Method and system for optimizing electrostatic dust collection efficiency of thermal power generating unit based on multivariable prediction

Abstract

The invention relates to the technical field of thermal power generating unit dust removal, and discloses a thermal power generating unit electrostatic dust removal efficiency optimization method and system based on multivariable prediction, and the method comprises the steps: constructing a flue three-dimensional model from a boiler air pre-heater outlet section to an electrostatic dust collector inlet section; performing fluid dynamic simulation on the flue gas flow field through the flue three-dimensional model, and evaluating the uniformity of the flue gas flow field of the inlet flue of the electrostatic dust collector according to a fluid dynamic simulation result; based on the uniformity of a flue gas flow field of an inlet flue of the electrostatic dust collector, a current flow distribution deviation cause is analyzed, and a flue gas flow field optimization measure is constructed in combination with the deviation cause. According to the method, in combination with refined simulation processes such as structured grid discretization, iterative collision transfer and macroscopic field reconstruction of the lattice Boltzmann technology, reliable flue gas flow field simulation data under the multi-load working condition can be output, and flow field topological feature and uniformity index analysis is performed based on the flue gas flow field simulation data; and thus, accurate formulation of electrostatic dust collection efficiency optimization measures is supported.

Description

Technical Field

[0001] This invention relates to the field of dust removal technology for thermal power units, and more specifically, to a method and system for optimizing the efficiency of electrostatic dust removal in thermal power units based on multivariate prediction. Background Technology

[0002] A thermal power unit is an electrical power plant that generates electricity by burning fossil fuels such as coal and using steam to drive a generator. A thermal power unit mainly consists of a boiler, steam turbine, generator, and cooling system. The large amount of flue gas generated during combustion needs to be purified by flue gas treatment devices to reduce the emission of harmful substances. Electrostatic precipitators are key equipment in thermal power units for removing solid particles from flue gas. They use a high-voltage electric field to charge the particles and collect them through the electric field force, thereby purifying the flue gas. Optimizing the efficiency of electrostatic precipitators in thermal power units is crucial because their efficiency directly affects the concentration of emissions and environmental pollution control.

[0003] Therefore, optimizing electrostatic precipitator efficiency can not only improve dust removal effect and reduce harmful emissions, but also enhance the overall environmental performance of thermal power units and meet increasingly stringent environmental regulations. If the uniformity of the flue gas flow field in the inlet flue of the electrostatic precipitator is not considered during efficiency optimization, it may lead to uneven flue gas flow, with airflow velocities in some areas being too high or too low, resulting in uneven dust removal efficiency. Specifically, uneven airflow will cause fly ash in some flue gas areas to be unable to be effectively captured, thereby reducing overall dust removal efficiency. It may even prevent the electrostatic precipitator from achieving the expected dust removal effect in certain areas, thus affecting the achievement of emission standards, reducing the system's energy utilization efficiency, and increasing operating costs.

[0004] No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0005] To address the problems in related technologies, this invention proposes a method and system for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction, in order to overcome the aforementioned technical problems existing in the existing related technologies.

[0006] Therefore, the specific technical solution adopted by the present invention is as follows: According to a first aspect of the present invention, a method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction is provided, the method comprising: S1. Based on the pre-acquired structural data of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue, construct a three-dimensional model of the flue from the boiler air preheater outlet section to the electrostatic precipitator inlet section. S2. Collect flue gas parameters at the inlet flue of the electrostatic precipitator and use them as boundary conditions for the three-dimensional model of the flue. Perform fluid dynamics simulation on the flue gas flow field using the three-dimensional model of the flue, and evaluate the uniformity of the flue gas flow field at the inlet flue of the electrostatic precipitator based on the fluid dynamics simulation results. S3. Based on the uniformity of the flue gas flow field in the inlet flue of the electrostatic precipitator, analyze the causes of the current flow distribution deviation, and then combine the causes of the deviation to construct flue gas flow field optimization measures to achieve flow field optimization from the boiler air preheater outlet flue to the electrostatic precipitator inlet flue.

[0007] Furthermore, based on the pre-acquired structural data of the boiler air preheater outlet flue and electrostatic precipitator inlet flue, a three-dimensional model of the flue from the boiler air preheater outlet section to the electrostatic precipitator inlet section is constructed, including: Structural data of the boiler air preheater outlet flue and electrostatic precipitator inlet flue were collected and standardized, and the flue geometry features of the boiler air preheater outlet flue and electrostatic precipitator inlet flue were extracted from the standardized structural data. An initial three-dimensional mesh model of the flue was constructed by using structured mesh generation technology and combining it with the geometric features of the flue. The fluid medium properties and constraints of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue are obtained. The boundary conditions and physical parameters of the initial flue 3D mesh model are coupled to obtain the flue 3D model from the boiler air preheater outlet section to the electrostatic precipitator inlet section.

[0008] Furthermore, the fluid medium properties of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue include fluid type, density, viscosity, specific heat capacity, thermal conductivity, flue gas component concentration, and dust particle characteristics; fluid medium constraints include boundary constraints, physical constraints, flow state constraints, and operating condition constraints.

[0009] Furthermore, flue gas parameters from the electrostatic precipitator inlet flue are collected and used as boundary conditions for the 3D flue model. The flue gas flow field is then simulated using the 3D flue model, and the uniformity of the flue gas flow field in the electrostatic precipitator inlet flue is evaluated based on the simulation results, including: Multi-dimensional flue gas parameters of the electrostatic precipitator inlet section were collected under different boiler operating conditions and used as input boundary conditions for the three-dimensional flue gas model. Using a 3D model of the flue as a carrier, a simulation scenario of the flue gas flow field in the inlet flue of the electrostatic precipitator is constructed. The lattice Boltzmann technique is then used to perform fluid dynamics simulation of the flue gas flow field in the simulation scenario, and output simulation data of the flue gas flow field under different boiler load operating conditions. Based on the simulation data of flue gas flow field under different boiler operating conditions, the topological characteristics and topological uniformity index of the flue gas flow field are analyzed, and the flue gas flow field uniformity of the inlet flue of the electrostatic precipitator is evaluated by combining the flow field topological characteristics and topological uniformity index.

[0010] Furthermore, the multi-dimensional flue gas parameters of the electrostatic precipitator inlet section under different boiler load operating conditions include the flue gas velocity, temperature, and pressure at different flue ducts at the electrostatic precipitator inlet and at different cross-sections of the same flue duct under different boiler load operating conditions.

[0011] Furthermore, using a 3D model of the flue gas duct as a carrier, a simulation scenario of the flue gas flow field in the inlet flue of the electrostatic precipitator is constructed. Then, combining lattice Boltzmann technology, fluid dynamics simulation of the flue gas flow field is performed within the simulation scenario, outputting simulated flue gas flow field data under different boiler operating conditions, including: Discretize the three-dimensional model of the flue into a structured mesh required by the lattice Boltzmann technique, and mark the boundary regions in the structured mesh; Construct simulation scenarios of flue gas flow field under different boiler operating conditions. In each flue gas flow field simulation scenario, multi-dimensional flue gas parameters are converted into distribution function form and the corresponding boundary region grids are assigned values, and several initial fluid grids are output. Local collision calculations and flow transfer between neighboring lattices are performed iteratively on all initial fluid lattices. The macroscopic field is reconstructed in real time after each iteration until the flow field of each boiler load operating condition meets the convergence state. The full flow field data of flue gas for each boiler load operating condition is output. The flue gas full flow field data is interpolated and mapped to the target section. The velocity vector, pressure, turbulence intensity and fly ash concentration at each location on the target section are calculated to generate the target section dataset for each boiler load operation condition and serve as the flue gas flow field simulation data.

[0012] Furthermore, local collision calculations and flow transfer between neighboring lattices are iteratively performed on all initial fluid lattices. The macroscopic field is reconstructed in real time after each iteration until the flow field for each boiler load operating condition meets the convergence state. The output flue gas full flow field data for each boiler load operating condition includes: Based on a preset relaxation time, the collision process of the distribution function is simulated at each initial fluid lattice node, so that the collision process relaxes to a local equilibrium state, and the distribution function after the collision is obtained. The post-collision distribution function is transferred to the neighboring fluid lattice nodes along the direction of the target discrete velocity, and the corresponding distribution function boundary conditions are applied at the boundary fluid lattice nodes to obtain the updated distribution function. The local collision calculation and the flow transfer process in neighboring lattice are performed iteratively, and the distribution function in all directions on the fluid lattice nodes is summed after each iteration to reconstruct the macroscopic field in real time. The iteration process stops when the flow field monitoring index for each boiler load operation condition reaches a steady state, and the full flow field data of the flue gas in the converged state under each boiler load operation condition is output.

[0013] Furthermore, based on the simulated flue gas flow field data under different boiler operating conditions, the topological characteristics and topological uniformity index of the flue gas flow field are analyzed. The flue gas flow field uniformity of the electrostatic precipitator inlet flue is then evaluated using the flow field topological characteristics and topological uniformity index, including: The streamline tracing algorithm is used to analyze the topological features of the flue gas flow field simulation data, identify the topological features of the flue gas flow field under different boiler operating conditions, and transform the topological features of the flue gas flow field into flue gas flow field feature vectors. The flue gas flow field feature vector is compared with a preset topological uniformity index, and the flue gas flow field uniformity of the electrostatic precipitator inlet flue is evaluated based on the comparison results.

[0014] Furthermore, the flue gas flow field feature vector is compared with a preset topological uniformity index, and the flue gas flow field uniformity of the electrostatic precipitator inlet flue is evaluated based on the comparison results, including: If the eigenvector of the flue gas flow field is greater than or equal to the topological uniformity index, it indicates that the flue gas flow field in the inlet flue of the electrostatic precipitator is in a uniform state. If the flue gas flow field eigenvector is less than the topological uniformity index, it indicates that the flue gas flow field at the inlet flue of the electrostatic precipitator is in a non-uniform state, and step S3 is executed.

[0015] According to a second aspect of the present invention, a system for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction is provided, the system comprising: The flue 3D model construction module is used to construct a 3D model of the flue from the boiler air preheater outlet section to the electrostatic precipitator inlet section based on the pre-acquired structural data of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue. The flue gas flow field simulation module is used to collect flue gas parameters from the inlet flue of the electrostatic precipitator and use them as boundary conditions for the three-dimensional model of the flue. The fluid dynamics of the flue gas flow field are simulated using the three-dimensional model of the flue, and the uniformity of the flue gas flow field in the inlet flue of the electrostatic precipitator is evaluated based on the results of the fluid dynamics simulation. The flow field optimization module is used to analyze the causes of current flow distribution deviations based on the uniformity of the flue gas flow field in the inlet flue of the electrostatic precipitator, and then construct flue gas flow field optimization measures based on the causes of deviations, so as to achieve flow field optimization from the boiler air preheater outlet flue to the electrostatic precipitator inlet flue.

[0016] The beneficial effects of this invention are as follows: 1. This invention constructs a three-dimensional model of the entire flue and conducts flow field simulation based on multivariate flue gas parameters. This allows for accurate assessment of the uniformity of the flow field at the inlet of the electrostatic precipitator, providing reliable data support for multivariate prediction. By analyzing the causes of flow distribution deviations and constructing targeted optimization measures, the flow field state of the entire flue can be effectively improved, thereby enhancing the efficiency of electrostatic precipitator and achieving precise optimization and efficient operation of the electrostatic precipitator system in thermal power units.

[0017] 2. This invention collects multi-section, multi-dimensional flue gas parameters at the inlet of the electrostatic precipitator under different boiler loads, providing accurate boundary conditions for the three-dimensional model of the flue. Combined with the structured mesh discretization, iterative collision transfer, and macroscopic field reconstruction of the lattice Boltzmann technique, it can output reliable flue gas flow field simulation data under multiple load conditions. Based on the flue gas flow field simulation data, flow field topology characteristics and uniformity index analysis are performed, thereby achieving a scientific assessment of the uniformity of the flue gas flow field and supporting the precise formulation of electrostatic precipitator efficiency optimization measures. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of a method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the principle of the electrostatic precipitator efficiency optimization system for thermal power units based on multivariate prediction according to an embodiment of the present invention.

[0020] In the picture: 1. 3D model construction module for flue gas duct; 2. Flue gas flow field simulation module; 3. Flow field optimization module. Detailed Implementation

[0021] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0022] According to embodiments of the present invention, a method and system for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction are provided.

[0023] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 As shown, according to an embodiment of the present invention, a method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction includes: S1. Based on the pre-acquired structural data of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue, construct a three-dimensional model of the flue from the boiler air preheater outlet section to the electrostatic precipitator inlet section.

[0024] Among them, based on the pre-acquired structural data of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue, a three-dimensional flue model is constructed from the boiler air preheater outlet section to the electrostatic precipitator inlet section, including: Structural data of the boiler air preheater outlet flue and electrostatic precipitator inlet flue were collected and standardized, and the flue geometry features of the boiler air preheater outlet flue and electrostatic precipitator inlet flue were extracted from the standardized structural data. An initial three-dimensional mesh model of the flue was constructed by using structured mesh generation technology and combining it with the geometric features of the flue. The fluid medium properties and constraints of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue are obtained. The boundary conditions and physical parameters of the initial flue 3D mesh model are coupled to obtain the flue 3D model from the boiler air preheater outlet section to the electrostatic precipitator inlet section.

[0025] The fluid medium properties of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue include fluid type, density, viscosity, specific heat capacity, thermal conductivity, flue gas component concentration and dust particle characteristics; fluid medium constraints include boundary constraints, physical constraints, flow state constraints and operating condition constraints.

[0026] It should be noted that by collecting and standardizing the structural data of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue, the consistency and comparability of the structural data can be ensured, and the geometric features of the flue can be extracted, providing a foundation for subsequent mesh generation and physical modeling. Based on this, using structured mesh generation technology, combined with the geometric features of the flue, an initial three-dimensional flue mesh model can be effectively generated, providing accurate spatial discretization for fluid dynamics simulation. By acquiring the fluid medium properties and constraints of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue, and coupling these parameters with the boundary conditions of the initial three-dimensional flue mesh model, accurate assignment of the physical field and boundary conditions of the model can be achieved, ultimately obtaining a complete three-dimensional flue model from the boiler air preheater outlet section to the electrostatic precipitator inlet section.

[0027] A three-dimensional model of the flue gas duct can accurately reflect the flow and physical characteristics of the flue gas, providing reliable basic data for subsequent flow field analysis, optimization design, and equipment performance evaluation. This process effectively improves the accuracy and reliability of flue gas duct design, providing a scientific basis for optimizing boiler operation, improving electrostatic precipitator efficiency, and reducing energy consumption.

[0028] S2. Collect flue gas parameters from the inlet flue of the electrostatic precipitator and use them as boundary conditions for the three-dimensional model of the flue. Perform fluid dynamics simulation on the flue gas flow field using the three-dimensional model of the flue, and evaluate the uniformity of the flue gas flow field in the inlet flue of the electrostatic precipitator based on the fluid dynamics simulation results.

[0029] The process includes collecting flue gas parameters from the electrostatic precipitator inlet flue and using them as boundary conditions for the flue gas 3D model. The flue gas flow field is then simulated using the 3D model, and the uniformity of the flue gas flow field in the electrostatic precipitator inlet flue is evaluated based on the simulation results. Multi-dimensional flue gas parameters of the electrostatic precipitator inlet section were collected under different boiler operating conditions and used as input boundary conditions for the three-dimensional flue gas model.

[0030] Among them, the multi-dimensional flue gas parameters of the electrostatic precipitator inlet section under different boiler load operating conditions include the flue gas velocity, temperature and pressure of different flue ducts at the electrostatic precipitator inlet and different cross sections of the same flue duct under different boiler load operating conditions.

[0031] Using a 3D model of the flue as a carrier, a simulation scenario of the flue gas flow field in the inlet flue of an electrostatic precipitator is constructed. The lattice Boltzmann technique is then used to perform fluid dynamics simulation of the flue gas flow field in the simulation scenario, and the simulation data of the flue gas flow field under different boiler load operating conditions are output.

[0032] Specifically, using a 3D model of the flue gas duct as a carrier, a simulation scenario of the flue gas flow field in the inlet flue of the electrostatic precipitator is constructed. Then, combining lattice Boltzmann technology, fluid dynamics simulation of the flue gas flow field is performed within the simulation scenario, outputting simulated flue gas flow field data under different boiler operating conditions, including: Discretize the three-dimensional model of the flue into a structured mesh required by the lattice Boltzmann technique, and mark the boundary regions in the structured mesh; Construct simulation scenarios of flue gas flow field under different boiler operating conditions. In each flue gas flow field simulation scenario, multi-dimensional flue gas parameters are converted into distribution function form and the corresponding boundary region grids are assigned values, and several initial fluid grids are output. Local collision calculations and flow transfer between neighboring lattices are performed iteratively on all initial fluid lattices. The macroscopic field is reconstructed in real time after each iteration until the flow field of each boiler load operating condition meets the convergence state, and the full flue gas flow field data of each boiler load operating condition is output.

[0033] It should be noted that the Lattice Boltzmann (LBM) method is a numerical computation method based on a discrete particle model for simulating fluid dynamics problems. It treats the fluid as composed of multiple discrete particles, each propagating along a predefined lattice path, and updating its velocity distribution function according to collision rules during collisions. Unlike traditional numerical fluid dynamics methods based on the assumption of a continuous medium (such as the finite volume method or the finite difference method), LBM, through lattice points and local collisions, can handle complex boundary conditions and non-uniform flows.

[0034] The 3D model of the flue gas duct is discretized into a structured mesh required by the lattice Boltzmann technique, and boundary regions are marked within the mesh to ensure correct handling of the flue gas wall and other physical boundaries in the simulation. Simulation scenarios of the flue gas flow field under different boiler load conditions are constructed. In each simulation scenario, multi-dimensional flue gas parameters are transformed into the distribution function form required by the lattice Boltzmann technique, and values ​​are assigned to the grid cells in the boundary regions, outputting several initial fluid grid cells. Then, local collision calculations and flow transfer between neighboring grid cells are performed iteratively, with the macroscopic field updated in real time after each iteration to ensure simulation accuracy. Finally, this process continues until the flue gas flow field for each boiler load operating condition meets the convergence state, outputting the full flue gas flow field data for each condition. This method can efficiently and accurately simulate and analyze the flue gas flow characteristics under different operating conditions, providing detailed flow field data, which helps optimize flue gas duct design, evaluate the performance of electrostatic precipitators, improve dust removal efficiency, and provide a scientific basis for subsequent flow field optimization and operational adjustments.

[0035] Specifically, local collision calculations and flow transfer between neighboring lattices are performed iteratively on all initial fluid lattices. The macroscopic field is reconstructed in real time after each iteration until the flow field for each boiler load operating condition meets the convergence state. The output flue gas full flow field data for each boiler load operating condition includes: Based on a preset relaxation time, the collision process of the distribution function is simulated at each initial fluid lattice node, so that the collision process relaxes to a local equilibrium state, and the distribution function after the collision is obtained. The post-collision distribution function is transferred to the neighboring fluid lattice nodes along the direction of the target discrete velocity, and the corresponding distribution function boundary conditions are applied at the boundary fluid lattice nodes to obtain the updated distribution function. The local collision calculation and the flow transfer process in neighboring lattice are performed iteratively, and the distribution function in all directions on the fluid lattice nodes is summed after each iteration to reconstruct the macroscopic field in real time. The iteration process stops when the flow field monitoring index for each boiler load operation condition reaches a steady state, and the full flow field data of the flue gas in the converged state under each boiler load operation condition is output.

[0036] The flue gas full flow field data is interpolated and mapped to the target section. The velocity vector, pressure, turbulence intensity and fly ash concentration at each location on the target section are calculated to generate the target section dataset for each boiler load operation condition and serve as the flue gas flow field simulation data.

[0037] It should be noted that at each initial fluid grid node, the collision process of the distribution function is simulated according to a preset relaxation time, gradually relaxing it towards a local equilibrium state to obtain the post-collision distribution function. Subsequently, the post-collision distribution function is transmitted to neighboring fluid grid nodes according to the target discrete velocity direction, and corresponding distribution function boundary conditions are applied to the boundary fluid grid nodes to update their distribution functions. By continuously iterating the local collision calculation and the flow transmission process to neighboring grids, the distribution functions in all directions on the fluid grid nodes are summed after each iteration, thereby reconstructing the macroscopic field (such as velocity, pressure, etc.) in real time. When the flow field monitoring indicators under boiler load operation conditions reach a steady state, the iteration stops and the full flue gas flow field data under convergence is output. Then, the full flue gas flow field data is interpolated and mapped to the target section, and the velocity vector, pressure, turbulence intensity, and fly ash concentration at each location on the section are calculated, generating the target section dataset for each boiler load condition as the flue gas flow field simulation data. The principle behind this process accurately reflects changes in flue gas flow, helps assess and optimize boiler operating conditions and the uniformity of the flue gas flow field, thereby improving the performance of equipment such as electrostatic precipitators and providing a scientific basis for further engineering design and optimization.

[0038] Based on the simulation data of flue gas flow field under different boiler operating conditions, the topological characteristics and topological uniformity index of the flue gas flow field are analyzed, and the flue gas flow field uniformity of the inlet flue of the electrostatic precipitator is evaluated by combining the flow field topological characteristics and topological uniformity index.

[0039] Among them, the analysis of flue gas flow field topological characteristics and topological uniformity index based on flue gas flow field simulation data under different boiler operating conditions, and the evaluation of flue gas flow field uniformity in the inlet flue of the electrostatic precipitator based on the flow field topological characteristics and topological uniformity index, include: The streamline tracing algorithm is used to analyze the topological features of the flue gas flow field simulation data, identify the topological features of the flue gas flow field under different boiler operating conditions, and transform the topological features of the flue gas flow field into flue gas flow field feature vectors.

[0040] Streamline tracing is a numerical method for calculating the trajectory of fluid particles based on fluid velocity field data. It draws continuous streamlines by solving the streamline differential equations in the velocity field, thereby visualizing and analyzing the flow field structure. Starting from a seed point and integrating stepwise along the velocity direction, the algorithm can identify key topological structures such as vortices, separated flows, and recirculation regions, specifically including: By employing a streamline tracing algorithm, the flue gas flow path is calculated based on the velocity information of the flue gas flow field. This identifies key flow structures such as vortices, recirculation zones, and flow separation zones. These structures represent the topological characteristics of the flow field, revealing its complexity, stability, and non-uniformity. The extracted topological features are then transformed into flue gas flow field feature vectors. By quantifying the characteristics of each topological region in the flow field, such as vortex intensity, flow separation zone area, and streamline density, the complex flow structures are converted into a set of numerically standardized feature parameters. This provides support for the evaluation and optimization design of flow field uniformity under boiler load conditions.

[0041] The flue gas flow field feature vector is compared with a preset topological uniformity index, and the flue gas flow field uniformity of the electrostatic precipitator inlet flue is evaluated based on the comparison results.

[0042] The process of comparing the flue gas flow field feature vector with a preset topological uniformity index and evaluating the flue gas flow field uniformity of the electrostatic precipitator inlet flue based on the comparison results includes: If the eigenvector of the flue gas flow field is greater than or equal to the topological uniformity index, it indicates that the flue gas flow field in the inlet flue of the electrostatic precipitator is in a uniform state. If the flue gas flow field eigenvector is less than the topological uniformity index, it indicates that the flue gas flow field at the inlet flue of the electrostatic precipitator is in a non-uniform state, and step S3 is executed.

[0043] S3. Based on the uniformity of the flue gas flow field in the inlet flue of the electrostatic precipitator, analyze the causes of the current flow distribution deviation, and then combine the causes of the deviation to construct flue gas flow field optimization measures to achieve flow field optimization from the boiler air preheater outlet flue to the electrostatic precipitator inlet flue.

[0044] Furthermore, this invention also includes constructing a multivariate prediction model based on multidimensional flue gas parameters (flow velocity, temperature, etc.) and flow field simulation data (uniformity index, topological features, etc.), inputting historical operating data to train the multivariate prediction model to establish the correlation between flow field uniformity, flue gas parameters and electrostatic precipitator efficiency, using the trained multivariate prediction model to predict electrostatic precipitator efficiency and optimal flow field parameter range under different operating conditions, and simultaneously formulating a scheme based on the optimal parameter range output by the multivariate prediction model when constructing flue gas flow field optimization measures.

[0045] Multivariate prediction models can employ architectures including multilayer perceptrons and convolutional neural networks (RNNs) based on neural networks. These models can learn complex nonlinear relationships from historical data to predict the efficiency of electrostatic precipitators. Their working principle involves using parameters such as flue gas velocity and temperature, as well as flow field simulation results, as input features. The model is trained to obtain the relationship between the inputs and the electrostatic precipitator efficiency, enabling multi-dimensional predictions. Through model training, subtle changes in flow field characteristics and dust removal efficiency under different operating conditions can be captured, thus achieving optimized prediction of dust removal efficiency. Multivariate prediction models can accurately predict the efficiency of electrostatic precipitators under different loads and operating conditions, identify potential performance bottlenecks in advance, optimize process parameters, and ultimately improve dust removal efficiency, reduce energy consumption, and enhance the environmental benefits of boilers.

[0046] It should be noted that, based on the CFD verification calculation results and the optimal parameter range output by the multivariate prediction model, flow field optimization is carried out in the flue gas duct from the air preheater outlet to the electrostatic precipitator inlet. Relevant flow field optimization measures are constructed to ensure that, after flow field optimization, the relative deviation of flue gas flow rate between different channels at the same electrostatic precipitator inlet under different loads is ≤10%. The flue gas flow field optimization measures include ensuring that the flue gas flow field does not become skewed or turbulent during flow field optimization, which would cause equipment scouring and wear, and adopting effective measures to prevent secondary dust generation.

[0047] The measures to optimize the flue gas flow field include: rationally arranging the porosity of the distribution plates to ensure the uniformity of the flow field in the dust collector; replacing all the distribution plates inside the air inlet with new distribution plates made of Q355B wear-resistant material; replacing the distribution plate supports; adding a pre-positioned airflow guiding and dust suppression device at the bottom of the new airflow distribution plate; and using the new airflow distribution plate bottom airflow guiding and dust suppression device together with the "W"-shaped ridge-type new airflow distribution plate to maximize the uniformity of airflow and ensure that the airflow uniformity coefficient is <0.15, etc.

[0048] like Figure 2 As shown, according to another embodiment of the present invention, a system for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction is also provided, the system comprising: The flue 3D model construction module 1 is used to construct a 3D flue model from the boiler air preheater outlet section to the electrostatic precipitator inlet section based on the pre-acquired structural data of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue. The flue gas flow field simulation module 2 is used to collect flue gas parameters of the inlet flue of the electrostatic precipitator and use them as boundary conditions for the three-dimensional model of the flue. The flue gas flow field is simulated by the three-dimensional model of the flue, and the uniformity of the flue gas flow field of the inlet flue of the electrostatic precipitator is evaluated based on the results of the fluid dynamics simulation. The flow field optimization module 3 is used to analyze the causes of the current flow distribution deviation based on the uniformity of the flue gas flow field in the inlet flue of the electrostatic precipitator, and then combine the causes of the deviation to construct flue gas flow field optimization measures to achieve flow field optimization from the boiler air preheater outlet flue to the electrostatic precipitator inlet flue.

[0049] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction, characterized in that, The method includes: S1. Based on the pre-acquired structural data of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue, construct a three-dimensional model of the flue from the boiler air preheater outlet section to the electrostatic precipitator inlet section. S2. Collect flue gas parameters at the inlet flue of the electrostatic precipitator and use them as boundary conditions for the three-dimensional model of the flue. Perform fluid dynamics simulation on the flue gas flow field using the three-dimensional model of the flue, and evaluate the uniformity of the flue gas flow field at the inlet flue of the electrostatic precipitator based on the fluid dynamics simulation results. S3. Based on the uniformity of the flue gas flow field in the inlet flue of the electrostatic precipitator, analyze the causes of the current flow distribution deviation, and then combine the causes of the deviation to construct flue gas flow field optimization measures to achieve flow field optimization from the boiler air preheater outlet flue to the electrostatic precipitator inlet flue.

2. The method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction according to claim 1, characterized in that, The construction of a three-dimensional flue model from the boiler air preheater outlet section to the electrostatic precipitator inlet section, based on pre-acquired structural data of the boiler air preheater outlet flue and electrostatic precipitator inlet flue, includes: Structural data of the boiler air preheater outlet flue and electrostatic precipitator inlet flue were collected and standardized, and the flue geometry features of the boiler air preheater outlet flue and electrostatic precipitator inlet flue were extracted from the standardized structural data. An initial three-dimensional mesh model of the flue was constructed by using structured mesh generation technology and combining it with the geometric features of the flue. The fluid medium properties and constraints of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue are obtained. The boundary conditions and physical parameters of the initial flue 3D mesh model are coupled to obtain the flue 3D model from the boiler air preheater outlet section to the electrostatic precipitator inlet section.

3. The method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction according to claim 2, characterized in that, The fluid medium properties of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue include fluid type, density, viscosity, specific heat capacity, thermal conductivity, flue gas component concentration, and dust particle characteristics; the fluid medium constraints include boundary constraints, physical constraints, flow state constraints, and operating condition constraints.

4. The method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction according to claim 1, characterized in that, The process of collecting flue gas parameters from the inlet flue of the electrostatic precipitator and using them as boundary conditions for the 3D model of the flue, performing fluid dynamics simulation of the flue gas flow field using the 3D model, and evaluating the uniformity of the flue gas flow field in the inlet flue of the electrostatic precipitator based on the fluid dynamics simulation results includes: Multi-dimensional flue gas parameters of the electrostatic precipitator inlet section were collected under different boiler operating conditions and used as input boundary conditions for the three-dimensional flue gas model. Using a 3D model of the flue as a carrier, a simulation scenario of the flue gas flow field in the inlet flue of the electrostatic precipitator is constructed. The lattice Boltzmann technique is then used to perform fluid dynamics simulation of the flue gas flow field in the simulation scenario, and output simulation data of the flue gas flow field under different boiler load operating conditions. Based on the simulation data of flue gas flow field under different boiler operating conditions, the topological characteristics and topological uniformity index of the flue gas flow field are analyzed, and the flue gas flow field uniformity of the inlet flue of the electrostatic precipitator is evaluated by combining the flow field topological characteristics and topological uniformity index.

5. The method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction according to claim 4, characterized in that, The multi-dimensional flue gas parameters of the electrostatic precipitator inlet section under different boiler load operating conditions include flue gas velocity, temperature, and pressure at different flue ducts at the electrostatic precipitator inlet and at different cross-sections of the same flue duct under different boiler load operating conditions.

6. The method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction according to claim 5, characterized in that, The simulation scenario of the flue gas flow field in the inlet flue of the electrostatic precipitator is constructed using a three-dimensional model of the flue gas duct as a carrier. The simulation is then combined with lattice Boltzmann technology to perform fluid dynamics simulation of the flue gas flow field within the simulation scenario, outputting simulated flue gas flow field data under different boiler operating conditions, including: Discretize the three-dimensional model of the flue into a structured mesh required by the lattice Boltzmann technique, and mark the boundary regions in the structured mesh; Construct simulation scenarios of flue gas flow field under different boiler operating conditions. In each flue gas flow field simulation scenario, multi-dimensional flue gas parameters are converted into distribution function form and the corresponding boundary region grids are assigned values, and several initial fluid grids are output. Local collision calculations and flow transfer between neighboring lattices are performed iteratively on all initial fluid lattices. The macroscopic field is reconstructed in real time after each iteration until the flow field of each boiler load operating condition meets the convergence state. The full flow field data of flue gas for each boiler load operating condition is output. The flue gas full flow field data is interpolated and mapped to the target section. The velocity vector, pressure, turbulence intensity and fly ash concentration at each location on the target section are calculated to generate the target section dataset for each boiler load operation condition and serve as the flue gas flow field simulation data.

7. The method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction according to claim 6, characterized in that, The process involves iteratively performing local collision calculations and flow transfer between neighboring lattices on all initial fluid lattices, reconstructing the macroscopic field in real time after each iteration, until the flow field for each boiler load operating condition satisfies the convergence state. The output of the full flue gas flow field data for each boiler load operating condition includes: Based on a preset relaxation time, the collision process of the distribution function is simulated at each initial fluid lattice node, so that the collision process relaxes to a local equilibrium state, and the distribution function after the collision is obtained. The post-collision distribution function is transferred to the neighboring fluid lattice nodes along the direction of the target discrete velocity, and the corresponding distribution function boundary conditions are applied at the boundary fluid lattice nodes to obtain the updated distribution function. The local collision calculation and the flow transfer process in neighboring lattice are performed iteratively, and the distribution function in all directions on the fluid lattice nodes is summed after each iteration to reconstruct the macroscopic field in real time. The iteration process stops when the flow field monitoring index for each boiler load operation condition reaches a steady state, and the full flow field data of the flue gas in the converged state under each boiler load operation condition is output.

8. The method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction according to claim 4, characterized in that, The analysis of flue gas flow field topological characteristics and topological uniformity index based on simulated flue gas flow field data under different boiler operating conditions, and the evaluation of flue gas flow field uniformity in the electrostatic precipitator inlet flue gas duct based on flow field topological characteristics and topological uniformity index, includes: The streamline tracing algorithm is used to perform flow field topology analysis on flue gas flow field simulation data, identify the topology features of flue gas flow field under different boiler load operating conditions, and transform the topology features of flue gas flow field into flue gas flow field feature vectors. The flue gas flow field feature vector is compared with a preset topological uniformity index, and the flue gas flow field uniformity of the electrostatic precipitator inlet flue is evaluated based on the comparison results.

9. The method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction according to claim 8, characterized in that, The step of comparing the flue gas flow field feature vector with a preset topological uniformity index and evaluating the flue gas flow field uniformity of the electrostatic precipitator inlet flue based on the comparison results includes: If the flue gas flow field eigenvector is greater than or equal to the topological uniformity index, it indicates that the flue gas flow field in the inlet flue of the electrostatic precipitator is in a uniform state. If the flue gas flow field eigenvector is less than the topological uniformity index, it indicates that the flue gas flow field at the inlet flue of the electrostatic precipitator is in a non-uniform state, and step S3 is executed.

10. A system for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction, used to implement the method for optimizing the electrostatic precipitator efficiency of thermal power units based on multivariate prediction as described in any one of claims 1-9, characterized in that, The system includes: The flue 3D model construction module is used to construct a 3D model of the flue from the boiler air preheater outlet section to the electrostatic precipitator inlet section based on the pre-acquired structural data of the boiler air preheater outlet flue and the electrostatic precipitator inlet flue. The flue gas flow field simulation module is used to collect flue gas parameters from the inlet flue of the electrostatic precipitator and use them as boundary conditions for the three-dimensional model of the flue. The fluid dynamics of the flue gas flow field are simulated using the three-dimensional model of the flue, and the uniformity of the flue gas flow field in the inlet flue of the electrostatic precipitator is evaluated based on the results of the fluid dynamics simulation. The flow field optimization module is used to analyze the causes of current flow distribution deviations based on the uniformity of the flue gas flow field in the inlet flue of the electrostatic precipitator, and then construct flue gas flow field optimization measures based on the causes of deviations, so as to achieve flow field optimization from the boiler air preheater outlet flue to the electrostatic precipitator inlet flue.

Images