CN116983819A - Flue gas desulfurization washing tower and method thereof - Google Patents
Flue gas desulfurization washing tower and method thereof Download PDFInfo
- Publication number
- CN116983819A CN116983819A CN202311008799.4A CN202311008799A CN116983819A CN 116983819 A CN116983819 A CN 116983819A CN 202311008799 A CN202311008799 A CN 202311008799A CN 116983819 A CN116983819 A CN 116983819A
- Authority
- CN
- China
- Prior art keywords
- time sequence
- ship
- training
- flue gas
- flow
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 title claims abstract description 124
- 239000003546 flue gas Substances 0.000 title claims abstract description 124
- 238000006477 desulfuration reaction Methods 0.000 title claims abstract description 67
- 230000023556 desulfurization Effects 0.000 title claims abstract description 67
- 238000005406 washing Methods 0.000 title abstract description 40
- 238000000034 method Methods 0.000 title abstract description 26
- 238000005507 spraying Methods 0.000 claims abstract description 63
- 230000008859 change Effects 0.000 claims abstract description 29
- 238000012300 Sequence Analysis Methods 0.000 claims abstract description 8
- 239000000779 smoke Substances 0.000 claims description 208
- 239000013598 vector Substances 0.000 claims description 141
- 238000012549 training Methods 0.000 claims description 79
- 238000009825 accumulation Methods 0.000 claims description 54
- 239000007921 spray Substances 0.000 claims description 53
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 28
- 230000001052 transient effect Effects 0.000 claims description 21
- 238000013527 convolutional neural network Methods 0.000 claims description 18
- 238000004458 analytical method Methods 0.000 claims description 15
- 230000003247 decreasing effect Effects 0.000 claims description 13
- 238000000605 extraction Methods 0.000 claims description 11
- 238000003062 neural network model Methods 0.000 claims description 9
- 238000010219 correlation analysis Methods 0.000 claims description 8
- 230000004927 fusion Effects 0.000 claims description 7
- 238000005201 scrubbing Methods 0.000 claims description 7
- 230000001186 cumulative effect Effects 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 3
- 230000002123 temporal effect Effects 0.000 claims 1
- 230000003009 desulfurizing effect Effects 0.000 abstract description 24
- 239000003795 chemical substances by application Substances 0.000 abstract description 16
- 238000012545 processing Methods 0.000 abstract description 9
- 239000007788 liquid Substances 0.000 description 26
- 230000006870 function Effects 0.000 description 17
- 230000000694 effects Effects 0.000 description 15
- RAHZWNYVWXNFOC-UHFFFAOYSA-N Sulphur dioxide Chemical compound O=S=O RAHZWNYVWXNFOC-UHFFFAOYSA-N 0.000 description 14
- 230000008569 process Effects 0.000 description 11
- 238000010586 diagram Methods 0.000 description 10
- 238000009826 distribution Methods 0.000 description 8
- 238000011176 pooling Methods 0.000 description 7
- 230000000737 periodic effect Effects 0.000 description 6
- 230000002159 abnormal effect Effects 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 5
- 238000012856 packing Methods 0.000 description 5
- 230000004913 activation Effects 0.000 description 4
- 210000002569 neuron Anatomy 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 239000012530 fluid Substances 0.000 description 3
- 239000007789 gas Substances 0.000 description 3
- 239000002245 particle Substances 0.000 description 3
- QAOWNCQODCNURD-UHFFFAOYSA-L Sulfate Chemical compound [O-]S([O-])(=O)=O QAOWNCQODCNURD-UHFFFAOYSA-L 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 230000014509 gene expression Effects 0.000 description 2
- 238000012417 linear regression Methods 0.000 description 2
- 238000005086 pumping Methods 0.000 description 2
- 238000000746 purification Methods 0.000 description 2
- 238000004064 recycling Methods 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 238000011282 treatment Methods 0.000 description 2
- VHUUQVKOLVNVRT-UHFFFAOYSA-N Ammonium hydroxide Chemical compound [NH4+].[OH-] VHUUQVKOLVNVRT-UHFFFAOYSA-N 0.000 description 1
- 241000288105 Grus Species 0.000 description 1
- 235000019738 Limestone Nutrition 0.000 description 1
- 238000003723 Smelting Methods 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 239000002253 acid Substances 0.000 description 1
- 238000003916 acid precipitation Methods 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 239000003513 alkali Substances 0.000 description 1
- 235000011114 ammonium hydroxide Nutrition 0.000 description 1
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 238000005311 autocorrelation function Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 239000011152 fibreglass Substances 0.000 description 1
- 239000000945 filler Substances 0.000 description 1
- 238000011049 filling Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 239000002608 ionic liquid Substances 0.000 description 1
- 239000006028 limestone Substances 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000011328 necessary treatment Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 239000002002 slurry Substances 0.000 description 1
- 229910001220 stainless steel Inorganic materials 0.000 description 1
- 239000010935 stainless steel Substances 0.000 description 1
- 238000003756 stirring Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 239000000725 suspension Substances 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
- B01D—SEPARATION
- B01D53/00—Separation of gases or vapours; Recovering vapours of volatile solvents from gases; Chemical or biological purification of waste gases, e.g. engine exhaust gases, smoke, fumes, flue gases, aerosols
- B01D53/34—Chemical or biological purification of waste gases
- B01D53/74—General processes for purification of waste gases; Apparatus or devices specially adapted therefor
- B01D53/77—Liquid phase processes
- B01D53/78—Liquid phase processes with gas-liquid contact
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
- B01D—SEPARATION
- B01D53/00—Separation of gases or vapours; Recovering vapours of volatile solvents from gases; Chemical or biological purification of waste gases, e.g. engine exhaust gases, smoke, fumes, flue gases, aerosols
- B01D53/34—Chemical or biological purification of waste gases
- B01D53/46—Removing components of defined structure
- B01D53/48—Sulfur compounds
- B01D53/50—Sulfur oxides
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
- B01D—SEPARATION
- B01D53/00—Separation of gases or vapours; Recovering vapours of volatile solvents from gases; Chemical or biological purification of waste gases, e.g. engine exhaust gases, smoke, fumes, flue gases, aerosols
- B01D53/34—Chemical or biological purification of waste gases
- B01D53/74—General processes for purification of waste gases; Apparatus or devices specially adapted therefor
- B01D53/77—Liquid phase processes
- B01D53/79—Injecting reactants
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
- B01D—SEPARATION
- B01D2258/00—Sources of waste gases
- B01D2258/02—Other waste gases
- B01D2258/0283—Flue gases
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Environmental & Geological Engineering (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Analytical Chemistry (AREA)
- General Chemical & Material Sciences (AREA)
- Oil, Petroleum & Natural Gas (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Treating Waste Gases (AREA)
Abstract
The application discloses a flue gas desulfurization washing tower and a method thereof, wherein a sensor is used for collecting ship flue gas flow values at a plurality of preset time points, and a data processing and analyzing algorithm is introduced at the rear end to perform time sequence analysis of the ship flue gas flow so as to judge the time sequence change trend of the ship flue gas flow, so that the self-adaptive adjustment of the spraying quantity is performed in real time, and the effective utilization of a desulfurizing agent and the maximization of the desulfurizing effect are ensured.
Description
Technical Field
The application relates to the field of intelligent control, and more particularly relates to a flue gas desulfurization washing tower and a method thereof.
Background
At present, the total emission amount of sulfur dioxide in China is the first place in the world, and if the sulfur dioxide in the flue gas is directly emitted in the air, acid rain can be formed, so that the environment is seriously affected. Thus, flue gas desulfurization becomes a necessary treatment process. The more advanced technology of smelting flue gas desulfurization treatment is an ionic liquid desulfurization technology. However, the current flue gas desulfurization washing tower can only carry out one-time spray purification on flue gas, and the spray purification mode is difficult to remove harmful substance particles in the flue gas, so that a small amount of harmful substance particles still exist in the washed flue gas. In addition, the washing tower is easy to block the spraying system after being washed by circulating water, and once the washing tower is blocked, the washing tower is required to be stopped for cleaning and overhauling, so that economic loss is caused.
In the desulfurization washing tower, it is very important to adjust the spray amount. Excessive spraying amount can cause excessive consumption of desulfurizing agent, increasing cost. Too small a spray amount does not cover the entire flue gas flow, reducing the desulfurization effect. However, the conventional scheme sets a fixed spraying amount in the whole spraying process based on manual experience, and does not pay attention to the suitability with the flue gas flow, so that the excessive consumption of the desulfurizing agent and the low desulfurizing effect are caused, and the practical application requirements are difficult to meet.
Accordingly, an optimized flue gas desulfurization scrubber is desired.
Disclosure of Invention
The present application has been made to solve the above-mentioned technical problems. The embodiment of the application provides a flue gas desulfurization washing tower and a method thereof, wherein a sensor is used for collecting ship flue gas flow values at a plurality of preset time points, and a data processing and analyzing algorithm is introduced into the rear end to perform time sequence analysis of ship flue gas flow so as to judge the time sequence change trend of the ship flue gas flow, so that the self-adaptive adjustment of spraying quantity is performed in real time, and the effective utilization of a desulfurizing agent and the maximization of a desulfurizing effect are ensured.
According to one aspect of the present application, there is provided a flue gas desulfurization scrubber comprising:
A circulation water tank;
the tower body is provided with a flue gas outlet formed at the upper part of the tower body and a flue gas inlet formed at the lower part of the tower body;
a main spray pipe communicated with the circulating water tank and extending into the tower body;
the circulating water pump is arranged on the main spraying pipeline;
the spraying branch pipe is connected with the main spraying pipe;
the spray gun is arranged at the head part of the spray branch pipe; and
and a controller for controlling the spraying amount of the spray gun.
According to another aspect of the present application, there is provided a flue gas desulfurization scrubbing method comprising:
acquiring flow values of ship smoke at a plurality of preset time points in a preset time period;
the smoke flow time sequence analysis module is used for performing time sequence correlation analysis on the flow values of the ship smoke at the plurality of preset time points to obtain ship smoke flow time sequence characteristics; and
and the spraying speed control module is used for determining that the spraying speed at the current time point should be increased or decreased based on the ship smoke flow time sequence characteristics.
Compared with the prior art, the flue gas desulfurization washing tower and the method thereof provided by the application have the advantages that the sensor is used for collecting the ship flue gas flow values at a plurality of preset time points, and the data processing and analysis algorithm is introduced at the rear end to perform time sequence analysis of the ship flue gas flow so as to judge the time sequence change trend of the ship flue gas flow, so that the self-adaptive adjustment of the spraying quantity is performed in real time, and the effective utilization of the desulfurizing agent and the maximization of the desulfurizing effect are ensured.
Drawings
The above and other objects, features and advantages of the present application will become more apparent by describing embodiments of the present application in more detail with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of embodiments of the application and are incorporated in and constitute a part of this specification, illustrate the application and together with the embodiments of the application, and not constitute a limitation to the application. In the drawings, like reference numerals generally refer to like parts or steps.
FIG. 1 is a block diagram of a flue gas desulfurization scrubber in accordance with an embodiment of the present application;
FIG. 2 is a block diagram of a controller in a flue gas desulfurization scrubber in accordance with an embodiment of the present application;
FIG. 3 is a system architecture diagram of a flue gas desulfurization scrubber in accordance with an embodiment of the present application;
FIG. 4 is a block diagram of a training phase of a flue gas desulfurization scrubber in accordance with an embodiment of the present application;
FIG. 5 is a block diagram of a flue gas flow timing analysis module in a flue gas desulfurization scrubber according to an embodiment of the present application;
FIG. 6 is a flow chart of a flue gas desulfurization scrubbing method according to an embodiment of the present application;
Detailed Description
Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
As used in the specification and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Although the present application makes various references to certain modules in a system according to embodiments of the present application, any number of different modules may be used and run on a user terminal and/or server. The modules are merely illustrative, and different aspects of the systems and methods may use different modules.
A flowchart is used in the present application to describe the operations performed by a system according to embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in order precisely. Rather, the various steps may be processed in reverse order or simultaneously, as desired. Also, other operations may be added to or removed from these processes.
Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
In the desulfurization washing tower, it is very important to adjust the spray amount. Excessive spraying amount can cause excessive consumption of desulfurizing agent, increasing cost. Too small a spray amount does not cover the entire flue gas flow, reducing the desulfurization effect. However, the conventional scheme sets a fixed spraying amount in the whole spraying process based on manual experience, and does not pay attention to the suitability with the flue gas flow, so that the excessive consumption of the desulfurizing agent and the low desulfurizing effect are caused, and the practical application requirements are difficult to meet. Accordingly, an optimized flue gas desulfurization scrubber is desired.
In the technical scheme of the application, a flue gas desulfurization washing tower is provided. Fig. 1 is a block diagram of a flue gas desulfurization scrubber in accordance with an embodiment of the present application. As shown in fig. 1, a flue gas desulfurization scrubber 300 according to an embodiment of the present application includes: a circulation tank 310; a tower 320 having a flue gas outlet formed at an upper portion thereof and a flue gas inlet formed at a lower portion thereof; a main shower pipe 330 connected to the circulation tank and extending into the tower; a circulating water pump 340 provided to the main spray pipe; a shower nozzle 350 connected to the main shower pipe; a spray gun 360 arranged at the head of the spray branch pipe; and a controller 370 for controlling the spray amount of the spray gun.
Specifically, the circulation tank 310. The circulating water tank of the flue gas desulfurization washing tower is equipment for storing and recycling washing liquid used in the desulfurization process. In a flue gas desulfurization system, a desulfurizing agent (such as limestone slurry or ammonia water) is sprayed into flue gas through a spraying layer to react with sulfur dioxide in the flue gas to form sulfate. During this process, the washing liquid is gradually consumed and contaminated, and thus needs to be circulated and treated. Notably, the circulation tank is typically located at or near the bottom of the scrubber and is a closed vessel for storing and circulating the scrubber. The circulation tank is generally internally provided with stirring means to maintain uniform mixing of the washing liquid and suspension of suspended solids. Meanwhile, the circulating water tank is also provided with a pump station and a pipeline system, and the pump station and the pipeline system are used for pumping the washing liquid from the water tank to the spraying layer for recycling. In addition, the circulating water tank also needs to be subjected to regular water quality monitoring and treatment so as to ensure that the quality of the washing liquid meets the requirements. This includes monitoring and adjusting solid particles, ph, dissolved oxygen, etc. in the wash liquor to maintain proper operation and desulfurization effects of the desulfurization system.
Specifically, the tower 320. The tower body of the flue gas desulfurization washing tower is one of key equipment for a flue gas desulfurization system. It is a device for removing harmful gases such as sulfur dioxide (SO 2) in flue gas. The tower body of the flue gas desulfurization washing tower is generally in a cylindrical or square structure and is made of acid and alkali resistant materials (such as glass fiber reinforced plastics, stainless steel and the like). The interior of the tower is typically provided with a spray layer, a packing layer, and an exhaust system. In a specific example, after the flue gas enters the scrubber, the desulfurizing agent is sprayed into the flue gas through the spray layer to form gas-liquid contact. The desulfurizing agent reacts with sulfur dioxide in the flue gas to generate sulfate, thereby achieving the purpose of desulfurization. The filler layer is used for increasing the contact area of gas and liquid and improving the desulfurization efficiency. The exhaust system is used for exhausting the treated flue gas.
Specifically, the main shower pipe 330 is connected to the circulation tank and extends into the tower. Wherein, the main spray pipe which is communicated with the circulating water tank and extends into the tower body plays a key role. The device is mainly used for conveying the washing liquid in the circulating water tank to a spraying layer in the tower body so as to realize spraying operation in the desulfurization process. The main spraying pipeline is connected with the circulating water tank and the tower body, so that the washing liquid can smoothly flow to the spraying layer. The spray layer is usually positioned at the top of the tower body and consists of spray pipes or spray heads. When the washing liquid enters the spraying layer through the main spraying pipe, the spraying pipe or the spray head can uniformly spray the washing liquid on the packing layer. It is noted that the purpose of the spraying operation is to bring the scrubbing liquid into sufficient contact with the flue gas to promote the reaction and absorption of harmful gases such as sulfur dioxide with the scrubbing liquid. The main spraying pipe is used for ensuring that the washing liquid can effectively cover the whole packing layer and providing enough contact area and time to realize efficient desulfurization effect.
Specifically, the circulating water pump 340 is disposed on the main spray pipe. Wherein, the circulating water pump arranged on the main spraying pipeline plays an important role. The circulating water pump is mainly used for pumping the washing liquid in the circulating water tank to the spraying main pipeline, so that the washing liquid can smoothly flow to the spraying layer in the tower body. The circulating water pump draws the washing liquid from the circulating water tank by providing sufficient pressure and flow and pushes the washing liquid to the main spray pipe. This ensures that the spray layer is able to receive enough scrubbing liquid to cover the entire packing layer and make sufficient contact with the flue gas. The circulating water pump is used for ensuring the normal operation and effect of the desulfurization system. The device can maintain the circulating flow of the washing liquid in the circulating water tank, prevent the washing liquid from being detained or blocked, and maintain the continuity and stability of spraying operation. Meanwhile, the circulating water pump can also adjust and control the flow and pressure of the washing liquid so as to adapt to different operation requirements and working condition changes. It is worth noting that the desulfurization system can realize the cyclic utilization of the washing liquid by the action of the circulating water pump, thereby improving the utilization rate of resources and reducing the consumption of water. Meanwhile, the circulating water pump can also keep the freshness and quality of the washing liquid, and ensure the stability and reliability of the desulfurization effect.
Specifically, the spray manifold 350 is connected to the main spray pipe. Wherein, spray the branch pipe and connect in spraying the trunk line, mainly played the important effect of distributing the washing liquid, controlling the volume of spraying and adjusting the mode of spraying, ensure that whole packing layer is sprayed uniformly to improve desulfurization effect.
Specifically, the spray gun 360 is disposed at the head of the spray manifold. Wherein the head of the spray branch pipe is usually provided with a spray gun, which is a device for controlling the spray direction and range of the spray liquid. The method can provide more flexible and accurate injection control, help optimize desulfurization operation, improve efficiency and reduce resource waste.
Specifically, the controller 370 for controlling the spray amount of the spray gun. In particular, in one specific example of the present application, FIG. 2 is a block diagram of a controller in a flue gas desulfurization scrubber in accordance with an embodiment of the present application. Fig. 3 is a system architecture diagram of a flue gas desulfurization scrubber according to an embodiment of the present application. As shown in fig. 2 and 3, a controller 370 of a spray amount of a spray gun according to an embodiment of the present application includes: the smoke flow acquisition module 371 is used for acquiring flow values of ship smoke at a plurality of preset time points in a preset time period; the smoke flow time sequence analysis module 372 is used for performing time sequence correlation analysis on the flow values of the ship smoke at the plurality of preset time points to obtain ship smoke flow time sequence characteristics; and a spray rate control module 373 for determining, based on the marine vessel smoke flow timing characteristics, that the spray rate at the current point in time should be increased or decreased.
In particular, the flue gas flow collection module 371 is configured to obtain flow values of ship flue gas at a plurality of predetermined time points within a predetermined time period. It will be appreciated that adjusting the amount of spray can control the amount of desulfurizing agent sprayed into the flue gas per unit time. If the spraying amount is too large, the desulfurizing agent may be excessively consumed in a short time, wasting costs. Conversely, if the spray amount is too small, the entire flue gas flow may not be covered, reducing the desulfurization effect. Accordingly, it is desirable to adaptively adjust the spray amount based on the time-series variation of the ship's flue gas flow rate to optimize the desulfurization effect.
According to the embodiment of the application, the flow sensor can be used for acquiring the flow value of the ship smoke at a plurality of preset time points in the preset time period. Wherein the flow sensor is a device for measuring the flow rate or velocity of a fluid. It can determine the magnitude of the flow by monitoring the change in speed or pressure of the fluid passing through the sensor. More specifically, the flow sensor plays an important role in fluid flow monitoring, flow control, fault diagnosis, energy saving optimization and the like, and is widely applied to various fields.
Accordingly, in one possible implementation, the flow value of the ship flue gas at a plurality of predetermined time points within a predetermined period of time may be obtained by: determining a predetermined period of time; and a flow sensor is arranged at a proper position of the flue gas desulfurization washing tower. The sensor can accurately measure the flue gas flow; the flow sensor is connected with a data acquisition system. The data acquisition system can record the flow value measured by the sensor; and selecting a plurality of preset time points in a preset time period to sample the flue gas flow. For example, you can choose to sample once an hour, or other time intervals as needed; at each predetermined point in time, the data acquisition system is activated to record the smoke flow rate value. The sensor measures the flow of the flue gas and transmits the flow to the data acquisition system; the data acquisition system stores the recorded flue gas flow values in a database or other storage medium. The safe and reliable data storage is ensured; after all the data at the predetermined time points are collected, you can analyze the data. You can calculate the average flow, peak flow or other interesting indicators to know the change of the flue gas flow.
In particular, the smoke flow timing analysis module 372 is configured to perform timing correlation analysis on flow values of the ship smoke at the plurality of predetermined time points to obtain a ship smoke flow timing characteristic. In particular, in one specific example of the present application, as shown in fig. 5, the flue gas flow timing analysis module 372 includes: a smoke flow time sequence arrangement unit 3721, configured to arrange flow values of the ship smoke at the plurality of predetermined time points into a ship smoke flow time sequence input vector according to a time dimension; the increment coding unit 3722 is configured to perform increment coding on the ship smoke flow time sequence input vector to obtain a ship smoke accumulation time sequence input vector; the smoke flow time sequence change feature extraction unit 3723 is configured to perform feature extraction on the ship smoke flow time sequence input vector and the ship smoke accumulation time sequence input vector through a time sequence feature extractor based on a deep neural network model, so as to obtain a ship smoke flow time sequence feature vector and a ship smoke accumulation time sequence feature vector; the smoke flow multi-scale time sequence feature fusion unit 3724 is configured to fuse the ship smoke flow time sequence feature vector and the ship smoke accumulation time sequence feature vector to obtain a ship smoke flow transient state and an accumulation state time sequence feature vector as the ship smoke flow time sequence feature.
Correspondingly, the smoke flow time sequence arrangement unit 3721 is configured to arrange the flow values of the ship smoke at the plurality of predetermined time points into a ship smoke flow time sequence input vector according to a time dimension. The method takes into consideration that the flow value of the ship smoke has a dynamic change rule in the time dimension, namely, the flow value of the ship smoke has time-sequence association characteristic information between each preset time point in time sequence. Therefore, in the technical scheme of the application, firstly, the flow values of the ship smoke at a plurality of preset time points are further arranged into the time sequence input vector of the ship smoke flow according to the time dimension, so that the distribution information of the flow values of the ship smoke on the time sequence is integrated.
Accordingly, in one possible implementation manner, the flow values of the ship smoke at the plurality of predetermined time points may be arranged into a ship smoke flow time sequence input vector according to a time dimension, for example: installing a proper flow sensor in a ship flue gas pipeline so as to measure the flow speed or flow of the flue gas; connecting the flow sensor to a data acquisition system to ensure that the output of the sensor can be accurately recorded and stored; a point in time is determined at which the flow of flue gas needs to be recorded. These time points may be predetermined fixed time intervals or specific time points which are adjusted as needed; the smoke flow rate value is recorded beginning at a predetermined point in time. The data acquisition system reads the output of the flow sensor and stores the output as time series data; storing the collected flue gas flow data in a suitable data storage device, such as a database or file system; and carrying out time sequence correlation analysis on the stored flue gas flow data. The following are some common analytical steps: and (3) data collection: collecting flue gas flow data to be analyzed; data preprocessing: cleaning and processing the data, including removing abnormal values, filling missing values and the like; and (3) drawing a time sequence chart: the change of the flue gas flow rate with time is visualized by using a suitable graph tool, such as a line graph or a graph; calculating a statistical index: calculating statistical indexes of the flue gas flow data, such as average value, maximum value, minimum value and the like; analysis of periodicity: detecting a periodic pattern in the data by applying a periodic analysis method (such as fourier transform); and (3) detecting trend: detecting trend patterns in the data by applying trend analysis methods (such as linear regression); finding an abnormal point: identifying and analyzing outliers or outliers in the data; explaining the timing characteristics: and (5) explaining time sequence characteristics in the flue gas flow data, such as periodical change, trend change and abnormal events, according to the analysis result.
Correspondingly, the incremental encoding unit 3722 is configured to perform incremental encoding on the ship smoke flow time sequence input vector to obtain a ship smoke accumulation time sequence input vector. Considering that in the process of performing the self-adaptive control of the spraying amount, since the data information of each position in the time sequence input vector of the ship smoke flow represents the ship smoke flow value at each preset time point, if the self-adaptive control of the spraying speed is performed only by using the transient time sequence characteristic of the ship smoke flow, the time sequence characteristic of the smoke flow at the past time point is ignored, so that the spraying speed and the adaptation degree of the ship smoke flow are reduced, and the desulfurization effect is reduced. Therefore, in the technical scheme of the application, the ship smoke flow time sequence input vector is further subjected to incremental coding to obtain the ship smoke accumulation time sequence input vector, so that the information of the past moment is fused into the characteristic representation of the current moment, the evolution process and the change trend of the smoke flow are better reflected, and the self-adaptive control of the spraying speed is facilitated. It should be noted that, here, the characteristic value of each position in the ship smoke accumulation time sequence input vector is the sum of the characteristic value of the corresponding position in the ship smoke flow time sequence input vector and the characteristic value of the previous position.
Accordingly, in one possible implementation, the ship smoke flow timing input vector may be incrementally encoded to obtain a ship smoke accumulation timing input vector, for example, by: first, an initial value needs to be defined for representing the cumulative value of the flue gas flow. This initial value may be 0 or an accumulated value of the previous time point; traversing the flow value of each time point in the time sequence input vector of the flue gas flow in sequence from the first time point; for the flow value of each time point, adding the flow value of each time point with the accumulated value of the previous time point to obtain the accumulated value of the current time point; adding the accumulated value of the current time point into the ship smoke accumulated time sequence input vector; repeating the step 3 and the step 4 until the complete flow time sequence input vector is traversed.
Correspondingly, the smoke flow time sequence variation feature extraction unit 3723 is configured to perform feature extraction on the ship smoke flow time sequence input vector and the ship smoke accumulation time sequence input vector through a time sequence feature extractor based on a deep neural network model, so as to obtain a ship smoke flow time sequence feature vector and a ship smoke accumulation time sequence feature vector. In particular, in one specific example of the present application, the deep neural network model is a one-dimensional convolutional neural network model. That is, for the transient state and the accumulated state information of the ship flue gas flow, there are respective dynamic change rules in the time dimension, so that in order to more fully describe the time sequence change characteristics of the ship flue gas flow, in the technical scheme of the application, the time sequence change characteristics of the transient state and the accumulated state of the ship flue gas flow need to be extracted and fused. Specifically, in the technical scheme of the application, the ship smoke flow time sequence input vector and the ship smoke accumulation time sequence input vector are respectively subjected to feature mining in a time sequence feature extractor based on a one-dimensional convolutional neural network model so as to respectively extract time sequence related feature distribution information of transient state and accumulation state of the ship smoke flow, thereby obtaining a ship smoke flow time sequence feature vector and a ship smoke accumulation time sequence feature vector. Specifically, each layer of the time sequence feature extractor based on the one-dimensional convolutional neural network model is used for respectively carrying out the forward transfer on input data: carrying out convolution processing on input data to obtain a convolution characteristic diagram; pooling the convolution feature images based on a feature matrix to obtain pooled feature images; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the time sequence feature extractor based on the one-dimensional convolutional neural network model is the ship smoke flow time sequence feature vector and the ship smoke accumulation time sequence feature vector, and the input of the first layer of the time sequence feature extractor based on the one-dimensional convolutional neural network model is the ship smoke flow time sequence input vector and the ship smoke accumulation time sequence input vector.
A one-dimensional convolutional neural network (1D CNN) is a deep learning model for processing sequence data. Unlike conventional convolutional neural networks, 1D CNNs perform one-dimensional convolutional operations on input data, which can effectively capture local patterns and features in sequence data. The 1D CNN is typically composed of multiple convolutional layers, an activation function, a pooling layer, and a fully-connected layer. The convolution layer carries out convolution operation on the input sequence in a sliding window mode, and features at different positions are extracted. The activation function introduces nonlinear factors that enhance the expressive power of the network. The pooling layer is used to reduce the size of the feature map and preserve the most important features. The fully connected layer maps the characteristics of the pooling layer output to final output categories or values.
It should be noted that, in other specific examples of the present application, the time sequence feature extractor based on the deep neural network model may perform feature extraction on the time sequence input vector of the ship smoke flow and the time sequence input vector of the ship smoke accumulation respectively to obtain the time sequence feature vector of the ship smoke flow and the time sequence feature vector of the ship smoke accumulation in other manners, for example: data preparation: the ship smoke flow time sequence data and the ship smoke accumulation time sequence data are arranged into a proper input vector form, so that the data alignment and the correct format are ensured; the data is preprocessed, such as normalized, smoothed, etc., to better accommodate the input requirements of the model. Model selection: an appropriate deep neural network model is selected for time series feature extraction. Common models include Recurrent Neural Networks (RNNs), long-short-term memory networks (LSTM), gated loop units (GRUs), etc.; appropriate model structures and parameter settings are selected based on the specific task and data characteristics. Feature extraction: inputting the time sequence input vector of the ship smoke flow into a deep neural network model; gradually extracting time sequence features in the input vector by the model through forward propagation calculation; features are extracted in the appropriate layers and the output of the intermediate or final layer can be selected as a feature representation as desired. Feature vector generation: and converting the time sequence features extracted by the model into feature vectors. The time sequence features can be converted into feature vectors with fixed lengths by using methods such as pooling operation, global average pooling or global maximum pooling; the feature vectors are normalized to better accommodate subsequent feature analysis and model training. Outputting a result: obtaining a ship smoke flow time sequence feature vector and a ship smoke accumulation time sequence feature vector; the feature vectors may be used for subsequent tasks of timing analysis, anomaly detection, classification, etc., or as input for training and prediction of other models.
Correspondingly, the smoke flow multi-scale time sequence feature fusion unit 3724 is configured to fuse the ship smoke flow time sequence feature vector and the ship smoke accumulation time sequence feature vector to obtain a ship smoke flow transient state and an accumulation state time sequence feature vector as the ship smoke flow time sequence feature. That is, the ship smoke flow time sequence feature vector and the ship smoke accumulation time sequence feature vector are fused, so that the transient time sequence change feature information of the ship smoke flow and the accumulated state time sequence change feature information of the ship smoke flow are fused, and the ship smoke flow transient state and accumulated state time sequence feature vector with the time sequence fusion association feature information of the transient state and the accumulated state of the ship smoke flow is obtained, so that the time sequence change trend and the evolution process of the ship smoke flow are more accurately represented.
It should be noted that, in other specific examples of the present application, the time-series correlation analysis may be performed on the flow values of the ship flue gas at the plurality of predetermined time points in other manners to obtain the time-series characteristics of the ship flue gas flow, for example: and acquiring ship smoke flow value data at a plurality of preset time points from the data acquisition system. Ensuring that the data is accurate and complete and contains time stamp information; the data is preprocessed, including outlier removal, missing value processing and data smoothing. Ensuring data quality and reliability; and drawing the change of the ship smoke flow value along with time into a time sequence chart. The horizontal axis represents time, and the vertical axis represents the smoke flow value. A line graph or graph may be used to reveal timing characteristics; and calculating various statistical indexes to describe the time sequence characteristics of the ship smoke flow. Common statistical indicators include mean, standard deviation, maximum, minimum, etc. These indicators may provide information about the central trend, degree of dispersion and extremum of the flue gas flow; detecting whether a periodic variation exists. The periodic characteristics of the data may be analyzed using fourier transforms or autocorrelation functions, or the like. Periodic analysis can reveal repeated patterns of flue gas flow over different time scales; and detecting trend change of the flue gas flow. Linear regression, moving average, etc. methods can be used to fit the trend lines and analyze the trend changes. The trend analysis can reveal the long-term change trend of the flue gas flow; by comparing the actual observed value with the expected pattern or statistical law, the possible abnormal points are found. An outlier may represent a system failure, an operational anomaly, or other special condition; and according to the analysis result, explaining the time sequence characteristics of the ship smoke flow. For example, a periodic change in flow rate, a trend of change in trend, a cause of an abnormal point, or the like is indicated.
In particular, the spray rate control module 373 is configured to determine, based on the marine vessel smoke flow timing characteristics, whether the spray rate at the current point in time should be increased or decreased. In particular, in one specific example of the present application, the spray rate control module 373 includes: and (3) passing the transient state and cumulative state time sequence feature vectors of the ship smoke flow through a classifier to obtain a classification result, wherein the classification result is used for indicating that the spraying speed at the current time point should be increased or decreased. That is, the classification processing is performed by fusing the associated characteristic information with the time sequence of the transient state and the accumulated state of the ship flue gas flow, so that the self-adaptive control of the spraying speed is accurately performed in real time based on the time sequence change condition of the ship flue gas flow, thereby enabling the desulfurizing agent and the sulfur dioxide to fully react to ensure the effective utilization of the desulfurizing agent and the maximization of the desulfurizing effect. Specifically, using a plurality of full-connection layers of the classifier to carry out full-connection coding on the transient state and the accumulated state time sequence feature vectors of the ship smoke flow so as to obtain coded classification feature vectors; and passing the coding classification feature vector through a Softmax classification function of the classifier to obtain the classification result.
A Classifier (Classifier) refers to a machine learning model or algorithm that is used to classify input data into different categories or labels. The classifier is part of supervised learning, which performs classification tasks by learning mappings from input data to output categories.
The fully connected layer (Fully Connected Layer) is one type of layer commonly found in neural networks. In the fully connected layer, each neuron is connected to all neurons of the upper layer, and each connection has a weight. This means that each neuron in the fully connected layer receives inputs from all neurons in the upper layer, and weights these inputs together, and then passes the result to the next layer.
The Softmax classification function is a commonly used activation function for multi-classification problems. It converts each element of the input vector into a probability value between 0 and 1, and the sum of these probability values equals 1. The Softmax function is commonly used at the output layer of a neural network, and is particularly suited for multi-classification problems, because it can map the network output into probability distributions for individual classes. During the training process, the output of the Softmax function may be used to calculate the loss function and update the network parameters through a back propagation algorithm. Notably, the output of the Softmax function does not change the relative magnitude relationship between elements, but rather normalizes them. Thus, the Softmax function does not change the characteristics of the input vector, but simply converts it into a probability distribution form.
It should be noted that, in other specific examples of the present application, it may also be determined that the spraying speed at the current time point should be increased or decreased based on the ship smoke flow timing characteristics in other manners, for example: acquiring a ship smoke flow value at the current time point: acquiring a ship smoke flow value at the current time point by using a flow sensor; analyzing the spraying speed at the current time point: determining whether the spraying speed at the current time point should be increased or decreased according to the ship smoke flow value at the current time point and the previous time sequence feature analysis result; judging the spray speed adjusting direction of the current time point: comparing the ship smoke flow value at the current time point with the previous time sequence characteristic, and if the ship smoke flow value at the current time point is larger and the trend is raised, increasing the spraying speed possibly needs to be carried out; if the ship smoke flow value at the current point in time is small and the trend is decreasing, it may be necessary to reduce the spraying speed; adjusting the spraying speed: according to the judged adjusting direction, correspondingly increasing or decreasing the spraying speed at the current time point; monitoring and adjusting effects: and according to the adjusted spraying speed, observing the change trend of the ship smoke flow, and evaluating the adjusting effect.
It will be appreciated that the one-dimensional convolutional neural network model-based timing feature extractor and the classifier need to be trained prior to the inference using the neural network model described above. That is, the flue gas desulfurization scrubber 300 according to an embodiment of the present application further comprises a training stage 400 for training the one-dimensional convolutional neural network model-based time series feature extractor and the classifier.
Fig. 4 is a block diagram of a training phase of a flue gas desulfurization scrubber in accordance with an embodiment of the present application. As shown in fig. 3, a flue gas desulfurization scrubber 300 according to an embodiment of the present application includes: training phase 400, comprising: a training data acquisition unit 410, configured to acquire training data, where the training data includes flow values of a training ship smoke at a plurality of predetermined time points within a predetermined period, and a spraying speed at the current time point should be increased or decreased; a training smoke flow time sequence arrangement unit 420, configured to arrange flow values of training ship smoke at the plurality of predetermined time points into training ship smoke flow time sequence input vectors according to a time dimension; the training increment encoding unit 430 is configured to perform increment encoding on the training ship smoke flow time sequence input vector to obtain a training ship smoke accumulation time sequence input vector, where the feature value of each position in the training ship smoke accumulation time sequence input vector is the sum of the feature value of the corresponding position in the training ship smoke flow time sequence input vector and the feature value of the previous position; a training smoke flow time sequence feature extraction unit 440, configured to pass the training ship smoke flow time sequence input vector and the training ship smoke accumulation time sequence input vector through the one-dimensional convolutional neural network model-based time sequence feature extractor to obtain a training ship smoke flow time sequence feature vector and a training ship smoke accumulation time sequence feature vector; a training smoke flow multi-scale time sequence feature fusion unit 450, configured to fuse the training ship smoke flow time sequence feature vector and the training ship smoke accumulation time sequence feature vector to obtain a training ship smoke flow transient state and an accumulation state time sequence feature vector; a classification loss unit 460, configured to pass the transient state and the accumulated state time sequence feature vectors of the training ship flue gas flow through the classifier to obtain a classification loss function value; a common manifold implicit similarity loss unit 470 for calculating a common manifold implicit similarity factor of the training ship flue gas flow timing feature vector and the training ship flue gas accumulation timing feature vector to obtain a common manifold implicit similarity loss function value; a model training unit 480 for training the one-dimensional convolutional neural network model-based timing feature extractor and the classifier with a weighted sum of the classification loss function value and the common manifold implicit similarity loss function value as a loss function value, and traveling in a gradient descent direction.
In particular, in the technical scheme of the application, the ship smoke flow time sequence characteristic vector and the ship smoke accumulation time are adoptedThe sequence feature vector expresses one-dimensional local correlation features of a flow value of ship smoke and an increment value of ship smoke flow along a time sequence in a time sequence direction respectively, so that the ship smoke flow sequence feature vector and the ship smoke accumulation time sequence feature vector also have more obvious feature distribution differences due to the difference of distribution modes of the flow value of the ship smoke and the time sequence enhancement of the flow value of the ship smoke in the time sequence direction, and geometric monotonicity of high-dimensional feature manifold expression of the ship smoke transient state and the accumulation state time sequence feature vector in a high-dimensional feature space obtained by fusing the ship smoke flow sequence feature vector and the ship smoke accumulation time sequence feature vector is used, so that the convergence effect of classification regression of the ship smoke through a classifier is influenced, namely, the training speed and the accuracy of training results are reduced. Based on this, the applicant of the present application addresses the marine vessel smoke flow time series feature vector, e.g. noted asAnd the ship smoke cumulative time sequence feature vector, for example, marked as +.>The common manifold implicit similarity factor is introduced as a loss function, specifically expressed as:
wherein , and />The time sequence feature vectors of the smoke flow of the training ship are respectivelyAnd said training vessel smoke accumulation timing feature vector, < > and>representing the two norms of the vector, and +.>The square root of the Frobenius norm of the matrix is represented, the training ship smoke flow time sequence characteristic vector and the training ship smoke accumulation time sequence characteristic vector are in column vector form, and the training ship smoke flow time sequence characteristic vector is in column vector form> 、/> 、/> and />For the weight super parameter, ++>Representing vector multiplication, ++>Representing multiplication by location +.>Representing difference by position +.>Representing the common manifold implicit similarity loss function value. Here, the common manifold implicit similarity factor may be represented by the ship flue gas flow timing feature vector +.>And the ship smoke accumulation time sequence characteristic vector +.>Structured association between to represent respective featuresManifold sharing manifold under cross dimension, and sharing the ship smoke flow time sequence characteristic vector by same factorization weight>And the ship smoke accumulation time sequence characteristic vector +.>And (3) the common constraint of manifold structural factors such as variability, correspondence, relevance and the like, so as to measure the distribution similarity of geometric derivative structure representations depending on common manifold, thereby realizing nonlinear geometric monotonicity of fusion characteristics among cross-modal characteristic distributions, improving the geometric monotonicity of high-dimensional characteristic manifold expressions of the ship smoke flow transient state and cumulative state time sequence characteristic vectors, and improving the convergence effect of classification regression of the ship smoke flow transient state and cumulative state time sequence characteristic vectors through the classifier, namely improving the training speed and the accuracy of training results. Therefore, the self-adaptive control of the spraying speed can be performed in real time based on the time sequence change trend of the ship smoke flow, so that the effective utilization of the desulfurizing agent and the maximization of the desulfurizing effect are ensured.
Further, the application also provides a flue gas desulfurization washing method. Fig. 6 is a flow chart of a flue gas desulfurization scrubbing method according to an embodiment of the present application. As shown in fig. 6, the flue gas desulfurization washing method according to an embodiment of the present application includes: s110, acquiring flow values of ship smoke at a plurality of preset time points in a preset time period; s120, carrying out time sequence correlation analysis on the flow values of the ship smoke at a plurality of preset time points to obtain the time sequence characteristics of the ship smoke flow; and S130, determining that the spraying speed at the current time point is increased or reduced based on the ship smoke flow time sequence characteristics.
In summary, the flue gas desulfurization washing method according to the embodiment of the application is explained, wherein the sensor is used for collecting the ship flue gas flow values at a plurality of preset time points, and a data processing and analyzing algorithm is introduced into the rear end to perform time sequence analysis of the ship flue gas flow so as to judge the time sequence change trend of the ship flue gas flow, thus the self-adaptive adjustment of the spraying quantity is performed in real time, and the effective utilization of the desulfurizing agent and the maximization of the desulfurizing effect are ensured.
The foregoing description of the embodiments of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the various embodiments described. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or the improvement of technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (10)
1. A flue gas desulfurization scrubber, comprising:
a circulation water tank;
the tower body is provided with a flue gas outlet formed at the upper part of the tower body and a flue gas inlet formed at the lower part of the tower body;
a main spray pipe communicated with the circulating water tank and extending into the tower body;
the circulating water pump is arranged on the main spraying pipeline;
the spraying branch pipe is connected with the main spraying pipe;
the spray gun is arranged at the head part of the spray branch pipe; and
and a controller for controlling the spraying amount of the spray gun.
2. The flue gas desulfurization scrubber of claim 1, wherein the controller comprises:
the smoke flow acquisition module is used for acquiring flow values of ship smoke at a plurality of preset time points in a preset time period;
the smoke flow time sequence analysis module is used for performing time sequence correlation analysis on the flow values of the ship smoke at the plurality of preset time points to obtain ship smoke flow time sequence characteristics; and
and the spraying speed control module is used for determining that the spraying speed at the current time point should be increased or decreased based on the ship smoke flow time sequence characteristics.
3. The flue gas desulfurization scrubber of claim 2, wherein the flue gas flow timing analysis module comprises:
The smoke flow time sequence arrangement unit is used for arranging the flow values of the ship smoke at the plurality of preset time points into ship smoke flow time sequence input vectors according to the time dimension;
the increment coding unit is used for performing increment coding on the ship smoke flow time sequence input vector to obtain a ship smoke accumulation time sequence input vector;
the smoke flow time sequence change feature extraction unit is used for respectively carrying out feature extraction on the ship smoke flow time sequence input vector and the ship smoke accumulation time sequence input vector through a time sequence feature extractor based on a deep neural network model so as to obtain a ship smoke flow time sequence feature vector and a ship smoke accumulation time sequence feature vector;
the smoke flow multi-scale time sequence feature fusion unit is used for fusing the ship smoke flow time sequence feature vector and the ship smoke accumulation time sequence feature vector to obtain a ship smoke flow transient state and accumulation state time sequence feature vector as the ship smoke flow time sequence feature.
4. A flue gas desulfurization scrubber according to claim 3, wherein the characteristic value of each position in the vessel flue gas accumulation timing input vector is a sum of the characteristic value of the corresponding position in the vessel flue gas flow timing input vector and the characteristic value of the previous position.
5. The flue gas desulfurization scrubber of claim 4, wherein the deep neural network model is a one-dimensional convolutional neural network model.
6. The flue gas desulfurization scrubber of claim 5, wherein the spray rate control module is configured to: and (3) passing the transient state and cumulative state time sequence feature vectors of the ship smoke flow through a classifier to obtain a classification result, wherein the classification result is used for indicating that the spraying speed at the current time point should be increased or decreased.
7. The flue gas desulfurization scrubber of claim 6, further comprising a training module for training the one-dimensional convolutional neural network model-based temporal feature extractor and the classifier.
8. The flue gas desulfurization scrubber of claim 7, wherein the training module comprises:
the training data acquisition unit is used for acquiring training data, wherein the training data comprise flow values of training ship smoke at a plurality of preset time points in a preset time period, and the spraying speed at the current time point is increased or decreased;
the training smoke flow time sequence arrangement unit is used for arranging the flow values of the training ship smoke at a plurality of preset time points into training ship smoke flow time sequence input vectors according to the time dimension;
The training increment coding unit is used for performing increment coding on the training ship smoke flow time sequence input vector to obtain a training ship smoke accumulation time sequence input vector, wherein the characteristic value of each position in the training ship smoke accumulation time sequence input vector is the sum of the characteristic value of the corresponding position in the training ship smoke flow time sequence input vector and the characteristic value of the previous position;
the training smoke flow time sequence feature extraction unit is used for enabling the training ship smoke flow time sequence input vector and the training ship smoke accumulation time sequence input vector to respectively pass through the time sequence feature extractor based on the one-dimensional convolutional neural network model so as to obtain a training ship smoke flow time sequence feature vector and a training ship smoke accumulation time sequence feature vector;
the training smoke flow multi-scale time sequence feature fusion unit is used for fusing the training ship smoke flow time sequence feature vector and the training ship smoke accumulation time sequence feature vector to obtain training ship smoke flow transient state and accumulation state time sequence feature vector;
the classification loss unit is used for enabling the transient state and the accumulated state time sequence feature vectors of the smoke flow of the training ship to pass through the classifier so as to obtain a classification loss function value;
The common manifold implicit similarity loss unit is used for calculating a common manifold implicit similarity factor of the training ship smoke flow time sequence feature vector and the training ship smoke accumulation time sequence feature vector to obtain a common manifold implicit similarity loss function value;
and the model training unit is used for training the time sequence feature extractor and the classifier based on the one-dimensional convolutional neural network model by taking the weighted sum of the classification loss function value and the common manifold implicit similarity loss function value as the loss function value and transmitting the time sequence feature extractor and the classifier in the gradient descending direction.
9. The flue gas desulfurization scrubber of claim 8, wherein said common manifold implicit similarity loss unit is configured to: calculating a common manifold implicit similarity factor of the training ship smoke flow time sequence feature vector and the training ship smoke accumulation time sequence feature vector by using the following loss formula to obtain a common manifold implicit similarity loss function value;
wherein, the loss formula is:
wherein , and />The time sequence feature vector of the smoke flow of the training ship and the time sequence feature vector of the smoke accumulation of the training ship are respectively +. >Representing the two norms of the vector, and +.>The square root of the Frobenius norm of the matrix is represented, the training ship smoke flow time sequence characteristic vector and the training ship smoke accumulation time sequence characteristic vector are in column vector form, and the training ship smoke flow time sequence characteristic vector is in column vector form> 、/> 、/> and />For the weight super parameter, ++>Representing vector multiplication, ++>Representing multiplication by location +.>Representing difference by position +.>Representing the common manifold implicit similarity loss function value.
10. A flue gas desulfurization scrubbing method, comprising:
acquiring flow values of ship smoke at a plurality of preset time points in a preset time period;
the smoke flow time sequence analysis module is used for performing time sequence correlation analysis on the flow values of the ship smoke at the plurality of preset time points to obtain ship smoke flow time sequence characteristics; and
and the spraying speed control module is used for determining that the spraying speed at the current time point should be increased or decreased based on the ship smoke flow time sequence characteristics.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311008799.4A CN116983819A (en) | 2023-08-11 | 2023-08-11 | Flue gas desulfurization washing tower and method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311008799.4A CN116983819A (en) | 2023-08-11 | 2023-08-11 | Flue gas desulfurization washing tower and method thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116983819A true CN116983819A (en) | 2023-11-03 |
Family
ID=88526422
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311008799.4A Pending CN116983819A (en) | 2023-08-11 | 2023-08-11 | Flue gas desulfurization washing tower and method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116983819A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117180952A (en) * | 2023-11-07 | 2023-12-08 | 湖南正明环保股份有限公司 | Multi-directional airflow material layer circulation semi-dry flue gas desulfurization system and method thereof |
| CN117244270A (en) * | 2023-11-20 | 2023-12-19 | 新疆凯龙清洁能源股份有限公司 | Recycling method of sulfur-containing low-pressure alkane gas |
-
2023
- 2023-08-11 CN CN202311008799.4A patent/CN116983819A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117180952A (en) * | 2023-11-07 | 2023-12-08 | 湖南正明环保股份有限公司 | Multi-directional airflow material layer circulation semi-dry flue gas desulfurization system and method thereof |
| CN117180952B (en) * | 2023-11-07 | 2024-02-02 | 湖南正明环保股份有限公司 | Multi-directional airflow material layer circulation semi-dry flue gas desulfurization system and method thereof |
| CN117244270A (en) * | 2023-11-20 | 2023-12-19 | 新疆凯龙清洁能源股份有限公司 | Recycling method of sulfur-containing low-pressure alkane gas |
| CN117244270B (en) * | 2023-11-20 | 2024-02-13 | 新疆凯龙清洁能源股份有限公司 | Recycling method of sulfur-containing low-pressure alkane gas |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116983819A (en) | Flue gas desulfurization washing tower and method thereof | |
| CN109034260B (en) | Desulfurization tower oxidation fan fault diagnosis system and method based on statistical principle and intelligent optimization | |
| CN112000084B (en) | Intelligent BIT design method of controller module based on 1D-CNN and GRU-SVM | |
| US20220365521A1 (en) | Virtual simulation manufacturing platform based on automatic control | |
| CN111241755A (en) | Power load prediction method | |
| CN110716512A (en) | Environmental protection equipment performance prediction method based on coal-fired power plant operation data | |
| Mathew et al. | Regression kernel for prognostics with support vector machines | |
| CN101706882A (en) | Embedded platform based neural network model online training method | |
| Wang et al. | Remaining useful life prediction based on improved temporal convolutional network for nuclear power plant valves | |
| CN117648643B (en) | Rigging predictive diagnosis method and device based on artificial intelligence | |
| CN114519923A (en) | Intelligent diagnosis and early warning method and system for power plant | |
| CN116688754A (en) | Ship flue gas desulfurization automatic control system and method thereof | |
| CN114225662A (en) | Flue gas desulfurization and denitrification optimization control method based on hysteresis model | |
| Shao et al. | Predictive scheduling of wet flue gas desulfurization system based on reinforcement learning | |
| Alvi et al. | Cost effective soft sensing for wastewater treatment facilities | |
| CN113505497B (en) | Method and system for monitoring slurry quality of wet flue gas desulfurization absorption tower | |
| CN117132089B (en) | Power utilization strategy optimization scheduling method and device | |
| Yang et al. | Teacher-student uncertainty autoencoder for the process-relevant and quality-relevant fault detection in the industrial process | |
| KR20220078456A (en) | Method for air purifier filter life prediction based on ensemble supervised learning | |
| CN116933643A (en) | Intelligent data monitoring method based on partial robust M regression and multiple interpolation | |
| Won et al. | Prediction of remaining useful lifetime of membrane using machine learning | |
| CN114414648A (en) | Automatic potentiometric titration method and system based on machine learning | |
| CN113688973A (en) | Water supply pump station scheduling method based on deep neural network | |
| Prabowo et al. | Improving Internet of Things Platform with Anomaly Detection for Environmental Sensor Data | |
| Ramos et al. | Visual supervision of a waste water biological reactor using artificial intelligence algorithms |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |