A zoned precise ammonia injection system based on big data model

Abstract

The present invention discloses a partitioned precise ammonia injection system based on a big data model, and belongs to the technical field of environmental protection and energy conservation of coal-fired power generation. The present invention solves the problem of inaccurate and unstable operation of the existing denitrification system, by establishing a steady-state precise ammonia injection data model under different working conditions, and performing feedback adjustment of precise ammonia injection, the adjusted precise ammonia injection data is recorded, collected and forms a new steady-state data model; a discrete limiting coefficient is introduced, and when the opening of the ammonia injection regulating valve is away from the big data model and reaches the value specified by the discrete limiting coefficient, the big data model that meets the current operating conditions is re-called; the influence of the linear adjustment error of the traditional partitioned ammonia injection and the influence of the hysteresis of the denitrification measuring instrument are effectively reduced, and the response speed and accuracy of the precise ammonia injection system are improved under the premise of ensuring the denitrification efficiency and reducing ammonia escape. The reasonable use of the big data model can still ensure the stable and efficient operation of the denitrification system under the condition of large fluctuations in working conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental protection and energy conservation in coal-fired power generation, and in particular to a zoned precise ammonia injection system based on a big data model. Background Art

[0002] The nitrogen oxides contained in the flue gas emissions of coal-fired boilers are an important component of air pollution. Controlling the nitrogen oxide (NOx) content in the flue gas emissions of coal-fired boilers is the focus of my country's environmental protection regulations. Selective catalytic reduction denitrification technology (SCR) has become the mainstream flue gas denitrification process for large domestic thermal power plants due to its high denitrification efficiency. However, in actual use, the ammonia concentration in the denitrification system is unevenly distributed, resulting in an increasingly serious problem of uneven distribution of outlet NOx. In order to ensure that the outlet NOx of denitrification meets environmental protection indicators, the impact of ammonia slip on the entire denitrification system and even the boiler system is often ignored. Excessive ammonia slip not only affects the catalyst life and flue gas system resistance, but also has a great impact on the unit economy. It has become a problem that needs to be solved urgently for the SCR denitrification system.

[0003] At present, the problem of ammonia escape and uneven outlet NOx concentration field is mainly solved by precise ammonia injection; the concentration of outlet ammonia escape is controlled by the method of zone detection and zone ammonia injection adjustment. This mode completely relies on the feedback of the detection instrument to realize the adjustment function, which will be affected by the hysteresis of the detection and produce adjustment deviation. There are many variabilities in the load and coal of domestic coal-fired units. The use of only traditional precise ammonia injection technology has unstable limitations, which is not conducive to the long-term guarantee of efficient and economic operation of denitrification units. Therefore, in order to improve the accuracy and stability of the operation of the denitrification system, it is urgently necessary to optimize the existing measurement methods, control methods, data utilization, etc. Summary of the invention

[0004] The purpose of the present invention is to provide a partitioned precise ammonia injection system based on a big data model. By establishing a steady-state precise ammonia injection data model under different operating conditions and introducing a discrete limiting coefficient, when the opening of the ammonia injection regulating valve is away from the big data model and reaches the value specified by the discrete limiting coefficient, the big data model that meets the current operating conditions is re-called; effectively reducing the influence of the linear adjustment error of traditional partitioned ammonia injection and the influence of the hysteresis of the denitrification measuring instrument, while ensuring the denitrification efficiency and reducing ammonia escape, the response speed and accuracy of the precise ammonia injection system are improved, and the problems raised in the above-mentioned background technology are solved.

[0005] To achieve the above object, the present invention provides the following technical solution: a partitioned precise ammonia injection system based on a big data model, comprising:

[0006] Model data collection and screening module for

[0007] Obtain the boiler load, coal type, flue gas flow, inlet NOx concentration, outlet zone NOx concentration, outlet zone NH3 escape concentration, and zone ammonia injection regulating valve opening data generated in each area, and filter each data to retain valid data under steady-state conditions;

[0008] Data modeling module for

[0009] Under different boiler load and coal conditions, according to the flue gas flow in each area and the NOx concentration range in different positions, the regulating valve opening model under different working conditions is established;

[0010] Intelligent computing module for

[0011] Introduce a discrete limit coefficient, and set the value specified by the discrete limit coefficient for the opening model of each regional control valve;

[0012] Opening value comparison module, used for

[0013] Compare the opening data of the zoned ammonia injection regulating valve with the value specified by the discrete limit coefficient of the big data model, and feed back the comparison result to the model calling module;

[0014] Model call module, used for

[0015] Read the operating data of each area, and according to different working conditions and the comparison results of discrete limit coefficients, call the big data model corresponding to the current working conditions to control the emission values ​​of denitrification NOx and ammonia slip;

[0016] Precision ammonia injection module for

[0017] Accurate ammonia injection is achieved by combining denitrification tail zone detection feedback with denitrification inlet zone adjustment ammonia injection.

[0018] Furthermore, the model data collection and screening module includes:

[0019] Data collection unit for

[0020] Collect multiple groups of historical production and operation processes in each area, including boiler load, coal type, flue gas flow, inlet NOx concentration, outlet zone NOx concentration, outlet zone NH3 escape concentration, and zone ammonia injection regulating valve opening;

[0021] Data acquisition unit for

[0022] Real-time collection of boiler load, coal type, flue gas flow, inlet NOx concentration, outlet zone NOx concentration, outlet zone NH 3 Fugitive concentration and zoned ammonia injection regulating valve opening data;

[0023] Data screening unit for

[0024] It is used to organize and screen multiple groups of historical operating data, delete the data under unstable conditions, retain the data under steady conditions, and transmit the retained data to the data modeling module for modeling processing.

[0025] Furthermore, the data modeling module includes:

[0026] Data receiving unit, used for

[0027] Receiving multiple groups of data under steady-state conditions, and classifying and storing the multiple groups of data to obtain a modeling database;

[0028] Data analysis unit for

[0029] Analyze multiple groups of data to obtain the production area and operating conditions represented by each group of data, and mark the production area and operating conditions of each group of data;

[0030] Model creation unit for

[0031] Different big data models are established according to the production area and operating conditions represented by each set of data.

[0032] Furthermore, the data analysis unit includes:

[0033] Data partitioning unit, used for

[0034] Each group of data is divided into training set and test set. The training set of each group of data is imported into the corresponding big data model for training. After the characteristic points of this type of big data model are determined, the test set of each group of data is imported into the corresponding big data model for testing to optimize the accuracy of each big data model.

[0035] Furthermore, the opening value comparison module compares the opening data of the partitioned ammonia injection regulating valve with the value specified by the discrete limit coefficient of the big data model, specifically:

[0036] The partition ammonia injection regulating valve opening data is collected in real time through the model data collection and screening module. When the partition ammonia injection regulating valve opening data is close to the big data model and reaches the value specified by the discrete coefficient limit number, there is no need to adjust the big data model; when the partition ammonia injection regulating valve opening data is far away from the big data model and reaches the value specified by the discrete coefficient limit number, it is necessary to re-call the big data model that meets the current operating conditions and then perform precise adjustment of the partition regulating valve.

[0037] Furthermore, the model calling module includes:

[0038] Keyword building unit for

[0039] Establish model call search keywords according to the names of each area and each data, and enter the search keywords according to the calling interface to optimize the model call efficiency.

[0040] Furthermore, the precise ammonia injection module comprises:

[0041] Data feedback unit for

[0042] After calling the corresponding steady-state model to control the emission values ​​of denitrification NOx and ammonia escape, the various data collected in real time by the model data acquisition and screening module are fed back to the data modeling module to optimize the operating data in the big data model.

[0043] Furthermore, the precise ammonia injection module adopts a method of combining denitrification tail zone detection feedback with denitrification inlet zone adjustment ammonia injection, specifically:

[0044] By setting up zoned automatic regulating valves and remote flow meters on the branch pipes of the ammonia injection grid, the cross-section of the denitrification inlet flue is divided into zones, and the zones correspond to the measurements of the denitrification tail zones. A zoned static mixer is set up in the inlet flue to make the concentration field of ammonia evenly distributed in the zoned area. The zoned regulating valve is then adjusted to level the ammonia-nitrogen molar ratio between the zones.

[0045] Furthermore, the denitrification tail zone detection is to detect NOx and NH3 in the denitrification tail zone using an in-situ extraction detector.

[0046] Furthermore, when the intelligent computing module introduces the discrete limiting coefficient, during the precise ammonia injection commissioning or the initial operation, a relatively large value of the discrete limiting coefficient is taken; after the big data model is established and improved, the discrete limiting coefficient is relatively lowered.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The present invention establishes a steady-state precise ammonia injection data model under different working conditions, performs feedback regulation of precise ammonia injection, and the adjusted precise ammonia injection data is recorded, collected, and forms a new steady-state data model; introduces a discrete limiting coefficient, and when the ammonia injection regulating valve opening is away from the big data model and reaches the value specified by the discrete limiting coefficient, the big data model that meets the current operating conditions is called again; effectively reduces the influence of the linear regulation error of traditional partition ammonia injection and the influence of the hysteresis of the denitration measuring instrument, and improves the response speed and accuracy of the precise ammonia injection system while ensuring the denitration efficiency and reducing ammonia escape. Reasonable use of the big data model can still ensure the stable and efficient operation of the system under the condition of large fluctuations in working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a logic diagram of the intelligent and precise ammonia injection control system based on the big data model of the present invention;

[0050] Figure 2 This is a composition diagram of the intelligent and precise ammonia injection control system based on the big data model of the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] In order to solve the technical problems that the existing mode completely relies on the feedback of detection instruments to realize the regulation function, which will be affected by the hysteresis of detection and produce regulation deviation; and the load and coal of domestic coal-fired units are variable, and only precise ammonia injection has unstable limitations, which is not conducive to the long-term guarantee of efficient and economic operation of denitrification units. Please refer to Figure 1-Figure 2 This embodiment provides the following technical solution

[0053] A partitioned precise ammonia injection system based on a big data model, comprising:

[0054] Model data collection and screening module for

[0055] The boiler load, coal type, flue gas flow, inlet NOx concentration, outlet zone NOx concentration, outlet zone NH3 escape concentration, and zone ammonia injection regulating valve opening data generated in each area are obtained, and the data are screened to retain the valid data under steady-state conditions. Specifically, by collecting various data in each area of ​​the production environment, a big data basis is provided for the establishment of a big data model. At the same time, different data samples can effectively reflect the operating conditions of the environment, so as to optimize the diversity of the simulation conditions of the big data model.

[0056] Model data collection and screening module, including:

[0057] Data collection unit for

[0058] Collect multiple sets of historical production and operation data of each area, including boiler load, coal type, flue gas flow, inlet NOx concentration, outlet zone NOx concentration, outlet zone NH3 escape concentration and zone ammonia injection regulating valve opening. Specifically, collect five or more sets of historical production and operation data to ensure that multiple sets of data include different operating conditions, thereby enriching the types of big data models so that the required big data models can be called later.

[0059] Data acquisition unit for

[0060] Real-time data collection of boiler load, coal type, flue gas flow, inlet NOx concentration, outlet zone NOx concentration, outlet zone NH3 escape concentration and zone ammonia injection regulating valve opening in each area during the current production operation; specifically, through real-time collection of various data under the current operating conditions, the client can understand the production status of each area in real time and quickly grasp the emissions of denitrification NOx and ammonia escape in each area; at the same time, after calling the required big data model according to the current operating conditions in any area to control the emission values ​​of denitrification NOx and ammonia escape, the data collection unit is used to collect various data in the area in real time, and the data collected after the call control is fed back to the data modeling module as soon as possible, so as to continuously accumulate and screen the operating data during the operation, and intelligently expand the steady-state big data model under different conditions, effectively improving the accuracy of the big data model.

[0061] Data screening unit for

[0062] It is used to organize and screen multiple groups of historical operating data, delete the data under unstable conditions, retain the data under steady conditions, and transmit the retained data to the data modeling module for modeling processing; specifically, by retaining the data under steady conditions, the multiple big data models established are all models of normal operation under different operating conditions, thereby ensuring that the NOx concentration under different operating conditions output by each big data model has a high reference value.

[0063] Data modeling module for

[0064] Under different boiler load and coal conditions, according to the flue gas flow in each area and the NOx concentration range in different positions, the regulating valve opening model under different working conditions is established;

[0065] Data modeling module, including:

[0066] Data receiving unit, used for

[0067] Receive multiple groups of data under steady-state conditions, and classify and store the multiple groups of data to obtain a modeling database; specifically, after receiving multiple groups of historical data, in the process of daily production operation, use the data acquisition unit to collect data from various areas in real time, and filter out the data under steady-state conditions for storage, thereby enriching the modeling database and improving the reference basis of the big data model.

[0068] Data analysis unit for

[0069] Analyze multiple groups of data to obtain the production area and operating conditions represented by each group of data, and mark the production area and operating conditions of each group of data; specifically, by analyzing the production area and operating conditions to which each data belongs, when storing each data, a corresponding folder can be created, and the belonging information and collection time can be marked in the folder, so that the client can quickly retrieve the required data in the modeling database, thereby improving data utilization.

[0070] Data Analysis Unit, including:

[0071] Data partitioning unit, used for

[0072] Divide each group of data into training sets and test sets, import the training set of each group of data into the corresponding big data model for training, and after determining the characteristic points of the big data model of this type, import the test set of each group of data into the corresponding big data model for testing, so as to optimize the accuracy of each big data model;

[0073] Model creation unit for

[0074] Different big data models are established according to the production areas and operating conditions represented by each set of data, so as to optimize the types of big data models, improve the diversity of big data models, and ensure the effectiveness of subsequent calls.

[0075] Intelligent computing module for

[0076] A discrete limiting coefficient is introduced, and the value specified by the discrete limiting coefficient is set for the opening model of each regional control valve; specifically, by setting the discrete limiting coefficient, the stable operation of the dilution fan is guaranteed, and the adverse effect of fluctuations in the opening of the zone control valve on the dilution air flow rate is avoided; during the precise ammonia injection commissioning or the initial operation, the discrete limiting coefficient takes a relatively large value, which is conducive to the establishment of the big data model; after the big data model is established and improved, the discrete limiting coefficient can be relatively lowered, which is conducive to the call of the big data steady-state model.

[0077] In this embodiment: During the initial commissioning and operation of the denitrification system, operation data are collected to establish a big data steady-state model, the discrete limit coefficient is increased, the precise ammonia injection zone regulating valve is differentially feedback-adjusted according to the outlet zone NOx and NH3 detection values, the zone regulating valve opening values ​​are collected, and a steady-state model is established through an intelligent algorithm; under the condition of using a certain type of coal, the boiler load is adjusted, for example: 100% rated load, 90% rated load, 80% rated load, ..., 50% rated load, 40% rated load, 30% rated load, to ensure steady-state operation for a period of time under different boiler loads in order to determine a big data steady-state model under typical operating conditions of a commonly used type of coal.

[0078] As the denitrification system continues to operate, the operating data continues to accumulate, and then through the big data model intelligent algorithm, a big data steady-state model of different flue gas flow ranges and NOx concentration ranges is formed.

[0079] In the later stage of the operation of the denitrification system, the discrete limit coefficient can be reduced, and the big data steady-state model can be used; when the boiler load is adjusted, the typical stable model of the boiler under this load is called, and under this typical stable model, the zone regulating valve is adjusted through the precise ammonia injection module; when the discrete value reaches the established discrete limit coefficient, the big data steady-state model under this boiler load is called according to the flue gas flow and NOx concentration.

[0080] Opening value comparison module, used for

[0081] The opening data of the partitioned ammonia injection regulating valve is compared with the value specified by the discrete limit coefficient of the big data model, and the comparison result is fed back to the model calling module; specifically:

[0082] The partition ammonia injection regulating valve opening data is collected in real time through the model data collection and screening module. When the partition ammonia injection regulating valve opening data is close to the big data model and reaches the value specified by the discrete coefficient limit number, there is no need to adjust the big data model; when the partition ammonia injection regulating valve opening data is far away from the big data model and reaches the value specified by the discrete coefficient limit number, it is necessary to re-call the big data model that meets the current operating conditions, and then accurately adjust the partition regulating valve;

[0083] Model call module, used for

[0084] Read the operating data of each area, and according to different working conditions and the comparison results of discrete limit coefficients, call the big data model corresponding to the current working conditions to control the emission values ​​of denitrification NOx and ammonia slip;

[0085] In this embodiment: for different boiler loads and coal types, a regulating valve opening model under different working conditions is established within the range of inlet NOx concentration and flue gas flow. For example:

[0086] Coal type 1:

[0087] When the boiler load is 100%: 1100000Nm3 / h<flue gas flow ≤1200000Nm3 / h, NOx concentration>400mg / Nm3;

[0088] When the boiler load is 100%: 1100000Nm3 / h<flue gas flow ≤1200000Nm3 / h, NOx concentration ≤400mg / Nm3;

[0089] When the boiler load is 100%: 1000000Nm3 / h<flue gas flow ≤1100000Nm3 / h, NOx concentration>400mg / Nm3;

[0090] When the boiler load is 100%: 1000000Nm3 / h<flue gas flow ≤1100000Nm3 / h, NOx concentration ≤400mg / Nm3;

[0091] Through the above settings, the NOx concentration values ​​under different operating conditions output by each big data model can be set as the NOx concentration comparison thresholds under different operating conditions; for example: in the first case mentioned above, the big data model outputs a NOx concentration of >400mg / Nm3. If the NOx concentration of 350mg / Nm3 under the operating condition is collected in real time through the data acquisition unit, it is compared with the threshold value under the operating condition, and it is concluded that the NOx concentration in the area under the current operating condition is normal, and there is no need to adjust the ammonia injection scheme; if the NOx concentration of 450mg / Nm3 under the operating condition is collected in real time through the data acquisition unit, it is compared with the threshold value under the operating condition, and it is concluded that the NOx concentration in the area under the current operating condition exceeds the threshold value, which is an abnormal state, then it is necessary to call the big data model corresponding to the current operating condition to control the emission values ​​of denitrification NOx and ammonia escape under the current operating condition.

[0092] Model calling module, including:

[0093] Keyword building unit for

[0094] A model is called to retrieve keywords based on the names of each area and each data item, and the retrieval keywords are input according to the calling interface to optimize the model calling efficiency. In this embodiment: the above-mentioned boiler load value, flue gas flow value and area name can be used as retrieval keywords to facilitate the client to perform retrieval based on the above keywords in the future, thereby improving the utilization efficiency of the big data model. The establishment of keywords includes but is not limited to the above categories.

[0095] Precision ammonia injection module for

[0096] Accurate ammonia injection is carried out by combining denitrification tail area detection feedback with denitrification inlet area adjustment ammonia injection, specifically:

[0097] By setting a partition automatic regulating valve and a remote flow meter on the branch pipe of the ammonia spraying grid, the cross section of the denitrification inlet flue is divided into zones, which correspond to the measurement of the denitrification tail zone; a partition static mixer is set in the inlet flue to make the concentration field of ammonia in the partition area uniform; and the ammonia nitrogen molar ratio between the zones is leveled by adjusting the partition regulating valve; wherein, the denitrification tail zone detection is to detect NOx and NH3 in the denitrification tail zone by using an in-situ extraction detector; in this embodiment: the denitrification inlet flue can be divided into zones, and the zone area is 3m2 -6m 2 A zone automatic regulating valve is set on the ammonia spray grid branch pipe of each zone, and a corresponding zone is also set at the tail outlet of the denitrification reactor. Each zone is equipped with an in-situ extraction measuring instrument for NOx and NH3; secondly, the precise ammonia spray zone regulating valve adopts differential feedback adjustment. First, the NOx value of the denitrification outlet zone is averaged, and the difference between the NOx value of all zones and the average value is taken. The opening of the regulating valve in the area with the largest difference is directionally adjusted. After the adjustment is completed, the above process is repeated to finally achieve the leveling purpose.

[0098] Precision ammonia injection module, including:

[0099] Data feedback unit for

[0100] After calling the corresponding steady-state model to control the emission values ​​of denitrification NOx and ammonia escape, the various data collected in real time by the model data acquisition and screening module are fed back to the data modeling module to optimize the operating data in the big data model; with the accumulation of operating data, the flue gas flow and NOx concentration partitions can be further subdivided through intelligent algorithms to form a more accurate big data steady-state model.

[0101] The beneficial effects achieved by the above content are as follows: Through the above method, the influence of the linear adjustment error of the traditional partition ammonia injection and the influence of the hysteresis of the denitration measuring instrument are effectively reduced, and the response speed and accuracy of the precise ammonia injection system are improved under the premise of ensuring the denitration efficiency and reducing ammonia escape. The reasonable use of big data models can still ensure the stable and efficient operation of the system under the condition of large fluctuations in operating conditions.

[0102] Working principle: The denitrification inlet flue is divided into zones, and the ammonia spraying grid main pipe of each zone is provided with a zone automatic regulating valve; corresponding zones are also set at the tail outlet of the denitrification reactor, and each zone is equipped with in-situ extraction measuring instruments for NOx and NH3; a zone static mixer is set at the inlet flue to make the flue gas and ammonia evenly distributed in the zone; the precise ammonia spraying module performs differential adjustment to level the NOx at the outlet of the denitrification reactor; the above operating data are converted into a big data steady-state model through the big data collection and screening module and the data modeling module, and the model is called according to the boiler load, operating parameters and discrete limit parameters; thereby realizing intelligent zone precise ammonia spraying, improving the denitrification efficiency and reducing ammonia escape, while ensuring the accuracy and stability of the denitrification system operation.

[0103] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0104] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A zoned precise ammonia injection system based on a big data model, comprising: Model data collection and screening module for Obtain the boiler load, coal type, flue gas flow, inlet NOx concentration, outlet zone NOx concentration, outlet zone NH3 escape concentration, and zone ammonia injection regulating valve opening data generated in each area, and filter each data to retain valid data under steady-state conditions; Data modeling module for Under different boiler load and coal conditions, according to the flue gas flow in each area and the NOx concentration range in different positions, the regulating valve opening model under different working conditions is established; Intelligent computing module for Introduce a discrete limit coefficient, and set the value specified by the discrete limit coefficient for the opening model of each regional control valve; Opening value comparison module, used for Compare the opening data of the zoned ammonia injection regulating valve with the value specified by the discrete limit coefficient of the big data model, and feed back the comparison result to the model calling module; Model call module, used for Read the operating data of each area, and according to different working conditions and the comparison results of discrete limit coefficients, call the big data model corresponding to the current working conditions to control the emission values ​​of denitrification NOx and ammonia slip; Precision ammonia injection module for Accurate ammonia injection is achieved by combining denitrification tail zone detection feedback with denitrification inlet zone adjustment ammonia injection.

2. According to the big data model-based partition precise ammonia injection system according to claim 1, it is characterized by: The model data collection and screening module includes: Data collection unit for Collect multiple groups of historical production and operation processes in each area, including boiler load, coal type, flue gas flow, inlet NOx concentration, outlet zone NOx concentration, outlet zone NH3 escape concentration, and zone ammonia injection regulating valve opening; Data acquisition unit for Real-time collection of boiler load, coal type, flue gas flow, inlet NOx concentration, outlet zone NOx concentration, outlet zone NH3 escape concentration and zone ammonia injection regulating valve opening data in each area during the current production operation; Data screening unit for It is used to organize and screen multiple groups of historical operating data, delete the data under unstable conditions, retain the data under steady conditions, and transmit the retained data to the data modeling module for modeling processing.

3. According to the big data model-based partition precise ammonia injection system of claim 1, it is characterized by: The data modeling module comprises: Data receiving unit, used for Receiving multiple groups of data under steady-state conditions, and classifying and storing the multiple groups of data to obtain a modeling database; Data analysis unit for Analyze multiple groups of data to obtain the production area and operating conditions represented by each group of data, and mark the production area and operating conditions of each group of data; Model creation unit for Different big data models are established according to the production area and operating conditions represented by each set of data.

4. According to claim 3, a partitioned precise ammonia injection system based on a big data model is characterized by: The data analysis unit comprises: Data partitioning unit, used for Each group of data is divided into training set and test set. The training set of each group of data is imported into the corresponding big data model for training. After the characteristic points of this type of big data model are determined, the test set of each group of data is imported into the corresponding big data model for testing to optimize the accuracy of each big data model.

5. According to claim 1, a partitioned precise ammonia injection system based on a big data model is characterized by: The opening value comparison module compares the opening data of the partitioned ammonia injection regulating valve with the value specified by the discrete limit coefficient of the big data model, specifically: The partition ammonia injection regulating valve opening data is collected in real time through the model data collection and screening module. When the partition ammonia injection regulating valve opening data is close to the big data model and reaches the value specified by the discrete coefficient limit number, there is no need to adjust the big data model; when the partition ammonia injection regulating valve opening data is far away from the big data model and reaches the value specified by the discrete coefficient limit number, it is necessary to re-call the big data model that meets the current operating conditions and then perform precise adjustment of the partition regulating valve.

6. According to the big data model-based partition precise ammonia injection system of claim 1, it is characterized by: The model calling module comprises: Keyword building unit for Establish model call search keywords according to the names of each area and each data, and enter the search keywords according to the calling interface to optimize the model call efficiency.

7. According to claim 1, a partitioned precise ammonia injection system based on a big data model is characterized by: The precise ammonia injection module comprises: Data feedback unit for After calling the corresponding steady-state model to control the emission values ​​of denitrification NOx and ammonia escape, the various data collected in real time by the model data acquisition and screening module are fed back to the data modeling module to optimize the operating data in the big data model.

8. The partition-based precise ammonia injection system based on a big data model according to claim 1 is characterized by: The precise ammonia injection module adopts a method of combining denitration tail zone detection feedback with denitration inlet zone adjustment ammonia injection, specifically: By setting up zoned automatic regulating valves and remote flow meters on the branch pipes of the ammonia injection grid, the cross-section of the denitrification inlet flue is divided into zones, and the zones correspond to the measurements of the denitrification tail zones. A zoned static mixer is set up in the inlet flue to make the concentration field of ammonia evenly distributed in the zoned area. The zoned regulating valve is then adjusted to level the ammonia-nitrogen molar ratio between the zones.

9. The partition-based precise ammonia injection system based on a big data model according to claim 8 is characterized by: The denitrification tail zone detection is to detect NOx and NH3 in the denitrification tail zone using an in-situ extraction detector.

10. The partition-based precise ammonia injection system based on a big data model according to claim 1, characterized in that: When the intelligent calculation module introduces the discrete limit coefficient, during the precise ammonia injection commissioning or the initial operation, a larger value of the discrete limit coefficient is taken; after the big data model is established and improved, the discrete limit coefficient is relatively lowered.

Images