulfide output, desulfurizer type, and desulfurizer spraying amount; the DBN-Stacking integrated model in the denitrifier prediction module is pre-trained with a data set consisting of historical data on nitrate output, denitrifier type, and denitrifier spraying amount. The two models adjust the network weights based on the validation function (MSE function) until the model error is minimized (i.e., the MSE loss is reduced to below 0.001).

[0064] The entire system achieves effective control of the sulfur oxide and nitrogen oxide content in industrial emissions through the coordinated work of the pollutant prediction module, the desulfurizer prediction module and the denitrifier prediction module.

[0065] This embodiment also provides a method for optimizing biomass oxygen-enriched combustion and intelligently controlling pollutants, which uses the above-mentioned biomass oxygen-enriched combustion optimization and intelligent pollutant control system to optimize biomass oxygen-enriched combustion and intelligently control pollutants; the method comprises the following steps:

[0066] The boiler temperature and pressure prediction system receives the load instruction and predicts the boiler operating pressure and boiler operating temperature corresponding to the load; the fuel identification module of the fuel prediction and control system confirms the fuel type; the fuel calorific value prediction module of the fuel prediction and control system predicts the fuel calorific value; the fuel demand prediction module of the fuel prediction and control system predicts the amount of fuel entering the furnace;

[0067] The primary air calculation module of the primary and secondary air prediction and control system takes the amount of fuel fed into the furnace, the operating temperature of the boiler and the historical data of the boiler bed pressure as input to calculate the primary air volume and adjust the opening or frequency of the primary fan in real time; The secondary air calculation module of the primary and secondary air prediction and control system takes the amount of fuel fed into the furnace and the operating temperature of the boiler as input to calculate the secondary air volume and adjust the opening or frequency of the secondary fan in real time;

[0068] The induced draft fan control module of the flue gas volume prediction system controls the opening degree or frequency of the induced draft fan; the flue gas prediction module of the flue gas volume prediction system takes the boiler operating pressure and the opening degree or frequency of the induced draft fan as input, predicts the flue gas volume and outputs it;

[0069] The recirculation volume prediction module of the flue gas recirculation control system takes the flue gas volume as input, predicts the carbon capture flue gas volume and the recirculation air volume, and outputs the predictions; the supplementary volume prediction module of the flue gas recirculation control system takes the recirculation air volume, primary air volume, and secondary air volume as input, predicts the air supplementary volume and the pure oxygen supplementary volume, and outputs the predictions, and controls the air fan and the oxygen generator in real time according to the output data;

[0070] The pollutant prediction module of the desulfurization and denitrification intelligent control system takes the fuel type and flue gas volume as input, predicts the sulfur compound output and nitrogen oxide output and outputs; the desulfurizer prediction module of the flue gas recirculation control system takes the sulfur compound output and the desulfurizer type as input, predicts the desulfurizer spray amount and outputs; the denitrifier prediction module of the flue gas recirculation control system takes the nitrogen oxide output and the denitrifier type as input, predicts the denitrifier spray amount and outputs.

[0071] The present embodiment provides a biomass oxygen-enriched combustion optimization and pollutant intelligent control system and method, which can solve the combustion system and pollutant intelligent control problems faced by biomass boilers under oxygen-enriched combustion conditions under variable fuels, multiple operating conditions, and large circulation rates, and realize dynamic optimization and precise control of biomass boilers under oxygen-enriched combustion conditions.

[0072] The above is only an embodiment of the present invention. The common sense such as the known specific structure and characteristics in the scheme is not described in detail here. The ordinary technicians in the relevant field are aware of all the common technical knowledge in the technical field of the invention before the application date or priority date, can obtain all the existing technologies in the field, and have the ability to apply the conventional experimental means before that date. The ordinary technicians in the relevant field can improve and implement the scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for the ordinary technicians in the relevant field to implement this application. It should be pointed out that for the technicians in this field, without departing from the structure of the present invention, several deformations and improvements can be made, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent.

Claims

1. Biomass oxygen-enriched combustion optimization and pollutant intelligent control system, characterized in that: Including boiler temperature and pressure prediction system, fuel prediction and control system, primary and secondary air prediction and control system, flue gas volume prediction system, flue gas recirculation control system and desulfurization and denitrification intelligent control system; The boiler temperature and pressure prediction system is used to receive load instructions and predict the boiler operating pressure and boiler operating temperature corresponding to the load; the fuel prediction and control system includes a fuel identification module for confirming the type of fuel, a fuel calorific value prediction module for predicting the calorific value of the fuel, and a fuel demand prediction module for predicting the amount of fuel entering the furnace; The primary and secondary air prediction and control system includes a primary air calculation module and a secondary air calculation module; the primary air calculation module takes the amount of fuel fed into the furnace, the operating temperature of the boiler and the historical data of the boiler bed pressure as input to calculate the primary air volume and adjust the opening or frequency of the primary fan in real time; the secondary air calculation module takes the amount of fuel fed into the furnace and the operating temperature of the boiler as input to calculate the secondary air volume and adjust the opening or frequency of the secondary fan in real time; The flue gas volume prediction system comprises an induced draft fan control module for controlling the opening or frequency of the flue gas induced draft fan and a flue gas prediction module for predicting the flue gas volume; the flue gas prediction module takes the boiler operating pressure and the opening or frequency of the flue gas induced draft fan as input and takes the predicted flue gas volume as output; The flue gas recirculation control system comprises a recirculation volume prediction module and a supplementary volume prediction module; the recirculation volume prediction module takes the flue gas volume as input, and takes the predicted carbon capture flue gas volume and recirculation air volume as output; the supplementary volume prediction module takes the recirculation air volume, primary air volume and secondary air volume as input, and takes the air supplementary volume and pure oxygen supplementary volume as output, and respectively controls the air blower and the oxygen generator in real time according to the output data; The desulfurization and denitrification intelligent control system includes a pollutant prediction module, a desulfurizer prediction module and a denitrifier prediction module; the pollutant prediction module takes the fuel type and the flue gas volume as input, and the sulfur compound output and the nitrogen oxide output as output; the desulfurizer prediction module takes the sulfur compound output and the desulfurizer type as input, and the desulfurizer spray amount as output; the denitrifier prediction module takes the nitrogen oxide output and the denitrifier type as input, and the denitrifier spray amount as output.

2. The biomass oxygen-enriched combustion optimization and pollutant intelligent control system according to claim 1 is characterized in that: The fuel prediction and control system is preset with a fuel database; the fuel database is set with a fuel calorific value comparison table, which lists the unit calorific value of the biomass fuel base material.

3. The biomass oxygen-enriched combustion optimization and pollutant intelligent control system according to claim 1 is characterized in that: When confirming the fuel type, the fuel identification module first collects the fuel image used for the biomass boiler and identifies the fuel type based on the image recognition model; the image recognition model includes an input layer, a first CNN layer, a second CNN layer, a fully connected layer and an output layer.

4. The biomass oxygen-enriched combustion optimization and pollutant intelligent control system according to claim 1 is characterized in that: The fuel types include the categories and proportions of biomass fuel substrates; the fuel calorific value prediction module calls the fuel database based on the fuel types to calculate the fuel calorific value.

5. The biomass oxygen-enriched combustion optimization and pollutant intelligent control system according to claim 1 is characterized in that: The fuel demand prediction module adopts a support vector regression model structure, forms a basic model by training historical data of fuel heat, boiler load, boiler operating temperature and pressure, and fuel input, and updates itself based on real-time data.

6. The biomass oxygen-enriched combustion optimization and pollutant intelligent control system according to claim 1 is characterized in that: The primary air calculation module and the secondary air calculation module both use the least square method to calculate the air volume.

7. The biomass oxygen-enriched combustion optimization and pollutant intelligent control system according to claim 1 is characterized in that: The recirculation quantity prediction module adopts a CNN-LSTM network structure; the supplementary quantity prediction module uses a GRU network structure; the flue gas recirculation control system also includes a PID control module; the PID control module adjusts the opening or frequency of the air fan in real time according to the output of the supplementary quantity prediction module to control the air supplementary quantity.

8. The biomass oxygen-enriched combustion optimization and pollutant intelligent control system according to claim 1 is characterized in that: A DBN-Stacking integrated model is respectively provided in the desulfurizer prediction module and the denitrifier prediction module; the DBN-Stacking integrated model includes an input layer, a CNN layer, an RNN layer, a DBN layer, a simplified CNN layer and an output layer.

9. Biomass oxygen-enriched combustion optimization and pollutant intelligent control method, characterized in that: Applying the biomass oxygen-enriched combustion optimization and pollutant intelligent control system as described in any one of claims 1 to 8 to perform biomass oxygen-enriched combustion optimization and pollutant intelligent control; comprising the following steps: The boiler temperature and pressure prediction system receives the load instruction and predicts the boiler operating pressure and boiler operating temperature corresponding to the load; the fuel identification module of the fuel prediction and control system confirms the fuel type; the fuel calorific value prediction module of the fuel prediction and control system predicts the fuel calorific value; the fuel demand prediction module of the fuel prediction and control system predicts the amount of fuel entering the furnace; The primary air calculation module of the primary and secondary air prediction and control system takes the amount of fuel fed into the furnace, the operating temperature of the boiler and the historical data of the boiler bed pressure as input to calculate the primary air volume and adjust the opening or frequency of the primary fan in real time; The secondary air calculation module of the primary and secondary air prediction and control system takes the amount of fuel fed into the furnace and the operating temperature of the boiler as input to calculate the secondary air volume and adjust the opening or frequency of the secondary fan in real time; The induced draft fan control module of the flue gas volume prediction system controls the opening degree or frequency of the induced draft fan; the flue gas prediction module of the flue gas volume prediction system takes the boiler operating pressure and the opening degree or frequency of the induced draft fan as input, predicts the flue gas volume and outputs it; The recirculation volume prediction module of the flue gas recirculation control system takes the flue gas volume as input, predicts the carbon capture flue gas volume and the recirculation air volume, and outputs the predictions; the supplementary volume prediction module of the flue gas recirculation control system takes the recirculation air volume, primary air volume, and secondary air volume as input, predicts the air supplementary volume and the pure oxygen supplementary volume, and outputs the predictions, and controls the air fan and the oxygen generator in real time according to the output data; The pollutant prediction module of the desulfurization and denitrification intelligent control system takes the fuel type and flue gas volume as input, predicts the sulfur compound output and nitrogen oxide output and outputs; the desulfurizer prediction module of the flue gas recirculation control system takes the sulfur compound output and the desulfurizer type as input, predicts the desulfurizer spray amount and outputs; the denitrifier prediction module of the flue gas recirculation control system takes the nitrogen oxide output and the denitrifier type as input, predicts the denitrifier spray amount and outputs.

Images