Adaptive method and system for ultra-low emission of thermal power generating unit under low load operation condition

Abstract

The application discloses a low-load operation condition ultra-low emission adaptability method and system for a thermal power unit, and relates to the technical field of denitration control.The method comprises the following steps: based on a first denitration target, optimizing a CFB boiler of a target thermal power unit from multiple dimensions to determine a first denitration optimization scheme, performing multi-working condition simulation of flue gas flow in the CFB boiler, and determining aggregated and screened flue gas data through working condition aggregation and screening analysis; combining a second denitration target, optimizing the structure of an SNCR denitration system to determine a target structure optimization scheme; and combining the first denitration optimization scheme to perform ultra-low emission adaptability denitration control.The application solves the technical problem that, in the prior art, the denitration effect of a thermal power unit under low-load operation conditions is poor, which leads to the fact that NOx emission cannot meet the requirement of ultra-low emission, and achieves the technical effect of realizing NOx ultra-low emission under low-load working conditions by optimizing the boiler combustion process and the structure of the SNCR denitration system.

Description

Technical Field

[0001] This invention relates to the field of denitrification control technology, specifically to a method and system for adapting thermal power units to ultra-low emissions under low-load operating conditions. Background Technology

[0002] In the operation of thermal power units, NOx emission control is a crucial step in achieving environmental compliance. Traditional denitrification technologies, such as selective non-catalytic reduction (SNCR) and selective catalytic reduction (SCR), are highly effective under high-load conditions. However, under low-load conditions, due to incomplete combustion and low flue gas temperature, NOx emissions struggle to meet increasingly stringent ultra-low emission standards. In existing technologies, CFB boilers experience increased NOx generation under low-load conditions due to uneven air distribution, decreased separator efficiency, and limited combustion adjustment. Furthermore, the nozzle layout and ammonia delivery parameters of the SNCR system are ill-suited to low-load conditions, further exacerbating the risk of exceeding NOx emission limits. Summary of the Invention

[0003] This application provides a method and system for adapting thermal power units to ultra-low emissions under low-load operating conditions, which is used to solve the technical problem in the prior art where the denitrification effect of thermal power units is poor under low-load operating conditions, resulting in NOx emissions failing to meet ultra-low emission requirements.

[0004] The first aspect of this application provides a method for ultra-low emission adaptation of thermal power units under low-load operation conditions. The method includes: optimizing the CFB boiler of the target thermal power unit from four dimensions—air distribution device, separator inlet flue, central cylinder, and boiler combustion adjustment—based on a first denitrification target, to determine a first denitrification optimization scheme; performing multi-condition simulation of flue gas flow within the CFB boiler based on the first denitrification optimization scheme to obtain multiple flue gas simulation data sets; traversing the multiple flue gas simulation data sets for condition aggregation and screening analysis to determine multiple aggregated and screened flue gas simulation data; optimizing the structure of the SNCR denitrification system of the target thermal power unit based on the multiple aggregated and screened flue gas simulation data and a second denitrification target, to determine a target structural optimization scheme, wherein the target structural optimization scheme includes the number of spray guns, the spray gun placement position, and the initial ammonia solution delivery parameters; and performing ultra-low emission adaptive denitrification control on the target thermal power unit based on the first denitrification optimization scheme and the target structural optimization scheme.

[0005] The second aspect of this application provides an ultra-low emission adaptive system for thermal power units under low-load operation conditions. The system includes: a denitrification optimization module, used to optimize the CFB boiler of the target thermal power unit from four dimensions—air distribution device, separator inlet flue, central cylinder, and boiler combustion adjustment—based on a first denitrification target, to determine a first denitrification optimization scheme; a multi-condition simulation module, used to perform multi-condition simulation of flue gas flow within the CFB boiler based on the first denitrification optimization scheme, obtaining multiple flue gas simulation data sets; a condition aggregation and screening module, used to traverse the multiple flue gas simulation data sets for condition aggregation and screening analysis, determining multiple aggregated and screened flue gas simulation data; a structure optimization module, used to perform structural optimization of the SNCR denitrification system of the target thermal power unit based on the multiple aggregated and screened flue gas simulation data and a second denitrification target, determining a target structure optimization scheme, wherein the target structure optimization scheme includes the number of spray guns, the spray gun placement position, and the initial ammonia solution delivery parameters; and an adaptive denitrification control module, used to perform ultra-low emission adaptive denitrification control of the target thermal power unit based on the first denitrification optimization scheme and the target structure optimization scheme.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The method and system for adapting thermal power units to ultra-low emissions under low-load operating conditions provided in this application relate to the field of denitrification control technology. By optimizing the air distribution device, separator inlet flue, central cylinder, and combustion adjustment of the CFB boiler, combined with optimizing the nozzle configuration and ammonia water delivery parameters of the SNCR system, ultra-low NOx emissions are achieved under low-load conditions. This solves the technical problem in the prior art where poor denitrification effect of thermal power units under low-load operating conditions leads to NOx emissions failing to meet ultra-low emission requirements. It achieves the technical effect of ultra-low NOx emissions under low-load conditions by optimizing the boiler combustion process and the structure of the SNCR denitrification system. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0009] Figure 1 This is a schematic diagram of the process for adapting thermal power units to ultra-low emissions under low-load operating conditions, provided in an embodiment of this application.

[0010] Figure 2 This is a schematic diagram of the structure of the ultra-low emission adaptability system for thermal power units under low load operation conditions provided in the embodiments of this application.

[0011] Figure labeling: Denitrification optimization module 11, multi-condition simulation module 12, condition aggregation screening module 13, structure optimization module 14, adaptive denitrification control module 15. Detailed Implementation

[0012] This application provides a method and system for adapting thermal power units to ultra-low emissions under low-load operating conditions, which is used to solve the technical problem in the prior art where the denitrification effect of thermal power units is poor under low-load operating conditions, resulting in NOx emissions failing to meet ultra-low emission requirements.

[0013] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0014] It should be noted that the terms "first," "second," etc., in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to such processes, methods, products, or devices.

[0015] Example 1, as Figure 1 As shown, this application provides a method for adapting thermal power units to ultra-low emissions under low-load operating conditions. The method includes:

[0016] P10: Based on the primary denitrification target, the CFB boiler of the target thermal power unit is optimized from four dimensions: air distribution device, separator inlet flue, central cylinder, and boiler combustion adjustment, to determine the primary denitrification optimization scheme. Specifically, the primary denitrification target is to reduce the NOx concentration at the furnace outlet of the CFB boiler from 400 mg / Nm³ to 250 mg / Nm³, and to achieve a separator inlet flue gas velocity of 25 m / s.

[0017] Furthermore, step P10 in this embodiment of the application also includes:

[0018] P11: After the CFB boiler of the target thermal power unit is shut down, dispersed multi-point image acquisition is performed on the air distribution plate and the central cylinder to obtain a multi-point air distribution plate image set and a multi-point central cylinder image set; P12: Based on the multi-point air distribution plate image set, coking and agglomeration anomaly identification is performed to obtain the air distribution plate anomaly identification result; P13: Based on the multi-point central cylinder image set, deformation anomaly identification is performed to obtain the central cylinder deformation anomaly identification result; P14: The historical combustion parameters of the boiler and the historical NOx concentration at the furnace outlet are obtained. Combined with the air distribution plate anomaly identification result and the central cylinder deformation anomaly identification result, the first denitrification optimization scheme is determined with the first denitrification target as a constraint.

[0019] It should be understood that, to achieve the ultra-low emission denitrification target for the circulating fluidized bed (CFB) boiler in the target thermal power unit, two denitrification targets were initially set. The first denitrification target is to reduce the NOx concentration at the boiler furnace outlet from 400 mg / Nm³ to 250 mg / Nm³, while ensuring that the flue gas velocity at the separator inlet reaches 25 m / s. Achieving the first denitrification target relies primarily on optimizing key boiler components, including the air distribution system, separator inlet flue, central cylinder, and boiler combustion adjustment. Optimizing these components improves boiler combustion efficiency and reduces NOx generation. Optimization of the air distribution system mainly focuses on adjusting the duct design and air distribution plate layout to ensure uniform air distribution, avoid localized over-combustion or oxygen deficiency, and thus reduce NOx generation. Optimizing the flow velocity in the separator inlet flue helps improve flue gas separation efficiency and maintain stable boiler airflow. For the central cylinder, optimizing its structural design improves airflow distribution and ensures uniform combustion. Finally, boiler combustion adjustment further improves combustion efficiency, optimizes combustion temperature and oxygen ratio, and reduces NOx emissions.

[0020] Specifically, the air distribution system of the CFB boiler was first optimized. After boiler shutdown, dispersed multi-point image acquisition was performed on the air distribution plate to obtain a multi-point image set. Image recognition technology was used to analyze these images to identify abnormal areas of coking and accumulation on the air distribution plate. Coking and accumulation can lead to uneven air distribution, which in turn affects combustion efficiency and NOx generation. Based on the identification results, the number and specific locations of the air caps and core tubes that need to be replaced were determined to increase the resistance of the air distribution plate and the air velocity of the air cap orifices, thereby enhancing the rigidity of the primary air. In this way, while reducing the primary air volume, the temperature in the dense phase zone can be effectively reduced, NOx generation can be reduced, the formation of fluidization dead zones in the air distribution plate can be alleviated, the uniformity of air distribution in the air distribution plate can be ensured, and the increase in NOx generation caused by local high oxygen levels can be avoided.

[0021] Secondly, the separator inlet flue was optimized. Similarly, after boiler shutdown, dispersed multi-point image acquisition was performed on the central cylinder to obtain a multi-point image set. Image recognition technology was used to analyze the central cylinder images to identify abnormal deformation. The central cylinder may deform during long-term operation, which affects its separation efficiency and circulating ash volume. Based on the identification results, targeted repairs or modifications were made to the central cylinder to improve separator efficiency. Simultaneously, wear-resistant refractory material was laid inside the separator inlet flue to reduce the flow area of ​​the flue and increase the flue gas velocity to 25 m / s. This increased velocity helps improve the separation efficiency of the cyclone separator, increases the circulating ash volume, thereby reducing bed temperature, reducing NOx formation, improving the fluidization state of materials in the furnace, and optimizing the combustion process.

[0022] Based on the aforementioned optimization measures, and combining historical combustion parameters and historical NOx concentration data at the furnace outlet, a comprehensive first denitrification optimization scheme is determined, using the primary denitrification target as a constraint. Historical combustion parameters and NOx concentration data reflect the boiler's actual performance under different operating conditions, providing crucial reference for formulating the optimization scheme. By comprehensively evaluating abnormal results from the air distributor and central cylinder, the boiler's combustion process can be further optimized to ensure the achievement of the primary denitrification target. Based on feedback from historical data, air distribution and temperature control during the boiler combustion process are adjusted to optimize combustion conditions, generating the primary denitrification optimization scheme, thereby ensuring NOx emissions are reduced to below 250 mg / Nm³.

[0023] Furthermore, step P11 in the embodiments of this application also includes:

[0024] P11-1: Determine the number of dispersed acquisition points for the air distribution panel based on its area size; P11-2: Randomly deploy acquisition points based on the number of dispersed acquisition points for the air distribution panel to obtain a set of dispersed acquisition points for the air distribution panel, wherein the interval between any two dispersed acquisition points in the set of dispersed acquisition points for the air distribution panel is greater than or equal to a preset interval threshold; P11-3: Determine the set of dispersed acquisition points for the central cylinder based on its diameter and a preset number of acquisition points for the central cylinder; P11-4: Perform dispersed multi-point image acquisition based on the set of dispersed acquisition points for the air distribution panel and the set of dispersed acquisition points for the central cylinder to obtain a multi-point set of air distribution panel images and a multi-point set of central cylinder images.

[0025] Optionally, the air distribution device of the CFB boiler can be optimized. First, after boiler shutdown, the number of distributed sampling points for the air distribution plate is determined based on its area. The area of ​​the air distribution plate can be calculated by measuring its length and width. Based on the area of ​​the air distribution plate, and according to a preset sampling point density (e.g., one sampling point per square meter), the number of distributed sampling points is determined. The number of points needs to be determined based on the specific dimensions of the air distribution plate and the required image resolution, ensuring that every area of ​​the air distribution plate surface is adequately covered.

[0026] Next, based on the area of ​​the wind distribution panel and the required number of data collection points, random data collection points are deployed. Random deployment prevents data collection points from concentrating in one area while ignoring other potentially anomaly-prone areas. The specific method for random deployment can employ a random number generation algorithm to generate a series of random coordinate points; these coordinate points will be the data collection points for the wind distribution panel. To ensure reasonable spacing between data collection points, the interval between any two data collection points in the dispersed data collection point set of the wind distribution panel should be greater than or equal to a preset interval threshold. The preset interval threshold can be set according to the specific size of the wind distribution panel and the required acquisition accuracy. For example, for a larger wind distribution panel, the preset interval threshold can be set to 1 meter to ensure sufficient independence of image data between data collection points and avoid image overlap or interference.

[0027] Similarly, the set of dispersed sampling points for the central cylinder is determined based on the diameter of the central cylinder and the preset number of sampling points. The diameter of the central cylinder can be determined by measuring its inner diameter, and the preset number of sampling points can be set according to the diameter of the central cylinder and the required sampling accuracy. For example, for a central cylinder with a larger diameter, more sampling points can be set to ensure comprehensive coverage of all areas of the central cylinder. The specific number of sampling points can be calculated to ensure reasonable spacing between them.

[0028] After determining the sets of dispersed sampling points for the air distribution panel and the central cylinder, professional image acquisition equipment was used to collect multi-point images of both panels according to the pre-set sampling points. During the acquisition process, the images of the air distribution panel were first collected from multiple points to ensure observation of its surface from different angles. High-precision cameras were used, ensuring that the area represented by each sampling point was included within the image acquisition range so that subsequent analysis could cover all critical parts of the air distribution panel. Similarly, for the image acquisition of the central cylinder, the sampling points should have broad coverage and diverse shooting angles. The image acquisition equipment should be able to capture images of the central cylinder from different angles, especially areas prone to deformation or carbon buildup. By integrating the images from multiple sampling points, a complete set of images for both the air distribution panel and the central cylinder was obtained. These image sets will provide necessary data support for subsequent anomaly identification and denitrification optimization schemes.

[0029] Furthermore, step P12 in this embodiment of the application also includes:

[0030] P12-1: Call the coking aggregation anomaly detector to perform coking aggregation anomaly identification on the multi-point wind distribution plate image set respectively, and obtain a set of multi-point initial wind distribution plate anomaly identification results; P12-2: Perform same-type aggregation on the set of multi-point initial wind distribution plate anomaly identification results to obtain multiple aggregated multi-point initial wind distribution plate anomaly identification result sets; P12-3: Calculate the mean of the multiple aggregated multi-point initial wind distribution plate anomaly identification result sets to obtain the mean of the multiple aggregated multi-point initial wind distribution plate anomaly identification results; P12-4: Perform weighted analysis based on the mean of the multiple aggregated multi-point initial wind distribution plate anomaly identification results to obtain the wind distribution plate anomaly identification result.

[0031] Specifically, the process of identifying anomalies in the air distribution plate can be further refined. This involves identifying coking and aggregation anomalies in the acquired air distribution plate images to detect the plate's condition and identify potential coking problems. Coking refers to the adhesion of solid substances such as carbon slag and dust to the surface of the air distribution plate due to high temperatures or coal dust accumulation during combustion, which affects airflow and combustion efficiency.

[0032] First, the coking and agglomeration anomaly detector is invoked to analyze the acquired multi-point air distribution plate image set point by point to identify potential coking and agglomeration anomalies on the air distribution plate. Based on image recognition technology, the coking and agglomeration anomaly detector analyzes features such as color, texture, and shape in the image to detect areas where coking materials such as slag and dust accumulate on the air distribution plate. These areas typically cause airflow obstruction, thus affecting the boiler's combustion efficiency. The detector processes the image at each acquisition point to obtain the initial anomaly identification result for each point, generating a multi-point initial air distribution plate anomaly identification result set, where each result corresponds to the anomaly identification situation at one acquisition point.

[0033] Subsequently, the anomaly identification results set of the multi-point initial air distribution panel are aggregated according to the same type. Due to the large size of the air distribution panel and the large number of collection points, there may be multiple similar anomaly identification results. Aggregating these similar results can reduce data redundancy and facilitate subsequent analysis and processing. Through the same type aggregation, multiple aggregated multi-point initial air distribution panel anomaly identification result sets can be obtained, each set containing a group of identification results with similar anomaly characteristics.

[0034] Next, the mean values ​​of multiple aggregated multi-point initial wind distribution panel anomaly identification result sets are calculated. The purpose of mean calculation is to further simplify the data and extract typical features of each aggregated set. By calculating the mean of all anomaly identification results in each aggregated set, the mean values ​​of multiple aggregated multi-point initial wind distribution panel anomaly identification results can be obtained. These mean values ​​can more accurately reflect the anomalies in different areas of the wind distribution panel, providing more representative data for subsequent analysis.

[0035] Finally, a weighted analysis was performed based on the average of multiple aggregated initial air distribution panel anomaly identification results. This weighted analysis, by considering the impact of anomalies in different regions on the overall performance of the air distribution panel and assigning different weights to different averages, can more accurately assess the overall anomaly status of the air distribution panel and obtain anomaly identification results. These results can clearly indicate the specific locations and extent of replacement or repair on the air distribution panel, providing a scientific basis for the optimization and transformation of the air distribution device.

[0036] P20: Based on the first denitrification optimization scheme, perform multi-condition simulation of flue gas flow in the CFB boiler to obtain multiple flue gas simulation data sets.

[0037] Optionally, after determining the first denitrification optimization scheme, in order to verify the effectiveness and feasibility of the scheme in actual operation, it is necessary to conduct multi-condition simulation of the flue gas flow in the CFB boiler. This simulation process aims to predict the impact of the optimized air distribution device, separator inlet flue, central cylinder, and combustion adjustment measures on flue gas flow and NOx generation under different operating conditions using computer simulation technology.

[0038] Specifically, a three-dimensional model of the CFB boiler is first established based on the specific measures for the air distribution device, separator inlet flue, central tube, and combustion adjustment determined in the first denitrification optimization scheme. This model should include in detail the geometry and dimensions of key components such as the air distribution plate, air cap, core tube, separator inlet flue, and central tube, as well as the structure and layout of the combustion chamber. Accurate modeling ensures that the simulation results accurately reflect the actual operation of the boiler.

[0039] Next, computational fluid dynamics (CFD) software was used to simulate the established 3D model under multiple operating conditions. During the simulation, different operating conditions were set, including different load levels, different fuel types, and different combustion parameters. Under each condition, the flow of flue gas within the boiler was simulated, including the velocity field, temperature field, and concentration field of the flue gas. Through simulation, detailed data on flue gas flow under different operating conditions can be obtained, which will serve as an important basis for subsequent analysis and optimization.

[0040] During the simulation, special attention was paid to the flow of flue gas in key areas such as the air distributor, the separator inlet flue, and the central cylinder. By analyzing the simulation data, the effectiveness of optimization measures in improving flue gas flow and their impact on NOx formation were evaluated. For example, it was examined whether optimizing the air distributor could achieve uniform air distribution, whether optimizing the separator inlet flue could achieve the expected flue gas velocity, and whether optimizing the central cylinder could improve separation efficiency.

[0041] Finally, through multi-condition simulation, multiple flue gas simulation datasets were obtained. Each dataset corresponds to a specific operating condition and contains detailed information on flue gas flow under that condition. These datasets will be used for subsequent analysis to evaluate the performance of the first denitrification optimization scheme under different operating conditions, ensuring that the optimization scheme can effectively reduce NOx emissions under various operating conditions while maintaining the boiler's efficient operation.

[0042] P30: Traverse the multiple flue gas simulation data sets to perform operating condition aggregation and screening analysis, and determine multiple aggregated and screened flue gas simulation data.

[0043] Furthermore, step P30 in this embodiment of the application also includes:

[0044] P31: Traverse the multiple flue gas simulation datasets to identify operating conditions, and divide them into multiple aggregated flue gas simulation datasets based on the identification results; P32: Extract multiple aggregation screening analysis centers from the multiple aggregated flue gas simulation datasets, wherein each aggregation screening analysis center is the mean of a aggregated flue gas simulation dataset; P33: Determine the initial neighborhood of the multiple aggregation screening analysis centers according to a preset aggregation screening radius, and obtain multiple initial aggregation screening analysis center neighborhoods; P34: Identify the expansion boundaries of the neighborhoods of the multiple initial aggregation screening analysis centers in the multiple aggregated flue gas simulation datasets according to the preset aggregation screening radius, and determine multiple expanded aggregation screening analysis center neighborhoods; P35: Calculate the mean of the multiple expanded aggregation screening analysis center neighborhoods to determine the multiple aggregated screening flue gas simulation datasets.

[0045] It should be understood that after obtaining multiple flue gas simulation datasets, in order to further analyze and select representative data, an operating condition aggregation and screening analysis can be performed to extract typical datasets that can reflect the flue gas flow characteristics under different operating conditions, providing a basis for subsequent optimization scheme evaluation.

[0046] First, multiple flue gas simulation datasets are traversed to identify operating conditions. Operating condition identification involves analyzing key parameters in each dataset, such as flue gas velocity, temperature, and NOx concentration, to determine its corresponding operating conditions. Based on the identification results, datasets with similar operating conditions are divided into multiple aggregated flue gas simulation datasets. Each aggregated dataset contains a set of flue gas simulation data generated under similar operating conditions, which facilitates subsequent analysis and processing.

[0047] Next, the mean of each of the multiple aggregated flue gas simulation datasets is extracted as the aggregation screening analysis center. The aggregation screening analysis center is a data point that represents the central characteristic of that aggregated dataset, i.e., the overall characteristic of that operating condition. By calculating the mean of all data points in each aggregated dataset, a typical data point that reflects the overall characteristics of that dataset can be obtained. These aggregation screening analysis centers will serve as the basis for subsequent neighborhood analysis.

[0048] Then, based on a preset aggregation filtering radius, the initial neighborhood of each aggregation filtering analysis center is determined. The preset aggregation filtering radius is a threshold used to define the neighborhood range around each analysis center. The initial neighborhood refers to the set of data points with similar characteristics to the analysis center within this radius. By determining the initial neighborhood, data points close to the analysis center can be initially filtered out.

[0049] Next, within each aggregated flue gas simulation dataset, the initial neighborhood is expanded and its boundaries identified according to a preset aggregation screening radius. During this expansion, potential boundary conditions within the neighborhood need to be identified and analyzed. The expanded neighborhood covers more simulation data, providing broader operating condition data for subsequent mean calculation and screening. Through this process, the expanded aggregation screening analysis center neighborhood can be determined; these neighborhoods contain a wider range of data points and can more comprehensively reflect the flue gas flow characteristics under that operating condition.

[0050] Finally, the mean values ​​of multiple expanded aggregation screening analysis center neighborhoods are calculated to determine the final aggregation screening flue gas simulation data. By calculating the mean of all data points in each expanded neighborhood, a set of typical data representing the flue gas flow characteristics under this operating condition can be obtained. These aggregation screening flue gas simulation data will reflect the flue gas flow characteristics and NOx emissions under different operating conditions, serving as an important basis for subsequent optimization scheme evaluation and verifying the effectiveness and feasibility of the optimization scheme under different operating conditions.

[0051] Furthermore, step P34 in this embodiment of the application also includes:

[0052] P34-1: Expand the neighborhood of the multiple initial aggregation screening analysis centers according to the preset aggregation screening radius to obtain multiple stage-expanded aggregation screening analysis center neighborhoods; P34-2: Determine whether the data volume of the multiple stage-expanded aggregation screening analysis center neighborhoods is greater than or equal to the data volume of the multiple initial aggregation screening analysis center neighborhoods. If so, continue to expand the neighborhood of the multiple stage-expanded aggregation screening analysis centers according to the preset aggregation screening radius until the preset expansion stop condition is met. Take the neighborhood boundary of the last edge expansion as the expansion boundary, and take the multiple stage-expanded aggregation screening analysis center neighborhoods obtained by the last expansion as the multiple expanded aggregation screening analysis center neighborhoods. The preset expansion stop condition is that the number of expansions meets the preset expansion number threshold.

[0053] Optionally, the process of expanding the boundary identification can be further refined by gradually expanding the neighborhood boundary and judging the changes in the amount of data, and finally determining the neighborhood of the expansion aggregation screening analysis center.

[0054] First, the neighborhoods of multiple initial aggregation screening analysis centers are expanded according to a preset aggregation screening radius. By gradually increasing the range of the neighborhood, the neighborhood boundaries are expanded outward to include more data points. During the expansion process, the boundaries of the neighborhoods gradually expand as the radius increases, ultimately forming multiple stages of expanded aggregation screening analysis center neighborhoods. These staged expanded neighborhoods contain more data points than the initial neighborhoods, and can more comprehensively reflect the flue gas flow characteristics under this operating condition.

[0055] Next, the amount of data in the neighborhood of the aggregation and screening analysis center in multiple stages is assessed. Specifically, the amount of data in the expanded neighborhood of each stage is compared with the amount of data in the initial neighborhood. If the amount of data in the expanded neighborhood of each stage is greater than or equal to the amount of data in the initial neighborhood, it indicates that more data points have been successfully included during the expansion process, better reflecting the flue gas flow characteristics under this operating condition. In this case, the expanded neighborhood of each stage continues to expand at the edge according to the preset aggregation and screening radius until the preset expansion stopping condition is met.

[0056] The preset expansion stopping condition is that the number of expansions meets a preset expansion number threshold. This threshold is set based on actual needs and computational resource limitations to control the number of iterations in the expansion process. When the number of expansions reaches the preset expansion number threshold, the expansion process stops. At this point, the neighborhood boundary of the last edge expansion is taken as the expansion boundary, and the neighborhood of the multiple expansion aggregation and screening analysis centers obtained from the last expansion is taken as the final multiple expansion aggregation and screening analysis center neighborhood.

[0057] This process not only effectively expands the neighborhood to include more data points, but also controls the expansion process through preset expansion stopping conditions, ensuring computational feasibility and efficiency. The final determined expanded aggregation screening analysis center neighborhood will provide more comprehensive and accurate data support for subsequent mean calculation and optimization scheme evaluation.

[0058] P40: Based on the multiple aggregated screening flue gas simulation data and the second denitrification target, the SNCR denitrification system of the target thermal power unit is structurally optimized to determine the target structural optimization scheme. The target structural optimization scheme includes the number of spray guns, the spray gun placement, and the initial ammonia solution delivery parameters. The second denitrification target is to reduce the NOx concentration at the outlet of the SNCR denitrification system from 250 mg / Nm³ to less than or equal to 50 mg / Nm³.

[0059] Furthermore, step P40 in this embodiment of the application also includes:

[0060] P41: Optimize the structure of the SNCR denitrification system based on the multiple polymer-screened flue gas simulation data and the second denitrification target, respectively, to obtain multiple initial structure optimization schemes; P42: Iterate through the denitrification results of each initial structure optimization scheme on the multiple polymer-screened flue gas simulation data, to obtain multiple denitrification results; P43: Determine whether there is a denitrification result that satisfies the second denitrification target among the multiple denitrification results. If so, take the initial structure optimization scheme corresponding to the maximum value among the multiple denitrification results as the target structure optimization scheme.

[0061] Specifically, after obtaining multiple simulation data of polymer-selective flue gas, in order to achieve the second denitrification target of the target thermal power unit, namely reducing the NOx concentration at the outlet of the SNCR denitrification system from 250 mg / Nm³ to less than or equal to 50 mg / Nm³, the SNCR denitrification system was structurally optimized. The goal of the structural optimization was to determine the number of spray guns, the placement of the spray guns, and the initial delivery parameters of the ammonia solution to ensure efficient denitrification under different operating conditions.

[0062] First, based on multiple simulation data of polymerization screening flue gas and a second denitrification target, the SNCR denitrification system was structurally optimized, resulting in several initial optimized structural schemes. Each initial optimized structural scheme includes the number of spray guns, the spray gun placement, and the initial ammonia solution delivery parameters. The optimization of these parameters is based on key parameters in the polymerization screening flue gas simulation data, such as flue gas velocity, temperature, and NOx concentration. Specifically, each initial optimized scheme will be tested under different simulated operating conditions. By adjusting the number and placement of the spray guns and optimizing the initial ammonia solution delivery parameters, multiple initial optimized structural schemes are generated.

[0063] Next, detailed simulations are needed for each initial structural optimization scheme to evaluate its denitrification performance under multiple operating conditions. Specifically, for each initial structural optimization scheme, the denitrification results are simulated under all polymer screening flue gas simulation data, yielding multiple denitrification results. These results reflect the denitrification effects of different structural optimization schemes under different operating conditions, providing data support for subsequent scheme evaluation. Through this process, the performance of each initial structural optimization scheme under different operating conditions can be comprehensively evaluated.

[0064] Finally, it is determined whether any of the multiple denitrification results satisfy the second denitrification objective. If such a result exists, the initial structure optimization scheme corresponding to the maximum value among these results is taken as the target structure optimization scheme. Here, the maximum value refers to the scheme with the lowest NOx concentration while satisfying the second denitrification objective. This screening process ensures that the final target structure optimization scheme achieves optimal denitrification performance under all operating conditions. If no denitrification result satisfies the second denitrification objective, the optimization parameters need to be readjusted, and simulations and evaluations conducted again until a scheme that meets the objective is found.

[0065] Furthermore, step P40 in this embodiment of the application also includes:

[0066] P44: If none of the multiple denitrification results satisfy the second denitrification target, then the multiple initial structure optimization schemes are randomly adjusted according to the preset adjustment step size to obtain multiple adjusted structure optimization schemes; P45: The target structure optimization scheme is determined again based on the multiple adjusted structure optimization schemes.

[0067] Optionally, if no denitrification result that meets the second denitrification target is found, the next processing step will enter the adjustment phase to optimize the existing initial structure optimization scheme.

[0068] Specifically, if no denitrification result meeting the second denitrification objective is found, adjustments to multiple initial structural optimization schemes are necessary. This adjustment process is carried out using a preset adjustment step size. The adjustment step size refers to the magnitude of change when adjusting parameters (such as the number of spray guns, their placement, or the ammonia solution delivery parameters). Depending on the size of the adjustment step size, multiple initial optimization schemes can be randomly adjusted. For example, the number of spray guns, their placement, and the initial ammonia solution delivery parameters can be randomly adjusted to generate new structural optimization schemes.

[0069] Next, the target structural optimization scheme is determined again based on multiple adjusted structural optimization schemes. This process is similar to the previous evaluation process. The denitrification results of each adjusted structural optimization scheme are simulated under all aggregated screening flue gas simulation data to obtain multiple new denitrification results. Then, it is determined whether any of these new denitrification results satisfy the second denitrification target. If so, the adjusted structural optimization scheme corresponding to the maximum value of these denitrification results is taken as the target structural optimization scheme. If no denitrification result satisfies the second denitrification target, the adjustment and evaluation process can be repeated until a scheme that satisfies the target is found. This method of randomly adjusting the initial optimization scheme can effectively expand the optimization space, provide more possibilities for satisfying the second denitrification target, and ultimately determine the optimal SNCR denitrification system structure scheme.

[0070] P50: Based on the first denitrification optimization scheme and the target structure optimization scheme, perform ultra-low emission adaptive denitrification control on the target thermal power unit.

[0071] Specifically, after determining the first denitrification optimization scheme and the target structure optimization scheme, in order to ensure that the target thermal power unit can achieve the denitrification target of ultra-low emissions in actual operation, it is necessary to implement ultra-low emission adaptive denitrification control for the thermal power unit. This control process aims to apply the various measures and parameters in the optimization scheme to actual operation, ensuring that ultra-low NOx emissions can be stably achieved under different operating conditions.

[0072] Specifically, the first step is to adjust and implement the air distribution device, separator inlet flue, central cylinder, and combustion adjustment measures of the CFB boiler according to the first denitrification optimization plan. This includes replacing the air caps and core tubes of the air distribution plate, optimizing the structure of the separator inlet flue, adjusting the shape and size of the central cylinder, and optimizing various parameters in the combustion process. The implementation of these measures will provide the physical basis for achieving the first denitrification target, ensuring that the NOx concentration at the furnace outlet can be reduced from 400 mg / Nm³ to 250 mg / Nm³, and that the flue gas velocity at the separator inlet reaches 25 m / s.

[0073] Next, the SNCR denitrification system was adjusted according to the target structure optimization scheme. This included adjusting the number and placement of the spray guns, as well as optimizing the initial delivery parameters of the ammonia solution. These adjustments ensured that the SNCR denitrification system could operate efficiently under different conditions, reducing the outlet NOx concentration from 250 mg / Nm³ to less than or equal to 50 mg / Nm³. Specifically, the existing spray gun system needed to be modified or reinstalled according to the number and placement of the spray guns in the optimization scheme. Simultaneously, based on the optimized initial delivery parameters of the ammonia solution, parameters such as the ammonia flow rate, pressure, and injection angle were adjusted to ensure that the ammonia solution was uniformly injected into the flue gas and reacted fully with the NOx.

[0074] After implementing the above optimization measures, ultra-low emission adaptive denitrification control is required for thermal power units. This control process includes real-time monitoring of key parameters such as NOx concentration, ammonia slip, and flue gas velocity in the flue gas, and dynamic adjustments to the denitrification system based on the real-time data of these parameters. For example, if an upward trend in NOx concentration is detected, the denitrification efficiency can be improved by increasing the ammonia flow rate or adjusting the spray angle of the spray gun; if an increase in ammonia slip is detected, the ammonia flow rate needs to be appropriately reduced to avoid excessive ammonia slip impacting the environment.

[0075] Furthermore, ultra-low emission adaptive denitrification control also needs to consider changes in the operating conditions of thermal power units. Under different load levels, fuel types, and combustion parameters, the flow characteristics of flue gas and NOx formation will vary. Therefore, it is necessary to automatically adjust the operating parameters of the denitrification system based on real-time monitoring data and changes in operating conditions to ensure stable ultra-low NOx emissions under various conditions. This adaptive adjustment can automatically adjust the denitrification control strategy according to different operating conditions such as unit load fluctuations and coal quality changes, ensuring the efficient operation of the denitrification system.

[0076] In summary, the embodiments of this application have at least the following technical effects:

[0077] This application optimizes the structure of the boiler combustion and denitrification systems, enabling thermal power units to stably meet ultra-low NOx emission standards even under low-load operating conditions. By employing multi-condition flue gas flow simulation and condition-based aggregation screening analysis, the adaptability and efficiency of the denitrification system under different operating conditions are improved, ensuring the stability of the denitrification effect. Through coordinated optimization of the combustion and denitrification systems, NOx generation is reduced while simultaneously improving the denitrification effect and enhancing the overall system operating efficiency. Based on the optimized denitrification scheme, real-time adjustment and precise control capabilities are provided, ensuring stable denitrification performance under different loads and operating conditions.

[0078] The technology has achieved ultra-low NOx emissions under low load conditions by optimizing the boiler combustion process and the structure of the SNCR denitrification system.

[0079] Example 2, based on the same inventive concept as the method for adapting thermal power units to ultra-low emissions under low-load operating conditions in the aforementioned examples, such as... Figure 2 As shown, this application provides an ultra-low emission adaptability system for thermal power units under low-load operation conditions. The system and method embodiments in this application are based on the same inventive concept. The system includes:

[0080] The denitrification optimization module 11 is used to optimize the CFB boiler of the target thermal power unit from four dimensions: air distribution device, separator inlet flue, central cylinder, and boiler combustion adjustment, based on the first denitrification target, to determine the first denitrification optimization scheme. The first denitrification target is to reduce the NOx concentration at the furnace outlet of the CFB boiler from 400 mg / Nm³ to 250 mg / Nm³, and to achieve a separator inlet flue gas velocity of 25 m / s. The second denitrification target is to reduce the NOx concentration at the outlet of the SNCR denitrification system from 250 mg / Nm³ to less than or equal to 50 mg / Nm³.

[0081] The multi-condition simulation module 12 is used to perform multi-condition simulation of flue gas flow in the CFB boiler based on the first denitrification optimization scheme, and obtain multiple flue gas simulation data sets.

[0082] The operating condition aggregation and filtering module 13 is used to traverse the multiple flue gas simulation data sets to perform operating condition aggregation and filtering analysis, and to determine multiple aggregated and filtered flue gas simulation data.

[0083] The structure optimization module 14 is used to optimize the structure of the SNCR denitrification system of the target thermal power unit based on the multiple aggregated screening flue gas simulation data and the second denitrification target, and to determine the target structure optimization scheme, wherein the target structure optimization scheme includes the number of spray guns, the spray gun layout position, and the initial delivery parameters of ammonia solution.

[0084] The adaptive denitrification control module 15 is used to perform ultra-low emission adaptive denitrification control on the target thermal power unit based on the first denitrification optimization scheme and the target structure optimization scheme.

[0085] Furthermore, the denitrification optimization module 11 is also used to perform the following steps:

[0086] After the CFB boiler of the target thermal power unit is shut down, dispersed multi-point image acquisition is performed on the air distribution plate and the central cylinder to obtain a multi-point air distribution plate image set and a multi-point central cylinder image set; coking and agglomeration anomaly identification is performed based on the multi-point air distribution plate image set to obtain the air distribution plate anomaly identification result; deformation anomaly identification is performed based on the multi-point central cylinder image set to obtain the central cylinder deformation anomaly identification result; historical combustion parameters of the boiler and historical NOx concentration at the furnace outlet are obtained, and combined with the air distribution plate anomaly identification result and the central cylinder deformation anomaly identification result, the first denitrification optimization scheme is determined with the first denitrification target as a constraint.

[0087] Furthermore, the denitrification optimization module 11 is also used to perform the following steps:

[0088] The number of dispersed acquisition points for the air distribution panel is determined based on its area size. Random point layout is then performed based on this number of points to obtain a set of dispersed acquisition points for the air distribution panel. The interval between any two dispersed acquisition points in this set is greater than or equal to a preset interval threshold. The number of acquisition points for the central cylinder is determined based on its diameter and a preset number of acquisition points. Dispersed multi-point image acquisition is then performed on both the air distribution panel acquisition point set and the central cylinder acquisition point set to obtain a multi-point air distribution panel image set and a multi-point central cylinder image set.

[0089] Furthermore, the denitrification optimization module 11 is also used to perform the following steps:

[0090] The coking and aggregation anomaly detector is invoked to identify coking and aggregation anomalies in the multi-point wind distribution plate image set, obtaining a set of multi-point initial wind distribution plate anomaly identification results; the set of multi-point initial wind distribution plate anomaly identification results is then aggregated to obtain multiple aggregated multi-point initial wind distribution plate anomaly identification result sets; the mean of the multiple aggregated multi-point initial wind distribution plate anomaly identification result sets is calculated to obtain the mean of the multiple aggregated multi-point initial wind distribution plate anomaly identification results; and a weighted analysis is performed based on the mean of the multiple aggregated multi-point initial wind distribution plate anomaly identification results to obtain the wind distribution plate anomaly identification result.

[0091] Furthermore, the working condition aggregation and filtering module 13 is also used to perform the following steps:

[0092] The process involves traversing multiple flue gas simulation datasets to identify operating conditions and dividing them into multiple aggregated flue gas simulation datasets based on the identification results. Multiple aggregation screening analysis centers are extracted from these datasets, where each aggregation screening analysis center represents the mean of one aggregated flue gas simulation dataset. Initial neighborhoods of these aggregation screening analysis centers are determined according to a preset aggregation screening radius, resulting in multiple initial aggregation screening analysis center neighborhoods. Expansion boundaries are identified for the neighborhoods of these initial aggregation screening analysis centers within each of the multiple aggregated flue gas simulation datasets based on the preset aggregation screening radius, determining multiple expanded aggregation screening analysis center neighborhoods. Finally, the mean of these expanded aggregation screening analysis center neighborhoods is calculated to determine the multiple aggregated screening flue gas simulation datasets.

[0093] Furthermore, the working condition aggregation and filtering module 13 is also used to perform the following steps:

[0094] The neighborhoods of the multiple initial aggregation filtering analysis centers are expanded according to a preset aggregation filtering radius to obtain multiple stage-expanded aggregation filtering analysis center neighborhoods. It is determined whether the data volume of the multiple stage-expanded aggregation filtering analysis center neighborhoods is greater than or equal to the data volume of the multiple initial aggregation filtering analysis center neighborhoods. If so, the multiple stage-expanded aggregation filtering analysis center neighborhoods are expanded according to the preset aggregation filtering radius until a preset expansion stop condition is met. The neighborhood boundary of the last edge expansion is taken as the expansion boundary, and the multiple stage-expanded aggregation filtering analysis center neighborhoods obtained by the last expansion are taken as the multiple expanded aggregation filtering analysis center neighborhoods. The preset expansion stop condition is that the number of expansions meets a preset expansion number threshold.

[0095] Furthermore, the structure optimization module 14 is also used to perform the following steps:

[0096] The SNCR denitrification system is structurally optimized based on the multiple polymer-screened flue gas simulation data and the second denitrification target, respectively, to obtain multiple initial structural optimization schemes. The denitrification results of each of the multiple initial structural optimization schemes on the multiple polymer-screened flue gas simulation data are traversed to obtain multiple denitrification results. It is determined whether there is a denitrification result that satisfies the second denitrification target among the multiple denitrification results. If so, the initial structural optimization scheme corresponding to the maximum value among the multiple denitrification results is taken as the target structural optimization scheme.

[0097] Furthermore, the structure optimization module 14 is also used to perform the following steps:

[0098] If none of the multiple denitrification results satisfy the second denitrification target, then the multiple initial structure optimization schemes are randomly adjusted according to a preset adjustment step size to obtain multiple adjusted structure optimization schemes; then the target structure optimization scheme is determined again based on the multiple adjusted structure optimization schemes.

[0099] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0100] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0101] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A method for ultra-low emission adaptability of a thermal power generating unit under low load operation conditions, characterized in that, The method comprises: Based on the first denitration target, the CFB boiler of the target thermal power generating unit is optimized from the air distribution device, the separator inlet flue, the center cylinder and the boiler combustion adjustment in four dimensions to determine a first denitration optimization scheme; Based on the first denitration optimization scheme, smoke flow multi-working condition simulation is carried out in the CFB boiler to obtain a plurality of smoke simulation data sets; Traverse the plurality of smoke simulation data sets to perform working condition aggregation filtering analysis to determine a plurality of aggregated filtered smoke simulation data; Based on the plurality of aggregated filtered smoke simulation data and a second denitration target, the structure of the SNCR denitration system of the target thermal power generating unit is optimized to determine a target structure optimization scheme, wherein the target structure optimization scheme includes the number of spray guns, the layout position of the spray guns, and the initial delivery parameters of the ammonia solution; Based on the first denitration optimization scheme and the target structure optimization scheme, the target thermal power generating unit is controlled for ultra-low emission adaptability denitration.

2. The method of claim 1, wherein the low load operation condition is a load of 30% to 60% of the rated load of the thermal power unit. The first denitration target is to reduce the NOx concentration at the furnace outlet of the CFB boiler from 400 mg / Nm³ to 250 mg / Nm³, and the smoke flow rate at the inlet of the separator reaches 25 m / s; The second denitration target is to reduce the NOx concentration at the outlet of the SNCR denitration system from 250 mg / Nm³ to less than or equal to 50 mg / Nm³.

3. The method of claim 1, wherein the method is characterized by, Based on the first denitration target, the CFB boiler of the target thermal power generating unit is optimized from the air distribution device, the separator inlet flue, the center cylinder and the boiler combustion adjustment in four dimensions to determine a first denitration optimization scheme, including: After the CFB boiler of the target thermal power generating unit is shut down, dispersed multi-point image acquisition is performed on the air distribution plate and the center cylinder to obtain a multi-point air distribution plate image set and a multi-point center cylinder image set; Based on the multi-point air distribution plate image set, coking aggregation anomaly recognition is performed to obtain an air distribution plate anomaly recognition result; Based on the multi-point center cylinder image set, deformation anomaly recognition is performed to obtain a center cylinder deformation anomaly recognition result; The historical combustion parameters and the historical NOx concentration at the furnace outlet of the boiler are obtained, and the air distribution plate anomaly recognition result and the center cylinder deformation anomaly recognition result are combined to determine the first denitration optimization scheme with the first denitration target as a constraint.

4. The method of claim 3, wherein the method further comprises: After the CFB boiler of the target thermal power generating unit is shut down, dispersed multi-point image acquisition is performed on the air distribution plate and the center cylinder to obtain a multi-point air distribution plate image set and a multi-point center cylinder image set, including: The number of dispersed acquisition points of the air distribution plate is determined based on the size of the area of the air distribution plate; Random point layout is performed based on the number of dispersed acquisition points of the air distribution plate to obtain a dispersed acquisition point set of the air distribution plate, wherein the interval between any two dispersed acquisition points of the air distribution plate in the dispersed acquisition point set of the air distribution plate is greater than or equal to a preset interval threshold; The dispersed acquisition point set of the center cylinder is determined based on the size of the diameter of the center cylinder and a preset number of acquisition points of the center cylinder; Dispersed multi-point image acquisition is performed based on the dispersed acquisition point set of the air distribution plate and the dispersed acquisition point set of the center cylinder to obtain the multi-point air distribution plate image set and the multi-point center cylinder image set.

5. The method of claim 3, wherein the method further comprises: Based on the multi-point air distribution plate image set, coking aggregation anomaly recognition is performed to obtain an air distribution plate anomaly recognition result, including: An initial coking aggregation anomaly recognizer is called to perform coking aggregation anomaly recognition on the multi-point air distribution plate image set to obtain a multi-point initial air distribution plate anomaly recognition result set; The multi-point initial air distribution plate anomaly recognition result set is aggregated to obtain a plurality of aggregated multi-point initial air distribution plate anomaly recognition result sets; The plurality of aggregated multi-point initial air distribution plate anomaly recognition result sets are subjected to mean value calculation to obtain a plurality of aggregated multi-point initial air distribution plate anomaly recognition result means; Based on the plurality of aggregated multi-point initial air distribution plate anomaly recognition result means, weighted analysis is performed to obtain an air distribution plate anomaly recognition result.

6. The method of claim 1, wherein the method is characterized by, The plurality of flue gas simulation data sets are traversed to perform working condition aggregation filtering analysis to determine a plurality of aggregated filtering flue gas simulation data, including: The plurality of flue gas simulation data sets are traversed to perform working condition recognition, and the same working condition is divided according to the recognition result to obtain a plurality of aggregated flue gas simulation data sets; A plurality of aggregated filtering analysis centers are extracted from the plurality of aggregated flue gas simulation data sets, wherein each aggregated filtering analysis center is a mean value of an aggregated flue gas simulation data set; According to a preset aggregation filtering radius, an initial neighborhood of the plurality of aggregated filtering analysis centers is determined to obtain a plurality of initial aggregated filtering analysis center neighborhoods; According to the preset aggregation filtering radius, the plurality of initial aggregated filtering analysis center neighborhoods are subjected to boundary expansion recognition in the plurality of aggregated flue gas simulation data sets to determine a plurality of expanded aggregated filtering analysis center neighborhoods; The plurality of expanded aggregated filtering analysis center neighborhoods are subjected to mean value calculation to determine the plurality of aggregated filtering flue gas simulation data.

7. The method of claim 6, wherein the method is characterized by, According to the preset aggregation filtering radius, the plurality of initial aggregated filtering analysis center neighborhoods are subjected to boundary expansion recognition in the plurality of aggregated flue gas simulation data sets to determine a plurality of expanded aggregated filtering analysis center neighborhoods, including: According to the preset aggregation filtering radius, the plurality of initial aggregated filtering analysis center neighborhoods are subjected to edge expansion to obtain a plurality of stage expanded aggregated filtering analysis center neighborhoods; It is judged whether the data quantity of the plurality of stage expanded aggregated filtering analysis center neighborhoods is greater than or equal to the data quantity of the plurality of initial aggregated filtering analysis center neighborhoods. If yes, then the edge expansion is continued according to the preset aggregation filtering radius, until a preset expansion stop condition is met. The neighborhood boundary of the last edge expansion is taken as the expanded boundary, and the plurality of stage expanded aggregated filtering analysis center neighborhoods obtained by the last expansion are taken as the plurality of expanded aggregated filtering analysis center neighborhoods. The preset expansion stop condition is that the expansion times satisfy a preset expansion times threshold.

8. The method of claim 1, wherein the method is characterized by, Based on the plurality of aggregated filtering flue gas simulation data and a second denitration target, a structure optimization of an SNCR denitration system of the target thermal power generating unit is performed to determine a target structure optimization scheme, including: The SNCR denitration system is subjected to structure optimization according to the plurality of aggregated filtering flue gas simulation data and the second denitration target to obtain a plurality of initial structure optimization schemes; obtaining a plurality of denitration results by traversing the plurality of initial structure optimization schemes to simulate the denitration results of the plurality of aggregated screening flue gas simulation data; determining whether there is a denitration result meeting the second denitration target in the plurality of denitration results, and if so, taking the initial structure optimization scheme corresponding to the maximum value in the plurality of denitration results as the target structure optimization scheme.

9. The method of claim 8, wherein the method further comprises: If there is no denitration result meeting the second denitration target in the plurality of denitration results, the plurality of initial structure optimization schemes are randomly adjusted according to a preset adjustment step to obtain a plurality of adjusted structure optimization schemes; determining the target structure optimization scheme based on the plurality of adjusted structure optimization schemes again.

10. A low-load operation condition ultra-low emission adaptive system for a thermal power generating unit, characterized in that, The system comprises: The denitration optimization module is configured to optimize the CFB boiler of the target thermal power generating unit from four dimensions of air distribution device, separator inlet flue, center cylinder and boiler combustion based on a first denitration target, and determine a first denitration optimization scheme. The multi-working condition simulation module is configured to perform flue gas flow multi-working condition simulation in the CFB boiler based on the first denitration optimization scheme, and obtain a plurality of flue gas simulation data sets. The working condition aggregation screening module is configured to perform working condition aggregation screening analysis by traversing the plurality of flue gas simulation data sets, and determine a plurality of aggregated screening flue gas simulation data. The structure optimization module is configured to perform structure optimization on the SNCR denitration system of the target thermal power generating unit based on the plurality of aggregated screening flue gas simulation data and a second denitration target, and determine a target structure optimization scheme, wherein the target structure optimization scheme comprises the number of spray guns, the arrangement position of the spray guns and the initial conveying parameters of the ammonia solution. The adaptive denitration control module is configured to perform ultra-low emission adaptive denitration control on the target thermal power generating unit based on the first denitration optimization scheme and the target structure optimization scheme.

Images