A digital twin method for a boiler combustion chamber

Abstract

This invention discloses a digital twin method for boiler combustion chambers, comprising the following steps: (1) obtaining the structural parameters of the combustion chamber and the physicochemical properties of the fuel; (2) dividing the internal space of the combustion chamber into fluid computing grids and fluid computing nodes; (3) selecting at least three operating load conditions as calculation conditions to obtain the thermal parameter calculation results of all fluid computing nodes in the combustion chamber; (4) training the calculation results based on the numerical simulation technology of digital twins to obtain the target model; (5) changing the boundary conditions of the target model to realize the digital twin of the flow reaction process in the combustion chamber. This invention enables operators to quickly obtain the characteristics and correlations of the flow combustion process in the boiler combustion chamber under different loads, which is convenient for the formulation of subsequent energy-saving and emission-reduction schemes.

Description

Technical Field

[0001] This invention relates to the field of boiler combustion technology, and more specifically to a method for creating a digital twin of a boiler combustion chamber. Background Technology

[0002] Coal combustion involves multiple processes, including turbulent flow, solid coal particle motion and combustion, homogeneous volatile matter combustion, heterogeneous semi-coke combustion, and radiative heat transfer, making it a complex physicochemical process. To achieve carbon emission reduction, optimized operation of coal-fired power plant units is urgently needed. However, current methods for obtaining and optimizing coal-fired boiler combustion chamber parameters through experiments or computational fluid dynamics (CFD) simulations suffer from drawbacks such as long development cycles and poor effectiveness, failing to quickly obtain distribution data of flow fields, temperature fields, and component fields under the required operating conditions. With the development of computers and the continuous enhancement of floating-point computing capabilities, digital twin technology has been widely applied in various fields through big data analysis. However, due to the significant differences in the geometric structures of boiler combustion processes studied in academia, the application of combining CFD calculations with digital twin technology is still limited. For boiler users, establishing a digital twin system for a boiler combustion chamber with a defined structure can quickly predict thermal parameters under arbitrary loads on the characteristic plane, which is of great significance for guiding the production process. Summary of the Invention

[0003] To address the aforementioned technical shortcomings, the purpose of this invention is to provide a method for creating a digital twin of a boiler combustion chamber.

[0004] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0005] This invention provides a method for creating a digital twin of a boiler combustion chamber, comprising the following steps:

[0006] (1) Obtain the structural parameters of the combustion chamber and the physicochemical properties of the fuel;

[0007] (2) Divide the internal space of the combustion chamber into a fluid computation grid and fluid computation nodes;

[0008] (3) Select at least three operating load conditions as calculation conditions and determine the corresponding boundary conditions; combine the structural parameters of the combustion chamber, the physicochemical properties of the fuel, the fluid computing grid and the fluid computing nodes to perform numerical calculations on the calculation conditions and obtain the thermal parameter calculation results of all fluid computing nodes in the combustion chamber.

[0009] (4) Using the boundary conditions and fluid computing node relationships as input parameters, and the thermal parameter calculation results of all fluid computing nodes as output parameters, the calculation results are trained based on the numerical simulation technology of digital twins to obtain the target model.

[0010] (5) Change the boundary conditions of the target model to obtain the thermal parameter calculation results of all corresponding fluid calculation nodes, and realize the digital twin of the combustion chamber flow reaction process.

[0011] Preferably, the structural parameters of the combustion chamber in step (1) include the structural dimensions of the combustion chamber and the number of burner nozzles at each corner of the combustion chamber.

[0012] Preferably, in step (2), when dividing the fluid computation grid, in order to reduce the influence of numerical pseudo-diffusion, the grid is constructed based on the flow characteristics, and the center line of the grid in the jet injection interval overlaps or is parallel to the flow direction.

[0013] Preferably, the boundary conditions in step (3) include flow, heat transfer, and mass transfer parameters.

[0014] Preferably, the thermal parameters in step (3) include temperature, pressure, speed, and component concentration.

[0015] Preferably, in step (3), when performing numerical calculations on the calculation conditions, the Euler-Lagrange model is used, with the gas as a continuous state and the particles as a discrete state, and ANSYS Fluent is used to calculate the multiphase reaction flow in the combustion chamber.

[0016] Preferably, step (4) specifically includes: extracting thermal data of temperature, pressure, velocity, and component concentration on the feature plane from the obtained whole-field data of the boiler combustion chamber, and outputting text data as point cloud; using a neural network fitting method, taking the load as the independent variable, the temperature, pressure, velocity, and component concentration parameters as the dependent variables, and the functions y=ax+b and y=ln(x) as the basis functions, and binding the obtained functions with the point coordinates to realize the digital twin function.

[0017] The beneficial effects of this invention are as follows:

[0018] 1. This invention greatly reduces the time required for CFD model calculation and post-processing, improves calculation efficiency, and enables real-time prediction of velocity, pressure, temperature, component concentration data and distribution in characteristic planes of boiler combustion chamber. This allows operators to quickly obtain the characteristics and correlations of the flow combustion process in boiler combustion chamber under different loads, facilitating the formulation of subsequent energy-saving and emission-reduction plans.

[0019] 2. This invention has low requirements for the amount of initial data, and only requires a very small amount of memory after training, saving data storage space and making it easy to operate.

[0020] 3. This invention has good scalability and can be trained on datasets constructed by different variable working condition simulation systems to achieve real-time and fast prediction function. Detailed Implementation

[0021] The present invention will be further described below with reference to specific embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] This embodiment provides a method for creating a digital twin of a boiler combustion chamber, including the following steps:

[0023] (1) Obtain the structural parameters of the combustion chamber and the physicochemical properties of the fuel. The structural parameters of the combustion chamber include the structural dimensions of the combustion chamber and the number of burner nozzles at each corner of the combustion chamber.

[0024] (2) Divide the internal space of the combustion chamber into fluid computing grids and fluid computing nodes. When dividing the fluid computing grid, in order to reduce the influence of numerical pseudo-diffusion, the grid is constructed based on the flow characteristics. The center line of the grid in the jet injection interval overlaps or is parallel to the flow direction.

[0025] (3) Select at least three operating load conditions as calculation conditions and determine the corresponding boundary conditions, including flow, heat transfer and mass transfer parameters; combine the structural parameters of the combustion chamber, the physicochemical properties of the fuel, the fluid computing grid and the fluid computing nodes, and perform numerical calculations on the calculation conditions to obtain the thermal parameter calculation results of all fluid computing nodes in the combustion chamber. The thermal parameters include temperature, pressure, velocity and component concentration. When performing numerical calculations on the calculation conditions, the Euler-Lagrange model is used, the gas is regarded as a continuous state and the particles are regarded as a discrete state, and the multiphase reaction flow in the combustion chamber is calculated using ANSYS Fluent.

[0026] (4) Using the boundary conditions and fluid computing node relationships as input parameters, and the thermal parameter calculation results of all fluid computing nodes as output parameters, the calculation results are trained based on the numerical simulation technology of digital twins to obtain the target model. Specifically, this includes: extracting thermal data of temperature, pressure, velocity, and component concentration on the feature plane from the obtained whole field data of the boiler combustion chamber, and outputting text data as point cloud; using a neural network fitting method, taking the load as the independent variable, the temperature, pressure, velocity, and component concentration parameters as the dependent variables, and the functions y=ax+b and y=ln(x) as the basis functions, and binding the obtained functions with the point coordinates to realize the digital twin function;

[0027] (5) Change the boundary conditions of the target model to obtain the thermal parameter calculation results of all corresponding fluid calculation nodes, and realize the digital twin of the combustion chamber flow reaction process.

[0028] To further verify the method of this embodiment, this example uses a combustion chamber with dimensions of 17.4485*19.558*73.600m (depth*width*height) as a reference, with 18 burner nozzles at each corner of the combustion chamber.

[0029] In this embodiment, coal was used as fuel and analyzed. The analysis revealed the following coal composition: carbon content 45.78%, hydrogen content 2.21%, oxygen content 3.63%, nitrogen content 1.06%, sulfur content 1.92%, ash content 35%, water content 10.4%, and lower calorific value 16209 kJ / kg.

[0030] To reduce the impact of numerical pseudo-diffusion, this embodiment constructs a grid based on flow characteristics, with the center line of the grid in the jet injection section overlapping or parallel to the flow direction.

[0031] Using the Euler-Lagrange model, the gas is treated as a continuous state and the particles as a discrete state. The multiphase reaction flow in the combustion chamber is calculated using the commercial software ANSYS Fluent.

[0032] Based on digital twin numerical simulation technology, the calculation results are used to train and establish a multi-parameter prediction model for the boiler combustion chamber:

[0033] The thermal data such as temperature, pressure, velocity, and component concentration on the feature plane are extracted from the whole field data of the boiler combustion chamber obtained in the previous steps, and the text data is output as a point cloud. The neural network fitting method is used to set the load as the independent variable and the parameters such as temperature, pressure, velocity, and component concentration as the dependent variables for the data of each point, and the functions y=ax+b and y=ln(x) as the basis functions. The obtained functions are then bound to the point coordinates to realize the digital twin function.

[0034] The training model and data are encapsulated in a GUI using C++, thereby enabling online rapid prediction of variable loads in boiler combustion chambers.

[0035] The system integrates real-time user input, target plane, parameter selection, and plotting functions into a GUI system using C++. This allows users to input load data, select the plane and thermal parameters, and quickly obtain and save the required parameter distribution. Ultimately, this results in a digital twin system that can quickly predict thermal parameters under any load on a characteristic plane.

[0036] Tests showed that in commercial software calculations, excluding the time for drawing the mesh and setting parameters, it takes 25 minutes to calculate a load condition; while this embodiment only takes 10 seconds to obtain the characteristic plane thermal parameters under any load, greatly speeding up the calculation while ensuring the accuracy of the results.

[0037] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for creating a digital twin of a boiler combustion chamber, characterized in that, Includes the following steps: (1) Obtain the structural parameters of the combustion chamber and the physicochemical properties of the fuel; (2) Divide the internal space of the combustion chamber into fluid computing grids and fluid computing nodes; when dividing the fluid computing grid, in order to reduce the influence of numerical pseudo-diffusion, the grid is constructed based on the flow characteristics, and the center line of the grid in the jet injection section overlaps or is parallel to the flow direction; (3) Select at least three operating load conditions as calculation conditions and determine the corresponding boundary conditions; combine the structural parameters of the combustion chamber, the physicochemical properties of the fuel, the fluid calculation grid and the fluid calculation nodes to perform numerical calculations on the calculation conditions and obtain the thermal parameter calculation results of all fluid calculation nodes in the combustion chamber; when performing numerical calculations on the calculation conditions, the Euler-Lagrange model is used, the gas is regarded as a continuous state and the particles are regarded as a discrete state, and the multiphase reaction flow in the combustion chamber is calculated using ANSYS Fluent; (4) Using the boundary conditions and fluid computing node relationships as input parameters, and the thermal parameter calculation results of all fluid computing nodes as output parameters, the calculation results are trained based on the numerical simulation technology of digital twin to obtain the target model; specifically, it includes: extracting thermal data of temperature, pressure, velocity, and component concentration on the feature plane from the obtained whole field data of the boiler combustion chamber, and outputting text data as point cloud; using the neural network fitting method, taking the load as the independent variable, the temperature, pressure, velocity, and component concentration parameters as the dependent variable, and the y=ax+b and y=ln(x) functions as the basis functions, and binding the obtained functions with the point coordinates to realize the digital twin function; (5) Change the boundary conditions of the target model to obtain the thermal parameter calculation results of all corresponding fluid calculation nodes and realize the digital twin of the combustion chamber flow reaction process.

2. The method for digital twinning a boiler combustion chamber as described in claim 1, characterized in that, The structural parameters of the combustion chamber in step (1) include the structural dimensions of the combustion chamber and the number of burner nozzles at each corner of the combustion chamber.

3. The method for digital twinning a boiler combustion chamber as described in claim 1, characterized in that, The boundary conditions in step (3) include flow, heat transfer, and mass transfer parameters.

4. The method for digital twinning a boiler combustion chamber as described in claim 1, characterized in that, The thermal parameters in step (3) include temperature, pressure, speed, and component concentration.