CN116392930A - Ship tail gas desulfurization process and system thereof - Google Patents

Ship tail gas desulfurization process and system thereof Download PDF

Info

Publication number
CN116392930A
CN116392930A CN202310445481.6A CN202310445481A CN116392930A CN 116392930 A CN116392930 A CN 116392930A CN 202310445481 A CN202310445481 A CN 202310445481A CN 116392930 A CN116392930 A CN 116392930A
Authority
CN
China
Prior art keywords
time sequence
training
gas
seawater
pressure
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.)
Granted
Application number
CN202310445481.6A
Other languages
Chinese (zh)
Other versions
CN116392930B (en
Inventor
郑浣琪
柴剑
沈海涛
陈煜�
方德忠
陈飞
王兴如
王汝能
程时明
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang Zheneng Mailing Environmental Technology Co ltd
Original Assignee
Zhejiang Zheneng Mailing Environmental Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Zhejiang Zheneng Mailing Environmental Technology Co ltd filed Critical Zhejiang Zheneng Mailing Environmental Technology Co ltd
Priority to CN202310445481.6A priority Critical patent/CN116392930B/en
Publication of CN116392930A publication Critical patent/CN116392930A/en
Application granted granted Critical
Publication of CN116392930B publication Critical patent/CN116392930B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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
    • B01D53/14—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 by absorption
    • B01D53/1456—Removing acid components
    • B01D53/1481—Removing sulfur dioxide or sulfur trioxide
    • 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
    • B01D53/14—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 by absorption
    • B01D53/1412—Controlling the absorption process
    • C—CHEMISTRY; METALLURGY
    • C02—TREATMENT OF WATER, WASTE WATER, SEWAGE, OR SLUDGE
    • C02F—TREATMENT OF WATER, WASTE WATER, SEWAGE, OR SLUDGE
    • C02F1/00—Treatment of water, waste water, or sewage
    • C02F1/72—Treatment of water, waste water, or sewage by oxidation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • C—CHEMISTRY; METALLURGY
    • C02—TREATMENT OF WATER, WASTE WATER, SEWAGE, OR SLUDGE
    • C02F—TREATMENT OF WATER, WASTE WATER, SEWAGE, OR SLUDGE
    • C02F2103/00—Nature of the water, waste water, sewage or sludge to be treated
    • C02F2103/08—Seawater, e.g. for desalination
    • C—CHEMISTRY; METALLURGY
    • C02—TREATMENT OF WATER, WASTE WATER, SEWAGE, OR SLUDGE
    • C02F—TREATMENT OF WATER, WASTE WATER, SEWAGE, OR SLUDGE
    • C02F2103/00—Nature of the water, waste water, sewage or sludge to be treated
    • C02F2103/18—Nature of the water, waste water, sewage or sludge to be treated from the purification of gaseous effluents
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A50/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE in human health protection, e.g. against extreme weather
    • Y02A50/20—Air quality improvement or preservation, e.g. vehicle emission control or emission reduction by using catalytic converters

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Analytical Chemistry (AREA)
  • General Chemical & Material Sciences (AREA)
  • Oil, Petroleum & Natural Gas (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • Biomedical Technology (AREA)
  • Evolutionary Computation (AREA)
  • Water Supply & Treatment (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Environmental & Geological Engineering (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Organic Chemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Hydrology & Water Resources (AREA)
  • Treating Waste Gases (AREA)

Abstract

The utility model relates to an intelligent control field, it specifically discloses a boats and ships tail gas desulfurization technology and system thereof, and it excavates boats and ships tail gas atmospheric pressure value with the differential pressure time sequence dynamic change characteristic information of sea water hydraulic pressure value, and boats and ships tail gas velocity value with the time sequence between the sea water velocity value changes associated feature information to this is based on boats and ships tail gas with the reinforcing of differential pressure time sequence change characteristic is carried out to the velocity time sequence associated feature of sea water to optimize the expression of differential pressure time sequence change characteristic, with improve the differential pressure control precision between sea water side liquid phase pressure and the boats and ships tail gas side gas phase pressure, optimize the desulfurization effect and the efficiency of boats and ships tail gas.

Description

Ship tail gas desulfurization process and system thereof
Technical Field
The application relates to the field of intelligent control, and more particularly, to a marine tail gas desulfurization process and a system thereof.
Background
Currently, main processes adopted by the desulfurization technology are a gypsum flue gas desulfurization method, a rotary spray drying desulfurization method, a seawater desulfurization method and the like. Wherein the seawater desulfurization method is a mature desulfurization technology developed in recent decades, and generally adopts an absorption tower process, and fully utilizes the acid-base buffering capacity and the strong acid gas neutralization capacity of natural seawater to effectively remove SO in the flue gas 2 . The seawater desulfurization process is simple in flow, high in efficiency, environment-friendly, high in reliability and economy, less in pollution to the ecological environment and considered as one of the ideal ship tail gas treatment methods. But also has the defects of large occupied space of equipment, poor effect when treating the tail gas discharged by the combustion of the high-sulfur fuel, and the like.
The membrane contactor technology is a new desulfurization and decarbonization technology developed in recent years, and is a membrane process for realizing interphase mass transfer without direct contact of two phases. In this process, the microporous membrane serves as only one interface between the two phases, separating the ship's tail gas phase from the absorbent seawater phase.SO in the gas phase 2 Can pass through the contact interface of the membrane and enter the absorbent phase to be taken away, thus achieving the purpose of desulfurization. The membrane contactor method has many outstanding advantages over the conventional absorber process, such as greatly increasing specific surface area, improving absorption effect, reducing equipment height and volume, reducing operating costs, etc. In addition, the membrane contact method has more flexible operability, does not depend on the gas-liquid flow rate, has stable performance, does not have the defects of bubbling, entrainment, flooding and the like, and has obvious advantages and good application prospects in the field of ship tail gas desulfurization and purification.
However, in the process of desulfurizing the tail gas of the ship by using the membrane contactor technology in practice, it is found that because the combustion of the diesel fuel of the ship is usually incomplete, the small solid particles and oil drops contained in the diesel fuel of the ship inevitably pollute the membrane holes and affect SO 2 Diffusion and mass transfer. And absorb SO 2 The sulfite is oxidized into sulfate before the post-seawater is discharged, and the conventional aeration method has low efficiency and long treatment time, and influences the practicability of the whole system.
Accordingly, an optimized marine exhaust gas desulfurization system is desired.
Disclosure of Invention
The present application has been made in order to solve the above technical problems. The embodiment of the application provides a ship tail gas desulfurization process and a system thereof, wherein the ship tail gas desulfurization process is characterized in that through an artificial intelligence technology of deep learning, differential pressure time sequence dynamic change characteristic information of a ship tail gas pressure value and a seawater hydraulic pressure value and time sequence change association characteristic information between a ship tail gas flow velocity value and a seawater flow velocity value are excavated, so that differential pressure time sequence change characteristic enhancement is carried out based on the ship tail gas and the seawater flow velocity time sequence association characteristic, the expression of the differential pressure time sequence change characteristic is optimized, the differential pressure control accuracy between the seawater side liquid phase pressure and the ship tail gas side gas phase pressure is improved, and the desulfurization effect and efficiency of the ship tail gas are optimized.
According to one aspect of the present application, there is provided a marine vessel exhaust gas desulfurization system comprising:
the ship tail gas and sea water collecting module is used for collecting the sulfur-containing ship tail gas and sea water;
the sulfur-containing ship tail gas pretreatment module is used for pretreating the sulfur-containing ship tail gas to obtain tail gas after dust removal and oil removal;
the seawater pretreatment module is used for pretreating the seawater to obtain pretreated seawater;
the membrane contactor desulfurization module is used for introducing the tail gas after dust removal and oil removal and the pretreated seawater into a membrane contactor to carry out marine tail gas desulfurization so as to obtain seawater after sulfur dioxide absorption;
and the seawater post-treatment module is used for carrying out oxidation treatment on the seawater absorbing sulfur dioxide to obtain oxidized sulfate.
In the above marine exhaust gas desulfurization system, the membrane contactor desulfurization module includes: the data acquisition unit is used for acquiring the air pressure value and the flow velocity value of the ship tail gas at a plurality of preset time points in a preset time period, and the hydraulic pressure value and the flow velocity value of the seawater at the preset time points; the data time sequence arrangement unit is used for arranging the air pressure values and the flow velocity values of the ship tail gas at a plurality of preset time points and the hydraulic pressure values and the flow velocity values of the seawater at a plurality of preset time points into an air pressure value time sequence input vector, an air flow velocity value time sequence input vector, a seawater pressure value time sequence vector and a seawater flow velocity value time sequence input vector according to the time dimension respectively; the pressure time sequence change feature extraction unit is used for respectively passing the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector through a multi-scale neighborhood feature extraction module to obtain a gas phase pressure time sequence feature vector and a liquid phase pressure time sequence feature vector; a pressure difference calculation unit configured to calculate a pressure difference feature vector between the gas phase pressure time series feature vector and the liquid phase pressure time series feature vector; the flow rate association coding unit is used for carrying out association coding on the gas flow rate value time sequence input vector and the seawater flow rate value time sequence input vector so as to obtain a gas-liquid flow rate synergy matrix; the flow velocity time sequence change feature extraction unit is used for enabling the gas-liquid flow velocity coordination matrix to pass through a convolutional neural network model serving as a filter so as to obtain a gas-liquid flow velocity coordination feature vector; a responsiveness correlation unit, configured to calculate responsiveness estimation of the gas-liquid flow rate cooperative feature vector relative to the differential pressure feature vector to obtain a classification feature matrix; and the differential pressure control unit is used for passing the classification characteristic matrix through a classifier to obtain a classification result, wherein the classification result is used for indicating that the differential pressure between the seawater side liquid phase pressure and the ship tail gas side gas phase pressure is increased or reduced.
In the above marine tail gas desulfurization system, the multi-scale neighborhood feature extraction module includes: the device comprises a first convolution layer, a second convolution layer parallel to the first convolution layer and a multi-scale feature fusion layer connected with the first convolution layer and the second convolution layer, wherein the first convolution layer uses a one-dimensional convolution kernel with a first length, and the second convolution layer uses a one-dimensional convolution kernel with a second length.
In the above marine vessel exhaust gas desulfurization system, the pressure time series variation feature extraction unit includes: a first neighborhood scale feature extraction subunit, configured to input the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector into a first convolution layer of the multi-scale neighborhood feature extraction module respectively to obtain the first neighborhood scale gas phase pressure time sequence feature vector and the first neighborhood scale liquid phase pressure time sequence feature vector, where the first convolution layer has a first one-dimensional convolution kernel with a first length; a second neighborhood scale feature extraction subunit, configured to input the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector into a second convolution layer of the multi-scale neighborhood feature extraction module respectively to obtain the second neighborhood scale gas phase pressure time sequence feature vector and the second neighborhood scale liquid phase pressure time sequence feature vector, where the second convolution layer has a second one-dimensional convolution kernel with a second length, and the first length is different from the second length; and the multi-scale cascading subunit is configured to cascade the first neighborhood-scale gas-phase pressure time sequence feature vector and the first neighborhood-scale liquid-phase pressure time sequence feature vector with the second neighborhood-scale gas-phase pressure time sequence feature vector and the second neighborhood-scale liquid-phase pressure time sequence feature vector to obtain the gas-phase pressure time sequence feature vector and the liquid-phase pressure time sequence feature vector. Wherein the first neighborhood scale feature extraction subunit is configured to: using a first convolution layer of the multi-scale neighborhood feature extraction module to respectively perform one-dimensional convolution coding on the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector according to the following one-dimensional convolution formula so as to obtain a first neighborhood scale gas phase pressure time sequence feature vector and a first neighborhood scale liquid phase pressure time sequence feature vector; wherein, the formula is:
Figure SMS_1
Wherein,,ais the first convolution kernelxWidth in the direction,
Figure SMS_2
For the first convolution kernel parameter vector, +.>
Figure SMS_3
For a local vector matrix that operates with a convolution kernel,wfor the size of the first one-dimensional convolution kernel,Xrepresenting the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector, +.>
Figure SMS_4
Representing that one-dimensional convolution coding is respectively carried out on the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector; the second neighborhood scale feature extraction subunit is configured to: using a second convolution layer of the multi-scale neighborhood feature extraction module to respectively perform one-dimensional convolution coding on the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector according to the following one-dimensional convolution formula so as to obtain a second neighborhood scale gas phase pressure time sequence feature vector and a second neighborhood scale liquid phase pressure time sequence feature vector; wherein, the formula is:
Figure SMS_5
wherein b is the second convolution kernelxWidth in the direction,
Figure SMS_6
For a second convolution kernel parameter vector, +.>
Figure SMS_7
For the local vector matrix to operate with the convolution kernel function, m is the size of the second one-dimensional convolution kernel,Xrepresenting the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector, +. >
Figure SMS_8
And respectively carrying out one-dimensional convolution coding on the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector.
In the above marine vessel exhaust gas desulfurization system, the flow velocity time sequence variation feature extraction unit is configured to: each layer of the convolutional neural network model used as the filter performs the following steps on input data in forward transfer of the layer: 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; wherein the output of the last layer of the convolutional neural network as a filter is the gas-liquid flow velocity cooperative eigenvector, and the input of the first layer of the convolutional neural network as a filter is the gas-liquid flow velocity cooperative matrix.
In the above marine vessel exhaust gas desulfurization system, the responsiveness correlation unit is configured to: calculating a responsiveness estimate of the gas-liquid flow rate synergy eigenvector relative to the pressure differential eigenvector to obtain a classification eigenvector by the formula; wherein, the formula is:
Figure SMS_9
wherein- >
Figure SMS_10
Representing the gas-liquid flow rate synergy characteristicsVector (S)>
Figure SMS_11
Representing the differential pressure characteristic vector,/->
Figure SMS_12
Representing the classification feature matrix.
The marine tail gas desulfurization system further comprises a training unit for training the multi-scale neighborhood feature extraction module, the convolutional neural network model serving as a filter and the classifier.
In the above marine vessel exhaust gas desulfurization system, the training unit includes: the system comprises a training data acquisition unit, a control unit and a control unit, wherein the training data acquisition unit is used for acquiring training data, the training data comprises training air pressure values and training flow rate values of ship tail gas at a plurality of preset time points in a preset time period, training hydraulic pressure values and training flow rate values of seawater at the preset time points, and a true value that the pressure difference between the seawater side liquid phase pressure and the ship tail gas side gas phase pressure is required to be increased or reduced; the training data time sequence arrangement unit is used for arranging the training air pressure values and the training flow velocity values of the ship tail gas at a plurality of preset time points and arranging the training hydraulic pressure values and the training flow velocity values of the seawater at a plurality of preset time points into a training air pressure value time sequence input vector, a training air flow velocity value time sequence input vector, a training seawater pressure value time sequence vector and a training seawater flow velocity value time sequence input vector according to time dimensions respectively; the training pressure time sequence change feature extraction unit is used for enabling the training gas pressure value time sequence input vector and the training seawater pressure value time sequence vector to respectively pass through the multi-scale neighborhood feature extraction module so as to obtain a training gas phase pressure time sequence feature vector and a training liquid phase pressure time sequence feature vector; the training pressure difference calculation unit is used for calculating a training pressure difference characteristic vector between the training gas phase pressure time sequence characteristic vector and the training liquid phase pressure time sequence characteristic vector; the training flow rate association coding unit is used for carrying out association coding on the training gas flow rate value time sequence input vector and the training seawater flow rate value time sequence input vector so as to obtain a training gas-liquid flow rate cooperative matrix; the training flow velocity time sequence change feature extraction unit is used for enabling the training gas-liquid flow velocity cooperative matrix to pass through the convolutional neural network model serving as a filter so as to obtain a training gas-liquid flow velocity cooperative feature vector; the training response association unit is used for calculating the response estimation of the training gas-liquid flow rate cooperative feature vector relative to the training differential pressure feature vector so as to obtain a training classification feature matrix; the feature optimization unit is used for performing feature redundancy optimization on the training classification feature matrix based on the stacking of the low-cost bottleneck mechanism so as to obtain an optimized training classification feature matrix; the classification loss unit is used for enabling the optimized training classification characteristic matrix to pass through the classifier to obtain a classification loss function value; and the model training unit is used for training the multi-scale neighborhood feature extraction module, the convolution neural network model serving as the filter and the classifier based on the classification loss function value and through the gradient descending direction ship.
In the above marine exhaust gas desulfurization system, the feature optimization unit is configured to: performing feature redundancy optimization on the training classification feature matrix based on low-cost bottleneck mechanism stacking by using the following optimization formula to obtain the optimized training classification feature matrix; wherein, the optimization formula is:
Figure SMS_13
Figure SMS_14
Figure SMS_15
wherein,,
Figure SMS_18
classifying a feature matrix for said training, +.>
Figure SMS_20
Representing a single layer rollAccumulation and manipulation of->
Figure SMS_22
、/>
Figure SMS_17
And->
Figure SMS_19
Respectively represent the position-by-position addition, subtraction and multiplication of the feature matrix, and +.>
Figure SMS_21
And->
Figure SMS_23
For biasing the feature matrix +.>
Figure SMS_16
And (5) optimizing the multi-scale associated feature map.
According to another aspect of the present application, there is provided a marine vessel exhaust gas desulfurization process comprising:
collecting tail gas and seawater of a sulfur-containing ship;
pretreating the tail gas of the sulfur-containing ship to obtain tail gas after dust removal and oil removal;
pretreating the seawater to obtain pretreated seawater;
introducing the tail gas after dust removal and oil removal and the pretreated seawater into a membrane contactor to carry out marine tail gas desulfurization so as to obtain seawater after sulfur dioxide absorption;
and (3) oxidizing the seawater after absorbing sulfur dioxide to obtain oxidized sulfate.
According to still another aspect of the present application, there is provided an electronic apparatus including: a processor; and a memory having stored therein computer program instructions that, when executed by the processor, cause the processor to perform a marine vessel exhaust gas desulfurization process as described above.
According to a further aspect of the present application there is provided a computer readable medium having stored thereon computer program instructions which, when executed by a processor, cause the processor to perform a marine vessel exhaust gas desulphurisation process as described above.
Compared with the prior art, the ship tail gas desulfurization process and the system thereof provided by the application have the advantages that through the artificial intelligence technology of deep learning, the differential pressure time sequence dynamic change characteristic information of the ship tail gas pressure value and the seawater hydraulic pressure value and the time sequence change association characteristic information between the ship tail gas flow velocity value and the seawater flow velocity value are excavated, so that the differential pressure time sequence change characteristic is enhanced based on the flow velocity time sequence association characteristic of the ship tail gas and the seawater, the expression of the differential pressure time sequence change characteristic is optimized, the differential pressure control accuracy between the seawater side liquid phase pressure and the ship tail gas side gas phase pressure is improved, and the desulfurization effect and efficiency of the ship tail gas are optimized.
Drawings
The foregoing and other objects, features and advantages of the present application will become more apparent from the following more particular description of embodiments of the present application, as illustrated in the accompanying 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 not constitute a limitation to the application. In the drawings, like reference numerals generally refer to like parts or steps.
FIG. 1 is a schematic view of a marine exhaust gas desulfurization system according to an embodiment of the present application;
FIG. 2 is a block diagram of a marine vessel exhaust gas desulfurization system according to an embodiment of the present application;
FIG. 3 is a block diagram of a membrane contactor desulfurization module in a marine tail gas desulfurization system according to an embodiment of the present application;
FIG. 4 is a block diagram of a training unit in a marine tail gas desulfurization system according to an embodiment of the present application;
FIG. 5 is a system architecture diagram of an inference unit in a marine exhaust gas desulfurization system according to an embodiment of the present application;
FIG. 6 is a system architecture diagram of a training unit in a marine exhaust gas desulfurization system according to an embodiment of the present application;
FIG. 7 is a block diagram of a pressure timing variation feature extraction unit in a marine exhaust gas desulfurization system according to an embodiment of the present application;
FIG. 8 is a flow chart of convolutional neural network coding in a marine tail gas desulfurization system according to an embodiment of the present application;
FIG. 9 is a flow chart of a marine vessel tail gas desulfurization process according to an embodiment of the present application;
fig. 10 is a block diagram of an electronic device according to an embodiment of the present application.
Detailed Description
Hereinafter, example 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 of the embodiments of the present application and not all of the embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
Scene overview
As described above, in the process of desulfurizing marine exhaust gas by using membrane contactor technology, it has been found that, because marine diesel fuel is usually not completely combusted, small solid particles and oil droplets are contained therein, which inevitably pollute the membrane pores and affect SO 2 Diffusion and mass transfer. And absorb SO 2 The sulfite is oxidized into sulfate before the post-seawater is discharged, and the conventional aeration method has low efficiency and long treatment time, and influences the practicability of the whole system. Accordingly, an optimized marine exhaust gas desulfurization system is desired.
Specifically, in the technical scheme of this application, propose a boats and ships tail gas desulfurization system, it includes: and collecting the tail gas and seawater of the sulfur-containing ship. And the sulfur-containing ship tail gas pretreatment module is used for pretreating the sulfur-containing ship tail gas to obtain tail gas after dust removal and oil removal. It should be understood that in the process of actually collecting the tail gas of the sulfur-containing ship, solid particles and liquid oil drops with incomplete combustion exist in the tail gas of the sulfur-containing ship, which may cause pollution and blockage of the hydrophobic microporous membrane in the subsequent membrane contactor, so that pretreatment of the tail gas is required. The seawater is pretreated to obtain pretreated seawater. It will be appreciated that there will be suspended matter in the seawater, which is to be filtered before the marine exhaust gas desulfurization can be performed. And introducing the tail gas after dust removal and oil removal and the pretreated seawater into a membrane contactor to carry out marine tail gas desulfurization so as to obtain seawater after sulfur dioxide absorption. In particular, the membrane contactor is a membrane contactor filled with polytetrafluoroethylene hollow fiber hydrophobic microporous membranes, which can provide a very large volume specific surface area and gas-liquid mass transfer efficiency. And (3) carrying out oxidation treatment on the seawater after absorbing sulfur dioxide so as to oxidize sulfite into sulfate, thereby avoiding pollution to the environment.
Accordingly, in consideration of the fact that in the process of desulfurizing the tail gas of the ship using the membrane contactor technology, in the membrane contactor desulfurization module, the key to ensuring the efficiency and effect of desulfurizing the tail gas of the ship is to control the pressure difference between the liquid phase pressure of the sea water side and the gas phase pressure of the tail gas side of the ship. Further, it is considered that the differential pressure between the sea water side liquid phase pressure and the ship exhaust gas side gas phase pressure is related to the differential time series change characteristic between the ship exhaust gas pressure value and the sea water hydraulic pressure value, but the time series change of the differential pressure is weak and is easily disturbed by the environmental noise, so if the time series change characteristic optimization of the differential pressure can be performed by the flow velocity value time series change related characteristic information of the differential pressure, the control accuracy of the differential pressure can be obviously enhanced. In the process, the difficulty is how to excavate the differential pressure time sequence dynamic change characteristic information of the ship tail gas pressure value and the seawater hydraulic pressure value and the time sequence change association characteristic information between the ship tail gas flow velocity value and the seawater flow velocity value, so that the differential pressure time sequence change characteristic is enhanced based on the flow velocity time sequence association characteristic of the ship tail gas and the seawater, the expression of the differential pressure time sequence change characteristic is optimized, the differential pressure control precision between the seawater side liquid phase pressure and the ship tail gas side gas phase pressure is improved, and the desulfurization effect and efficiency of the ship tail gas are optimized.
In recent years, deep learning and neural networks have been widely used in the fields of computer vision, natural language processing, text signal processing, and the like. The deep learning and the development of the neural network provide new solutions and schemes for mining differential pressure time sequence dynamic change characteristic information of the ship tail gas pressure value and the seawater hydraulic pressure value and time sequence change association characteristic information between the ship tail gas flow velocity value and the seawater flow velocity value.
Specifically, in the technical scheme of the application, firstly, the air pressure value and the flow velocity value of the ship tail gas at a plurality of preset time points in a preset time period and the hydraulic pressure value and the flow velocity value of the seawater at the preset time points are obtained. Next, considering that the pressure value and the flow rate value of the ship tail gas and the hydraulic value and the flow rate value of the sea water have respective dynamic change rules in the time dimension, in order to fully extract time sequence dynamic change characteristic information of the data, in the technical scheme of the application, the pressure value and the flow rate value of the ship tail gas at a plurality of preset time points and the hydraulic value and the flow rate value of the sea water at a plurality of preset time points are respectively arranged into a gas pressure value time sequence input vector, a gas flow rate value time sequence input vector, a sea water pressure value time sequence vector and a sea water flow rate value time sequence input vector according to the time dimension, so as to integrate the pressure value and the flow rate value of the ship tail gas and the distribution information of the hydraulic value and the flow rate value of the sea water in time sequence.
Then, it is considered that the pneumatic pressure value of the ship tail gas and the hydraulic pressure value of the seawater have fluctuation and uncertainty in the time dimension, so that the ship tail gas and the seawater have different dynamic change characteristics in different time period spans in time sequence. Therefore, in the technical scheme of the application, the gas pressure value time sequence input vector and the seawater pressure value time sequence vector are respectively subjected to feature extraction through a multi-scale neighborhood feature extraction module so as to respectively extract dynamic multi-scale neighborhood associated features of the gas pressure value of the ship tail gas and the hydraulic value of the seawater under different spans, thereby obtaining a gas pressure time sequence feature vector and a liquid pressure time sequence feature vector. And then, calculating a pressure difference characteristic vector between the gas phase pressure time sequence characteristic vector and the liquid phase pressure time sequence characteristic vector, namely, a pressure difference time sequence change characteristic of the air pressure value of the ship tail gas and the hydraulic pressure value of the seawater in a time dimension.
Further, regarding the flow velocity value of the ship tail gas and the flow velocity value of the seawater, the fact that the two flow velocity values have a cooperative implicit association relationship in the time dimension is considered, and the influence on the desulfurization of the ship tail gas is achieved. Therefore, in order to optimize the expression of the differential pressure time sequence variation characteristic based on the time sequence cooperative correlation characteristic of the flow velocity value of the ship tail gas and the flow velocity value of the seawater, so as to improve the efficiency and the effect of the desulfurization of the ship tail gas, in the technical scheme of the application, the gas flow velocity time sequence input vector and the seawater flow velocity time sequence input vector are further subjected to the correlation coding to obtain a gas-liquid flow velocity cooperative matrix, so that the correlation relation between the flow velocity value of the ship tail gas and the flow velocity value of the seawater in time sequence is established. And then, further performing feature mining of the gas-liquid flow velocity synergy matrix by using a convolution neural network model which is taken as a filter and has excellent performance in the aspect of local implicit correlation feature extraction so as to extract time sequence synergistic dynamic correlation feature information of the flow velocity value of the ship tail gas and the flow velocity value of the seawater in the time dimension, thereby obtaining a gas-liquid flow velocity synergistic feature vector.
Then, a response estimation of the gas-liquid flow rate cooperative feature vector with respect to the differential pressure feature vector is calculated to represent correlation feature distribution information between differential pressure time sequence dynamic multi-scale correlation features of the ship exhaust gas pressure value and the sea water hydraulic pressure value and time sequence cooperative correlation features of the ship exhaust gas flow rate value and the sea water flow rate value, that is, feature enhancement of differential pressure time sequence dynamic correlation features of the ship exhaust gas pressure value and the sea water hydraulic pressure value is performed with the time sequence cooperative correlation features of the ship exhaust gas flow rate value and the sea water flow rate value to perform an optimized expression of time sequence variation features of the differential pressure, thereby obtaining a classification feature matrix.
And further, classifying the classification characteristic matrix in a classifier to obtain a classification result for indicating that the pressure difference between the liquid phase pressure on the sea water side and the gas phase pressure on the ship tail gas side should be increased or decreased. That is, in the technical solution of the present application, the labels of the classifier include that the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be increased (first label), and that the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be decreased (second label), wherein the classifier determines to which classification label the classification feature matrix belongs by a soft maximum function. It is noted that the first tag p1 and the second tag p2 do not contain the concept of artificial setting, and in fact, during the training process, the computer model does not have the concept of "the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be increased or decreased", which is only two kinds of classification tags and the probability that the output characteristics are under the two classification tags, i.e. the sum of p1 and p2 is one. Therefore, the classification result that the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be increased or decreased is actually converted into the classification probability distribution conforming to the natural law through the classification label, and the physical meaning of the natural probability distribution of the label is essentially used instead of the language text meaning of "the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be increased or decreased". It should be understood that, in the technical solution of the present application, the classification label of the classifier is a control strategy label that the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be increased or decreased, so after the classification result is obtained, the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure at the current time point can be adaptively adjusted based on the classification result, so as to optimize the desulfurization effect and efficiency of the ship tail gas.
In particular, in the technical solution of the present application, when the classification feature matrix is obtained by calculating the estimation of the responsiveness of the gas-liquid flow velocity synergy feature vector with respect to the differential pressure feature vector, the channel dimension correlation feature of the convolutional neural network model of the cross-time domain correlation coding value of the gas flow velocity value and the seawater flow velocity value is considered by the gas-liquid flow velocity synergy feature vector, and the differential pressure feature vector expresses the position-by-position multi-scale time sequence neighborhood correlation feature difference value of the gas phase pressure value and the liquid phase pressure value, which corresponds to different distribution dimensions of the feature, so that the responsiveness between different dimension distributions can be fully expressed by using the feature responsiveness of the classification feature matrix in the cross dimension. On the other hand, considering that the feature distribution in each dimension is not completely orthogonal, redundant features are inevitably present in the classification feature matrix, so that the classification efficiency of the classification feature matrix through the classifier is affected, that is, the training speed of the model is reduced.
Thus, the applicant of the present application performs training on the matrix of classification features, e.g., denoted as
Figure SMS_24
Feature redundancy optimization based on low-cost bottleneck-mechanism stacking is performed to obtain an optimized classified feature matrix, for example, marked as +. >
Figure SMS_25
The method is specifically expressed as follows:
Figure SMS_26
Figure SMS_27
Figure SMS_28
Figure SMS_30
representing a single layer convolution operation,/->
Figure SMS_35
、/>
Figure SMS_36
And->
Figure SMS_31
Respectively represent the position-by-position addition, subtraction and multiplication of the feature matrix, and +.>
Figure SMS_32
And->
Figure SMS_33
For biasing the feature matrix, for example, a global mean feature matrix or a unit feature matrix of the classification feature matrix can be initially set, wherein the initial bias feature matrix +.>
Figure SMS_34
And->
Figure SMS_29
Different.
Here, the feature redundancy optimization based on the low-cost bottleneck-mechanism stacking can use the low-cost bottleneck mechanism of the multiply-add stacking of two low-cost transformation features to perform feature expansion, and match a residual path by biasing stacking channels with uniform values, so that hidden distribution information under intrinsic features is revealed in redundancy features through low-cost operation transformation similar to a basic residual module, and a more intrinsic expression of the features is obtained through a simple and effective convolution operation architecture, thereby optimizing the redundant feature expression of the classification feature matrix, improving the classification efficiency of the classification feature matrix through a classifier, namely improving the training speed of a model. Therefore, the self-adaptive control of the pressure difference between the liquid phase pressure at the sea water side and the gas phase pressure at the propagating tail gas side can be accurately performed in real time based on the flow velocity change condition of the ship tail gas and the sea water, so that the desulfurization effect and efficiency of the ship tail gas are optimized.
Based on this, this application proposes a boats and ships tail gas desulfurization system, it includes: the ship tail gas and sea water collecting module is used for collecting the sulfur-containing ship tail gas and sea water; the sulfur-containing ship tail gas pretreatment module is used for pretreating the sulfur-containing ship tail gas to obtain tail gas after dust removal and oil removal; the seawater pretreatment module is used for pretreating the seawater to obtain pretreated seawater; the membrane contactor desulfurization module is used for introducing the tail gas after dust removal and oil removal and the pretreated seawater into a membrane contactor to carry out marine tail gas desulfurization so as to obtain seawater after sulfur dioxide absorption; and the seawater post-treatment module is used for carrying out oxidation treatment on the seawater absorbing sulfur dioxide to obtain oxidized sulfate.
Fig. 1 is an application scenario diagram of a marine tail gas desulfurization system according to an embodiment of the present application. As shown in fig. 1, in this application scenario, the air pressure values of the marine exhaust gas at a plurality of predetermined time points in a predetermined period are acquired by an air pressure sensor (e.g., V1 as illustrated in fig. 1), the flow velocity values of the marine exhaust gas at a plurality of predetermined time points in the predetermined period are acquired by a flow velocity sensor (e.g., V2 as illustrated in fig. 1), and the flow velocity values of the seawater at the plurality of predetermined time points are acquired by a flow velocity sensor (e.g., V4 as illustrated in fig. 1), and the hydraulic pressure values of the seawater at the plurality of predetermined time points are acquired by a hydraulic pressure sensor (e.g., V3 as illustrated in fig. 1). The pretreated tail gas and pretreated seawater are then passed into a membrane contactor (e.g., K as illustrated in fig. 1) for desulfurization of marine tail gas. Then, the above-mentioned air pressure values and flow rate values of the marine exhaust gas at a plurality of predetermined time points and the hydraulic pressure values and flow rate values of the seawater at the plurality of predetermined time points are input to a server (e.g., S in fig. 1) deployed with a marine exhaust gas desulfurization algorithm, wherein the server is capable of processing the above-mentioned input data with the marine exhaust gas desulfurization algorithm to generate a classification result indicating that the pressure difference between the seawater side liquid phase pressure and the marine exhaust gas side gas phase pressure should be increased or decreased.
Having described the basic principles of the present application, various non-limiting embodiments of the present application will now be described in detail with reference to the accompanying drawings.
Exemplary System
Fig. 2 is a block diagram of a marine vessel exhaust gas desulfurization system according to an embodiment of the present application. As shown in fig. 2, a marine vessel exhaust gas desulfurization system 300 according to an embodiment of the present application includes: the ship tail gas and seawater collection module 310 comprises a sulfur-containing ship tail gas pretreatment module 320; a seawater pretreatment module 330; a membrane contactor desulfurization module 340; a seawater aftertreatment module 350.
Wherein, the ship tail gas and sea water collection module 310 is used for collecting sulfur-containing ship tail gas and sea water; the sulfur-containing ship tail gas pretreatment module 320 is configured to pretreat the sulfur-containing ship tail gas to obtain tail gas after dust removal and oil removal; the seawater pretreatment module 330 is configured to pretreat the seawater to obtain pretreated seawater; the membrane contactor desulfurization module 340 is configured to introduce the tail gas after dust removal and oil removal and the pretreated seawater into a membrane contactor to perform desulfurization on the ship tail gas, so as to obtain seawater after absorbing sulfur dioxide; the seawater post-treatment module 350 is configured to perform oxidation treatment on the seawater after absorbing sulfur dioxide to obtain oxidized sulfate.
Specifically, during operation of the marine vessel exhaust gas desulfurization system 300, the marine vessel exhaust gas and seawater collection module 310 is configured to collect sulfur-containing marine vessel exhaust gas and seawater. It should be understood that in the technical solution of the present application, the desulfurization of marine exhaust gas is accomplished using a membrane contactor technology, which implements a membrane process of inter-phase mass transfer by indirect contact of gas and liquid phases. In particular, microporous membranes act as only one interface between the two phases, separating the marine tail gas phase from the absorbent seawater phase. SO in the gas phase 2 Can pass through the contact interface of the membrane and enter the absorbent phase to be taken away, thus achieving the purpose of desulfurization. Therefore, first, sulfur-containing marine tail gas and seawater are collected.
Specifically, during the operation of the marine exhaust gas desulfurization system 300, the sulfur-containing marine exhaust gas pretreatment module 320 and the seawater pretreatment module 330 are configured to pretreat the sulfur-containing marine exhaust gas to obtain a tail gas after dust removal and oil removal, and pretreat the sulfur-containing marine exhaust gas to obtain a tail gas after dust removal and oil removal. It should be understood that, on the one hand, in the process of actually collecting the tail gas of the sulfur-containing ship, solid particles and incompletely burned liquid oil droplets are present in the tail gas of the sulfur-containing ship, which may cause pollution and blockage of the hydrophobic microporous membrane in the subsequent membrane contactor, and thus, pretreatment of the tail gas is required. On the other hand, suspended substances exist in the seawater, and the suspended substances in the seawater need to be filtered out before the desulfurization of the tail gas of the ship is performed, so that the seawater also needs to be pretreated.
Specifically, during the operation of the marine tail gas desulfurization system 300, the membrane contactor desulfurization module 340 and the seawater aftertreatment module 350 are configured to introduce the tail gas after dust removal and oil removal and the pretreated seawater into a membrane contactor to perform marine tail gas desulfurization, so as to obtain seawater after absorbing sulfur dioxide, and further, perform oxidation treatment on the seawater after absorbing sulfur dioxide to obtain oxidized sulfate. That is, after the tail gas after dust removal and oil removal and the pretreated seawater are obtained, the tail gas after dust removal and oil removal and the pretreated seawater are introduced into a membrane contactor to carry out desulfurization on the tail gas of the ship so as to obtain the seawater after sulfur dioxide absorption. In particular, the membrane contactor is a membrane contactor filled with polytetrafluoroethylene hollow fiber hydrophobic microporous membranes, which can provide a very large volume specific surface area and gas-liquid mass transfer efficiency. And (3) carrying out oxidation treatment on the seawater after absorbing sulfur dioxide so as to oxidize sulfite into sulfate, thereby avoiding pollution to the environment.
Fig. 3 is a block diagram of a membrane contactor desulfurization module in a marine exhaust gas desulfurization system according to an embodiment of the present application. As shown in fig. 3, the membrane contactor desulfurization module 340 includes: a data acquisition unit 341, configured to acquire air pressure values and flow velocity values of the ship tail gas at a plurality of predetermined time points in a predetermined time period, and hydraulic pressure values and flow velocity values of the seawater at the plurality of predetermined time points; a data timing arrangement unit 342, configured to arrange the air pressure values and the flow velocity values of the ship tail gas at the plurality of predetermined time points, and the hydraulic pressure values and the flow velocity values of the seawater at the plurality of predetermined time points into an air pressure value timing input vector, an air flow velocity value timing input vector, a seawater pressure value timing vector, and a seawater flow velocity value timing input vector according to a time dimension, respectively; the pressure time sequence change feature extraction unit 343 is configured to pass the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector through a multi-scale neighborhood feature extraction module to obtain a gas phase pressure time sequence feature vector and a liquid phase pressure time sequence feature vector; a pressure difference calculation unit 344 for calculating a pressure difference feature vector between the gas phase pressure time series feature vector and the liquid phase pressure time series feature vector; a flow rate correlation encoding unit 345, configured to perform correlation encoding on the gas flow rate value time sequence input vector and the seawater flow rate value time sequence input vector to obtain a gas-liquid flow rate synergy matrix; a flow velocity time-series variation feature extraction unit 346 for passing the gas-liquid flow velocity synergy matrix through a convolutional neural network model as a filter to obtain a gas-liquid flow velocity synergy feature vector; a responsiveness correlation unit 347, configured to calculate a responsiveness estimate of the gas-liquid flow velocity cooperative eigenvector relative to the differential pressure eigenvector to obtain a classification eigenvector; and a differential pressure control unit 348 for passing the classification feature matrix through a classifier to obtain a classification result, wherein the classification result is used for indicating that the differential pressure between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be increased or decreased.
Fig. 5 is a system architecture diagram of an inference unit in a marine exhaust gas desulfurization system according to an embodiment of the present application. As shown in fig. 5, in the system architecture of the marine vessel exhaust gas desulfurization system 300, in the process of inference, the data acquisition unit 341 first acquires the air pressure values and the flow velocity values of the marine vessel exhaust gas at a plurality of predetermined time points in a predetermined period, and the hydraulic pressure values and the flow velocity values of the seawater at the plurality of predetermined time points; next, the data timing arrangement unit 342 arranges the air pressure values and the flow velocity values of the ship tail gas at a plurality of predetermined time points acquired by the data acquisition unit 341, and the hydraulic pressure values and the flow velocity values of the seawater at the plurality of predetermined time points into an air pressure value timing input vector, an air flow velocity value timing input vector, a seawater pressure value timing vector and a seawater flow velocity value timing input vector according to a time dimension, respectively; the pressure time sequence variation feature extraction unit 343 respectively passes the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector obtained by the data time sequence arrangement unit 342 through a multi-scale neighborhood feature extraction module to obtain a gas phase pressure time sequence feature vector and a liquid phase pressure time sequence feature vector; the pressure difference calculation unit 344 calculates a pressure difference feature vector between the gas phase pressure time series feature vector and the liquid phase pressure time series feature vector obtained by the pressure time series change feature extraction unit 343; then, the flow rate correlation encoding unit 345 performs correlation encoding on the gas flow rate value time sequence input vector and the seawater flow rate value time sequence input vector obtained by the data time sequence arrangement unit 342 to obtain a gas-liquid flow rate synergy matrix; the flow velocity time sequence variation feature extraction unit 346 obtains a gas-liquid flow velocity synergy feature vector by passing the gas-liquid flow velocity synergy matrix obtained by the flow velocity correlation encoding unit 345 through a convolutional neural network model serving as a filter; the responsiveness correlation unit 347 calculates a responsiveness estimate of the gas-liquid flow rate synergy feature vector obtained by the flow rate time-series variation feature extraction unit 346 with respect to the differential pressure feature vector obtained by the differential pressure calculation unit 344 to obtain a classification feature matrix; further, the differential pressure control unit 348 passes the classification feature matrix through a classifier to obtain a classification result, which is used to indicate that the differential pressure between the seawater side liquid-phase pressure and the ship tail gas side gas-phase pressure should be increased or decreased.
More specifically, during operation of the marine vessel exhaust gas desulfurization system 300, the data acquisition unit 341 is configured to acquire the air pressure value and the flow rate value of the marine vessel exhaust gas at a plurality of predetermined time points within a predetermined period of time, and the hydraulic pressure value and the flow rate value of the seawater at the plurality of predetermined time points. In consideration of the fact that in the process of carrying out marine tail gas desulfurization by using a membrane contactor technology, in the membrane contactor desulfurization module, the key for ensuring the marine tail gas desulfurization efficiency and effect is to control the pressure difference between the seawater side liquid phase pressure and the marine tail gas side gas phase pressure. Further, it is considered that the differential pressure between the sea water side liquid phase pressure and the ship exhaust gas side gas phase pressure is related to the differential time series change characteristic between the ship exhaust gas pressure value and the sea water hydraulic pressure value, but the time series change of the differential pressure is weak and is easily disturbed by the environmental noise, so if the time series change characteristic optimization of the differential pressure can be performed by the flow velocity value time series change related characteristic information of the differential pressure, the control accuracy of the differential pressure can be obviously enhanced. Thus, in one specific example of the present application, first, the air pressure values of the marine exhaust gas at a plurality of predetermined time points in the predetermined period of time may be obtained by the air pressure sensor, then the flow velocity values of the marine exhaust gas at a plurality of predetermined time points in the predetermined period of time may be obtained by the flow velocity sensor, and the flow velocity values of the seawater at the plurality of predetermined time points may be obtained by the flow velocity sensor, and then the hydraulic pressure values of the seawater at the plurality of predetermined time points may be obtained by the hydraulic pressure sensor.
More specifically, during operation of the marine vessel exhaust gas desulfurization system 300, the data timing arrangement unit 342 is configured to arrange the gas pressure values and the flow rate values of the marine vessel exhaust gas at the plurality of predetermined time points, and the hydraulic pressure values and the flow rate values of the seawater at the plurality of predetermined time points into a gas pressure value timing input vector, a gas flow rate value timing input vector, a seawater pressure value timing vector, and a seawater flow rate value timing input vector, respectively, according to a time dimension. In order to fully extract time sequence dynamic change characteristic information of the data, in the technical scheme of the application, the air pressure values and the flow velocity values of the ship tail gas at a plurality of preset time points and the sea water pressure values and the flow velocity values at the preset time points are respectively arranged into an air pressure value time sequence input vector, an air flow velocity value time sequence input vector, a sea water pressure value time sequence vector and a sea water flow velocity value time sequence input vector according to the time dimension, so that the air pressure values and the flow velocity values of the ship tail gas and the distribution information of the sea water pressure values and the flow velocity values in time sequence are respectively integrated.
More specifically, during operation of the marine exhaust gas desulfurization system 300, the pressure timing variation feature extraction unit 343 and the pressure difference calculation unit 344 are configured to pass the gas pressure value timing input vector and the seawater pressure value timing input vector through a multi-scale neighborhood feature extraction module to obtain a gas phase pressure timing feature vector and a liquid phase pressure timing feature vector, respectively, and calculate a pressure difference feature vector between the gas phase pressure timing feature vector and the liquid phase pressure timing feature vector. It is considered that the pneumatic pressure value of the ship tail gas and the hydraulic pressure value of the seawater have fluctuation and uncertainty in the time dimension, so that the ship tail gas and the seawater have different dynamic change characteristics under different time period spans in time sequence. Therefore, in the technical scheme of the application, the gas pressure value time sequence input vector and the seawater pressure value time sequence vector are respectively subjected to feature extraction through a multi-scale neighborhood feature extraction module so as to respectively extract dynamic multi-scale neighborhood associated features of the gas pressure value of the ship tail gas and the hydraulic value of the seawater under different spans, thereby obtaining a gas pressure time sequence feature vector and a liquid pressure time sequence feature vector. Wherein, the multiscale neighborhood feature extraction module comprises: the device comprises a first convolution layer, a second convolution layer parallel to the first convolution layer and a multi-scale feature fusion layer connected with the first convolution layer and the second convolution layer, wherein the first convolution layer uses a one-dimensional convolution kernel with a first length, and the second convolution layer uses a one-dimensional convolution kernel with a second length. And then, calculating a pressure difference characteristic vector between the gas phase pressure time sequence characteristic vector and the liquid phase pressure time sequence characteristic vector, namely, a pressure difference time sequence change characteristic of the air pressure value of the ship tail gas and the hydraulic pressure value of the seawater in a time dimension.
Fig. 7 is a block diagram of a pressure time series variation feature extraction unit in a marine exhaust gas desulfurization system according to an embodiment of the present application. As shown in fig. 7, the pressure time series variation feature extraction unit 343 includes: a first neighborhood scale feature extraction subunit 3431 configured to input the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector into a first convolution layer of the multi-scale neighborhood feature extraction module, respectively, to obtain the first neighborhood scale gas phase pressure time sequence feature vector and the first neighborhood scale liquid phase pressure time sequence feature vector, where the first convolution layer has a first one-dimensional convolution kernel with a first length; a second neighborhood scale feature extraction subunit 3432 configured to input the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector into a second convolution layer of the multi-scale neighborhood feature extraction module, respectively, to obtain the second neighborhood scale gas phase pressure time sequence feature vector and the second neighborhood scale liquid phase pressure time sequence feature vector, where the second convolution layer has a second one-dimensional convolution kernel with a second length, and the first length is different from the second length; and a multi-scale cascading subunit 3433, configured to cascade the first neighborhood-scale gas-phase pressure time-sequence feature vector and the first neighborhood-scale liquid-phase pressure time-sequence feature vector with the second neighborhood-scale gas-phase pressure time-sequence feature vector and the second neighborhood-scale liquid-phase pressure time-sequence feature vector, respectively, so as to obtain the gas-phase pressure time-sequence feature vector and the liquid-phase pressure time-sequence feature vector. Wherein the first neighborhood scale feature extraction subunit 3431 is configured to: using a first convolution layer of the multi-scale neighborhood feature extraction module to respectively perform one-dimensional convolution coding on the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector according to the following one-dimensional convolution formula so as to obtain a first neighborhood scale gas phase pressure time sequence feature vector and a first neighborhood scale liquid phase pressure time sequence feature vector; wherein, the formula is:
Figure SMS_37
Wherein,,ais the first convolution kernelxWidth in the direction,
Figure SMS_38
For the first convolution kernel parameter vector, +.>
Figure SMS_39
For a local vector matrix that operates with a convolution kernel,wfor the size of the first one-dimensional convolution kernel,Xrepresenting the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector, +.>
Figure SMS_40
Representing that one-dimensional convolution coding is respectively carried out on the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector; the method comprises the steps of,the second neighborhood scale feature extraction subunit 3432 is configured to: using a second convolution layer of the multi-scale neighborhood feature extraction module to respectively perform one-dimensional convolution coding on the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector according to the following one-dimensional convolution formula so as to obtain a second neighborhood scale gas phase pressure time sequence feature vector and a second neighborhood scale liquid phase pressure time sequence feature vector; wherein, the formula is:
Figure SMS_41
wherein b is the second convolution kernelxWidth in the direction,
Figure SMS_42
For a second convolution kernel parameter vector, +.>
Figure SMS_43
For the local vector matrix to operate with the convolution kernel function, m is the size of the second one-dimensional convolution kernel,Xrepresenting the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector, +. >
Figure SMS_44
And respectively carrying out one-dimensional convolution coding on the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector.
More specifically, during operation of the marine vessel exhaust gas desulfurization system 300, the flow rate correlation encoding unit 345 is configured to perform correlation encoding on the gas flow rate value time sequence input vector and the seawater flow rate value time sequence input vector to obtain a gas-liquid flow rate synergy matrix. Considering that the gas flow velocity value time sequence input vector and the seawater flow velocity value time sequence input vector have a cooperative implicit association relation in the time dimension, in the technical scheme of the application, the gas flow velocity value time sequence input vector and the seawater flow velocity value time sequence input vector are further subjected to association coding to obtain a gas-liquid flow velocity cooperative matrix so as to establish the flow of the ship tail gasAnd the correlation between the speed value and the flow velocity value of the seawater in time sequence. More specifically, in one specific example of the present application, the gas flow rate value time series input vector and the seawater flow rate value time series input vector are correlatively encoded in the following formula to obtain a gas-liquid flow rate synergy matrix; wherein, the formula is:
Figure SMS_45
Wherein->
Figure SMS_46
Time sequence input vector representing the gas flow rate value, < >>
Figure SMS_47
A transpose vector representing the time sequence input vector of the gas flow rate value,/->
Figure SMS_48
Time sequence input vector representing the seawater flow velocity value, < >>
Figure SMS_49
Representing said gas-liquid flow rate co-matrix, < >>
Figure SMS_50
Representing vector multiplication.
More specifically, during operation of the marine vessel exhaust gas desulfurization system 300, the flow velocity time-series variation feature extraction unit 346 is configured to pass the gas-liquid flow velocity synergy matrix through a convolutional neural network model as a filter to obtain a gas-liquid flow velocity synergy feature vector. That is, feature mining of the gas-liquid flow velocity synergy matrix is performed using a convolutional neural network model as a filter having excellent performance in local implicit correlation feature extraction to extract time-series collaborative dynamic correlation feature information of the flow velocity value of the ship tail gas and the flow velocity value of the seawater in a time dimension, thereby obtaining a gas-liquid flow velocity synergy feature vector. In one particular example, the convolutional neural network includes a plurality of neural network layers that are cascaded with one another, wherein each neural network layer includes a convolutional layer, a pooling layer, and an activation layer. In the coding process of the convolutional neural network, each layer of the convolutional neural network carries out convolutional processing based on a convolutional kernel on input data by using the convolutional layer in the forward transmission process of the layer, carries out pooling processing on a convolutional feature map output by the convolutional layer by using the pooling layer and carries out activation processing on the pooling feature map output by the pooling layer by using the activation layer.
Fig. 8 is a flowchart of convolutional neural network coding in a marine exhaust gas desulfurization system according to an embodiment of the present application. As shown in fig. 8, in the encoding process of the convolutional neural network, it includes: each layer of the convolutional neural network model used as the filter performs the following steps on input data in forward transfer of the layer: s310, carrying out convolution processing on input data to obtain a convolution characteristic diagram; s320, pooling the convolution feature map based on a feature matrix to obtain a pooled feature map; s330, performing nonlinear activation on the pooled feature map to obtain an activated feature map; wherein the output of the last layer of the convolutional neural network as a filter is the gas-liquid flow velocity cooperative eigenvector, and the input of the first layer of the convolutional neural network as a filter is the gas-liquid flow velocity cooperative matrix.
More specifically, during operation of the marine vessel exhaust gas desulfurization system 300, the responsiveness correlation unit 347 is configured to calculate a responsiveness estimate of the gas-liquid flow rate synergy feature vector relative to the pressure difference feature vector to obtain a classification feature matrix. That is, after the gas-liquid flow velocity cooperative characteristic vector and the differential pressure characteristic vector are obtained, a response estimation of the gas-liquid flow velocity cooperative characteristic vector with respect to the differential pressure characteristic vector is further calculated to represent correlation characteristic distribution information between differential pressure time series dynamic multi-scale correlation characteristics of the ship exhaust gas pressure value and the sea water hydraulic pressure value and time series cooperative correlation characteristics of the ship exhaust gas flow velocity value and the sea water flow velocity value, that is, the differential pressure time series dynamic correlation characteristics of the ship exhaust gas pressure value and the sea water hydraulic pressure value are performed with the time series cooperative correlation characteristics of the ship exhaust gas flow velocity value and the sea water flow velocity value And (3) carrying out the optimized expression of the time sequence change characteristics of the differential pressure so as to obtain a classification characteristic matrix. In one specific example of the present application, a responsiveness estimate of the gas-liquid flow rate synergy eigenvector relative to the pressure differential eigenvector is calculated to obtain a classification eigenvector with the following formula; wherein, the formula is:
Figure SMS_51
wherein->
Figure SMS_52
Representing the gas-liquid flow rate synergy eigenvector, < >>
Figure SMS_53
Representing the differential pressure characteristic vector,/->
Figure SMS_54
Representing the classification feature matrix.
More specifically, during operation of the marine vessel exhaust gas desulfurization system 300, the pressure difference control unit 348 is configured to pass the classification feature matrix through a classifier to obtain a classification result, where the classification result is used to indicate that the pressure difference between the seawater side liquid phase pressure and the marine vessel exhaust gas side gas phase pressure should be increased or decreased. In one technical scheme of the application, the classification feature matrix is passed through a classifier to obtain a classification result for indicating that the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be increased or decreased, and specifically, the classifier comprises a plurality of fully connected layers and a Softmax layer cascaded with the last fully connected layer of the plurality of fully connected layers. In the classification process of the classifier, the classification feature matrix is first projected as a vector, for example, in a specific example, the classification feature matrix is expanded along a row vector or a column vector to form a classification feature vector; then, performing multiple full-connection coding on the classification feature vectors by using multiple full-connection layers of the classifier to obtain coded classification feature vectors; further, the encoded classification feature vector is input to a Softmax layer of the classifier, i.e., the encoded classification feature vector is classified using the Softmax classification function to obtain a classification label. In the technical solution of the present application, the labels of the classifier include that the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be increased (first label), and that the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be decreased (second label), wherein the classifier determines to which classification label the classification feature matrix belongs through a soft maximum function. It is noted that the first tag p1 and the second tag p2 do not contain the concept of artificial setting, and in fact, during the training process, the computer model does not have the concept of "the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be increased or decreased", which is only two kinds of classification tags and the probability that the output characteristics are under the two classification tags, i.e. the sum of p1 and p2 is one. Therefore, the classification result that the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be increased or decreased is actually converted into the classification probability distribution conforming to the natural law through the classification label, and the physical meaning of the natural probability distribution of the label is essentially used instead of the language text meaning of "the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be increased or decreased". It should be understood that, in the technical solution of the present application, the classification label of the classifier is a control strategy label that the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure should be increased or decreased, so after the classification result is obtained, the pressure difference between the sea water side liquid phase pressure and the ship tail gas side gas phase pressure at the current time point can be adaptively adjusted based on the classification result, so as to optimize the desulfurization effect and efficiency of the ship tail gas.
It should be appreciated that the multi-scale neighborhood feature extraction module, the convolutional neural network model as a filter, and the classifier need to be trained prior to the inference using the neural network model described above. That is, in the marine tail gas desulfurization system of the present application, the marine tail gas desulfurization system further includes a training module configured to train the multi-scale neighborhood feature extraction module, the convolutional neural network model as a filter, and the classifier. The training of deep neural networks mostly adopts a back propagation algorithm, and the back propagation algorithm updates the parameters of the current layer through errors transmitted by the later layer by using a chained method, which can suffer from the problem of gradient disappearance or more broadly, the problem of unstable gradient when the network is deep.
Fig. 4 is a block diagram of a training unit in a marine exhaust gas desulfurization system according to an embodiment of the present application. As shown in fig. 4, the marine vessel exhaust gas desulfurization system 300 according to the embodiment of the present application further includes a training unit 400 including: a training data acquisition unit 410; a training data timing arrangement unit 420; training the pressure timing variation feature extraction unit 430; training the differential pressure calculation unit 440; training the flow rate association encoding unit 450; training the flow rate time-series variation feature extraction unit 460; training the responsiveness association unit 470; a feature optimization unit 480; and a classification loss unit 490; model training unit 500.
The training data acquisition unit 410 is configured to acquire training data, where the training data includes training air pressure values and training flow rate values of the ship tail gas at a plurality of predetermined time points in a predetermined time period, training hydraulic pressure values and training flow rate values of the seawater at the plurality of predetermined time points, and a true value that a pressure difference between the seawater side liquid phase pressure and the ship tail gas side gas phase pressure should be increased or decreased; the training data time sequence arrangement unit 420 is configured to arrange the training air pressure values and the training flow velocity values of the ship tail gas at the plurality of predetermined time points, and the training hydraulic pressure values and the training flow velocity values of the seawater at the plurality of predetermined time points into a training air pressure value time sequence input vector, a training air flow velocity value time sequence input vector, a training seawater pressure value time sequence vector and a training seawater flow velocity value time sequence input vector according to a time dimension, respectively; the training pressure time sequence variation feature extraction unit 430 is configured to pass the training gas pressure value time sequence input vector and the training seawater pressure value time sequence vector through the multi-scale neighborhood feature extraction module to obtain a training gas phase pressure time sequence feature vector and a training liquid phase pressure time sequence feature vector; the training pressure difference calculating unit 440 is configured to calculate a training pressure difference feature vector between the training gas phase pressure time sequence feature vector and the training liquid phase pressure time sequence feature vector; the training flow rate association coding unit 450 is configured to perform association coding on the training gas flow rate value time sequence input vector and the training seawater flow rate value time sequence input vector to obtain a training gas-liquid flow rate synergy matrix; the training flow velocity time sequence variation feature extraction unit 460 is configured to pass the training gas-liquid flow velocity cooperative matrix through the convolutional neural network model serving as a filter to obtain a training gas-liquid flow velocity cooperative feature vector; the training responsiveness correlation unit 470 is configured to calculate a responsiveness estimate of the training gas-liquid flow rate cooperative feature vector relative to the training pressure difference feature vector to obtain a training classification feature matrix; the feature optimization unit 480 is configured to perform feature redundancy optimization on the training classification feature matrix based on stacking of a low-cost bottleneck mechanism to obtain an optimized training classification feature matrix; and, the classification loss unit 490 is configured to pass the optimized training classification feature matrix through the classifier to obtain a classification loss function value; the model training unit 500 is configured to train the multi-scale neighborhood feature extraction module, the convolutional neural network model as a filter, and the classifier based on the classification loss function value and by gradient descent direction vessels.
Fig. 6 is a system architecture diagram of a training unit in a marine exhaust gas desulfurization system according to an embodiment of the present application. As shown in fig. 6, in the system architecture of the marine exhaust gas desulfurization system 300, in the training module 400, training data is first acquired by the training data acquisition unit 410, where the training data includes training air pressure values and training flow rate values of marine exhaust gas at a plurality of predetermined time points in a predetermined period of time, training hydraulic pressure values and training flow rate values of seawater at the plurality of predetermined time points, and a true value that a pressure difference between the seawater side liquid phase pressure and the marine exhaust gas side gas phase pressure should be increased or decreased; next, the training data timing arrangement unit 420 arranges the training air pressure values and the training flow rate values of the ship tail gas at a plurality of predetermined time points acquired by the training data acquisition unit 410, and the training hydraulic pressure values and the training flow rate values of the seawater at the plurality of predetermined time points into a training air pressure value timing input vector, a training air flow rate value timing input vector, a training seawater pressure value timing vector and a training seawater flow rate value timing input vector according to a time dimension, respectively; the training pressure time sequence variation feature extraction unit 430 respectively passes the training gas pressure value time sequence input vector and the training seawater pressure value time sequence vector obtained by the training data time sequence arrangement unit 420 through the multi-scale neighborhood feature extraction module to obtain a training gas phase pressure time sequence feature vector and a training liquid phase pressure time sequence feature vector; the training pressure difference calculating unit 440 calculates a training pressure difference feature vector between the training gas phase pressure time sequence feature vector and the training liquid phase pressure time sequence feature vector obtained by the training pressure time sequence change feature extracting unit 430; then, the training flow rate association coding unit 450 performs association coding on the obtained training gas flow rate value time sequence input vector and the training seawater flow rate value time sequence input vector obtained by the training data time sequence arrangement unit 420 to obtain a training gas-liquid flow rate synergy matrix; the training flow velocity time sequence variation feature extraction unit 460 passes the training gas-liquid flow velocity synergy matrix obtained by the training flow velocity correlation encoding unit 450 through the convolutional neural network model as a filter to obtain a training gas-liquid flow velocity synergy feature vector; the training responsiveness correlation unit 470 calculates a responsiveness estimate of the training gas-liquid flow rate synergy feature vector obtained by the training flow rate time sequence variation feature extraction unit 460 with respect to the training pressure difference feature vector obtained by the training pressure difference calculation unit 440 to obtain a training classification feature matrix; the feature optimization unit 480 performs feature redundancy optimization based on stacking of a low-cost bottleneck mechanism on the training classification feature matrix calculated by the training response association unit 470 to obtain an optimized training classification feature matrix; the classification loss unit 490 passes the optimization training classification feature matrix obtained by the feature optimization unit 480 through the classifier to obtain a classification loss function value; the model training unit 500 trains the multi-scale neighborhood feature extraction module, the convolutional neural network model as a filter, and the classifier based on the classification loss function value obtained by the classification loss unit 490 and by gradient descent direction vessels.
In particular, in the technical solution of the present application, when the classification feature matrix is obtained by calculating the estimation of the responsiveness of the gas-liquid flow velocity synergy feature vector with respect to the differential pressure feature vector, the channel dimension correlation feature of the convolutional neural network model of the cross-time domain correlation coding value of the gas flow velocity value and the seawater flow velocity value is considered by the gas-liquid flow velocity synergy feature vector, and the differential pressure feature vector expresses the position-by-position multi-scale time sequence neighborhood correlation feature difference value of the gas phase pressure value and the liquid phase pressure value, which corresponds to different distribution dimensions of the feature, so that the responsiveness between different dimension distributions can be fully expressed by using the feature responsiveness of the classification feature matrix in the cross dimension. On the other hand, considering that the feature distribution in each dimension is not completely orthogonal, redundant features are inevitably present in the classification feature matrix, so that the classification efficiency of the classification feature matrix through the classifier is affected, that is, the training speed of the model is reduced. Thus, the applicant of the present application performs training on the matrix of classification features, e.g., denoted as
Figure SMS_55
Feature redundancy optimization based on low-cost bottleneck-mechanism stacking is performed to obtain an optimized classified feature matrix, for example, marked as +. >
Figure SMS_56
The method is specifically expressed as follows:
Figure SMS_57
/>
Figure SMS_58
Figure SMS_59
wherein,,
Figure SMS_61
classifying a feature matrix for said training, +.>
Figure SMS_64
Representing a single layer convolution operation,/->
Figure SMS_66
、/>
Figure SMS_60
And->
Figure SMS_62
Respectively represent the position-by-position addition, subtraction and multiplication of the feature matrix, and +.>
Figure SMS_63
And->
Figure SMS_65
For biasing the feature matrix +.>
Figure SMS_67
And (5) optimizing the multi-scale associated feature map. Here, the feature redundancy optimization based on the low-cost bottleneck-mechanism stacking can use the low-cost bottleneck mechanism of the multiply-add stacking of two low-cost transformation features to perform feature expansion, and match a residual path by biasing stacking channels with uniform values, so that hidden distribution information under intrinsic features is revealed in redundancy features through low-cost operation transformation similar to a basic residual module, and a more intrinsic expression of the features is obtained through a simple and effective convolution operation architecture, thereby optimizing the redundant feature expression of the classification feature matrix, improving the classification efficiency of the classification feature matrix through a classifier, namely improving the training speed of a model. In this way, the liquid phase pressure on the sea water side and the tail gas propagation side can be accurately performed in real time based on the flow velocity change conditions of the ship tail gas and the sea waterAnd the self-adaptive control of the pressure difference between the gas phase pressures is used for optimizing the desulfurization effect and efficiency of the ship tail gas.
In summary, the marine tail gas desulfurization system 300 according to the embodiment of the present application is illustrated, and by using an artificial intelligence technology of deep learning, differential pressure time sequence dynamic change characteristic information of a marine tail gas pressure value and a seawater hydraulic pressure value, and time sequence change association characteristic information between a marine tail gas flow velocity value and a seawater flow velocity value are excavated, so that differential pressure time sequence change characteristic enhancement is performed based on the flow velocity time sequence association characteristic of the marine tail gas and the seawater, thereby optimizing expression of the differential pressure time sequence change characteristic, improving differential pressure control accuracy between a seawater side liquid phase pressure and a marine tail gas side gas phase pressure, and optimizing desulfurization effect and efficiency of the marine tail gas.
As described above, the marine vessel exhaust gas desulfurization system according to the embodiment of the present application may be implemented in various terminal devices. In one example, the marine vessel exhaust gas desulfurization system 300 according to embodiments of the present application may be integrated into the terminal device as one software module and/or hardware module. For example, the marine exhaust gas desulfurization system 300 can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the marine exhaust gas desulfurization system 300 can also be one of a number of hardware modules of the terminal equipment.
Alternatively, in another example, the marine vessel exhaust gas desulfurization system 300 and the terminal device may be separate devices, and the marine vessel exhaust gas desulfurization system 300 may be connected to the terminal device via a wired and/or wireless network and communicate interactive information in accordance with a agreed data format.
Exemplary method
Fig. 9 is a flow chart of a marine vessel exhaust gas desulfurization process according to an embodiment of the present application. As shown in fig. 9, the marine vessel exhaust gas desulfurization process according to the embodiment of the present application includes the steps of: s110, collecting tail gas and seawater of the sulfur-containing ship; s120, preprocessing the tail gas of the sulfur-containing ship to obtain tail gas after dust removal and oil removal; s130, preprocessing the seawater to obtain preprocessed seawater; s140, introducing the tail gas after dust removal and oil removal and the pretreated seawater into a membrane contactor to carry out marine tail gas desulfurization so as to obtain seawater after sulfur dioxide absorption; and S150, carrying out oxidation treatment on the seawater after absorbing sulfur dioxide to obtain oxidized sulfate.
In one example, in the above marine vessel exhaust gas desulfurization process, the step S140 includes: acquiring air pressure values and flow velocity values of ship tail gas at a plurality of preset time points in a preset time period, and hydraulic pressure values and flow velocity values of seawater at the preset time points; arranging the air pressure values and the flow velocity values of the ship tail gas at a plurality of preset time points and the hydraulic pressure values and the flow velocity values of the seawater at a plurality of preset time points into an air pressure value time sequence input vector, an air flow velocity value time sequence input vector, a seawater pressure value time sequence vector and a seawater flow velocity value time sequence input vector according to a time dimension respectively; respectively passing the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector through a multi-scale neighborhood feature extraction module to obtain a gas phase pressure time sequence feature vector and a liquid phase pressure time sequence feature vector; calculating a pressure difference feature vector between the gas phase pressure time sequence feature vector and the liquid phase pressure time sequence feature vector; performing associated coding on the gas flow velocity value time sequence input vector and the seawater flow velocity value time sequence input vector to obtain a gas-liquid flow velocity synergy matrix; the gas-liquid flow rate cooperative matrix is passed through a convolutional neural network model serving as a filter to obtain a gas-liquid flow rate cooperative eigenvector; calculating a responsiveness estimate of the gas-liquid flow rate synergy eigenvector relative to the pressure differential eigenvector to obtain a classification eigenvector; and passing the classification feature matrix through a classifier to obtain a classification result, wherein the classification result is used for indicating that the pressure difference between the liquid phase pressure on the sea water side and the gas phase pressure on the ship tail gas side should be increased or decreased. The gas pressure value time sequence input vector and the seawater pressure value time sequence input vector are respectively passed through a multi-scale neighborhood feature extraction module to obtain a gas phase pressure time sequence feature vector and a liquid phase pressure time sequence feature vector, and the method comprises the following steps: respectively inputting the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector into a first convolution layer of the multi-scale neighborhood feature extraction module to obtain the first neighborhood scale gas phase pressure time sequence feature vector and the first neighborhood scale liquid phase pressure time sequence feature vector, wherein the first convolution layer is provided with a first one-dimensional convolution kernel with a first length; respectively inputting the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector into a second convolution layer of the multi-scale neighborhood feature extraction module to obtain a second neighborhood scale gas phase pressure time sequence feature vector and a second neighborhood scale liquid phase pressure time sequence feature vector, wherein the second convolution layer is provided with a second one-dimensional convolution kernel with a second length, and the first length is different from the second length; the first neighborhood scale gas phase pressure time sequence feature vector and the first neighborhood scale liquid phase pressure time sequence feature vector are respectively cascaded with the second neighborhood scale gas phase pressure time sequence feature vector and the second neighborhood scale liquid phase pressure time sequence feature vector to obtain the gas phase pressure time sequence feature vector and the liquid phase pressure time sequence feature vector; and passing the gas-liquid flow rate synergy matrix through a convolutional neural network model as a filter to obtain a gas-liquid flow rate synergy eigenvector, comprising: each layer of the convolutional neural network model used as the filter performs the following steps on input data in forward transfer of the layer: 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; wherein the output of the last layer of the convolutional neural network as a filter is the gas-liquid flow velocity cooperative eigenvector, and the input of the first layer of the convolutional neural network as a filter is the gas-liquid flow velocity cooperative matrix; more specifically, calculating a responsiveness estimate of the gas-liquid flow rate synergy feature vector relative to the pressure differential feature vector to obtain a classification feature matrix, comprising: calculating a responsiveness estimate of the gas-liquid flow rate synergy eigenvector relative to the pressure differential eigenvector to obtain a classification eigenvector by the formula; wherein, the formula is:
Figure SMS_68
Wherein the method comprises the steps of
Figure SMS_69
Representing the gas-liquid flow rate synergy eigenvector, < >>
Figure SMS_70
Representing the differential pressure characteristic vector,/->
Figure SMS_71
Representing the classification feature matrix.
In summary, the marine tail gas desulfurization process according to the embodiment of the application is clarified, by means of an artificial intelligence technology of deep learning, differential pressure time sequence dynamic change characteristic information of a marine tail gas pressure value and a seawater hydraulic pressure value and time sequence change association characteristic information between a marine tail gas flow velocity value and a seawater flow velocity value are excavated, so that differential pressure time sequence change characteristic enhancement is performed based on the flow velocity time sequence association characteristic of the marine tail gas and the seawater, expression of the differential pressure time sequence change characteristic is optimized, differential pressure control accuracy between the seawater side liquid phase pressure and the marine tail gas side gas phase pressure is improved, and desulfurization effect and efficiency of the marine tail gas are optimized.
Exemplary electronic device
Next, an electronic device according to an embodiment of the present application is described with reference to fig. 10.
Fig. 10 illustrates a block diagram of an electronic device according to an embodiment of the present application.
As shown in fig. 10, the electronic device 10 includes one or more processors 11 and a memory 12.
The processor 11 may be a Central Processing Unit (CPU) or other form of processing unit having data processing and/or instruction execution capabilities, and may control other components in the electronic device 10 to perform desired functions.
Memory 12 may include one or more computer program products that may include various forms of computer-readable storage media, such as volatile memory and/or non-volatile memory. The volatile memory may include, for example, random Access Memory (RAM) and/or cache memory (cache), and the like. The non-volatile memory may include, for example, read Only Memory (ROM), hard disk, flash memory, and the like. One or more computer program instructions may be stored on the computer readable storage medium that may be executed by the processor 11 to perform the functions in the marine exhaust gas desulfurization system of the various embodiments of the present application described above, and/or other desired functions. Various content, such as a gas-liquid flow rate co-matrix, may also be stored in the computer readable storage medium.
In one example, the electronic device 10 may further include: an input device 13 and an output device 14, which are interconnected by a bus system and/or other forms of connection mechanisms (not shown).
The input means 13 may comprise, for example, a keyboard, a mouse, etc.
The output device 14 may output various information including the classification result and the like to the outside. The output means 14 may include, for example, a display, speakers, a printer, and a communication network and remote output devices connected thereto, etc.
Of course, only some of the components of the electronic device 10 that are relevant to the present application are shown in fig. 10 for simplicity, components such as buses, input/output interfaces, etc. are omitted. In addition, the electronic device 10 may include any other suitable components depending on the particular application.
Exemplary computer program product and computer readable storage Medium
In addition to the methods and apparatus described above, embodiments of the present application may also be a computer program product comprising computer program instructions which, when executed by a processor, cause the processor to perform steps in the functions in a marine vessel exhaust gas desulfurization process according to various embodiments of the present application described in the "exemplary systems" section of the present specification.
The computer program product may write program code for performing the operations of embodiments of the present application in any combination of one or more programming languages, including an object oriented programming language such as Java, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device, partly on a remote computing device, or entirely on the remote computing device or server.
Furthermore, embodiments of the present application may also be a computer-readable storage medium, having stored thereon computer program instructions, which when executed by a processor, cause the processor to perform steps in the functions in a marine vessel exhaust gas desulfurization process according to various embodiments of the present application described in the above-mentioned "exemplary systems" section of the present specification.
The computer readable storage medium may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may include, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium would include the following: an electrical connection having one or more wires, a portable disk, a hard disk, random Access Memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The basic principles of the present application have been described above in connection with specific embodiments, however, it should be noted that the advantages, benefits, effects, etc. mentioned in the present application are merely examples and not limiting, and these advantages, benefits, effects, etc. are not to be considered as necessarily possessed by the various embodiments of the present application. Furthermore, the specific details disclosed herein are for purposes of illustration and understanding only, and are not intended to be limiting, as the application is not intended to be limited to the details disclosed herein as such.
The block diagrams of the devices, apparatuses, devices, systems referred to in this application are only illustrative examples and are not intended to require or imply that the connections, arrangements, configurations must be made in the manner shown in the block diagrams. As will be appreciated by one of skill in the art, the devices, apparatuses, devices, systems may be connected, arranged, configured in any manner. Words such as "including," "comprising," "having," and the like are words of openness and mean "including but not limited to," and are used interchangeably therewith. The terms "or" and "as used herein refer to and are used interchangeably with the term" and/or "unless the context clearly indicates otherwise. The term "such as" as used herein refers to, and is used interchangeably with, the phrase "such as, but not limited to.
It is also noted that in the apparatus, devices and methods of the present application, the components or steps may be disassembled and/or assembled. Such decomposition and/or recombination should be considered as equivalent to the present application.
The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
The foregoing description has been presented for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of the application to the form disclosed herein. Although a number of example aspects and embodiments have been discussed above, a person of ordinary skill in the art will recognize certain variations, modifications, alterations, additions, and subcombinations thereof.

Claims (10)

1. A marine vessel exhaust gas desulfurization system, comprising:
the ship tail gas and sea water collecting module is used for collecting the sulfur-containing ship tail gas and sea water;
The sulfur-containing ship tail gas pretreatment module is used for pretreating the sulfur-containing ship tail gas to obtain tail gas after dust removal and oil removal;
the seawater pretreatment module is used for pretreating the seawater to obtain pretreated seawater;
the membrane contactor desulfurization module is used for introducing the tail gas after dust removal and oil removal and the pretreated seawater into a membrane contactor to carry out marine tail gas desulfurization so as to obtain seawater after sulfur dioxide absorption;
and the seawater post-treatment module is used for carrying out oxidation treatment on the seawater absorbing sulfur dioxide to obtain oxidized sulfate.
2. The marine vessel exhaust gas desulfurization system of claim 1, wherein the membrane contactor desulfurization module comprises:
the data acquisition unit is used for acquiring the air pressure value and the flow velocity value of the ship tail gas at a plurality of preset time points in a preset time period, and the hydraulic pressure value and the flow velocity value of the seawater at the preset time points;
the data time sequence arrangement unit is used for arranging the air pressure values and the flow velocity values of the ship tail gas at a plurality of preset time points and the hydraulic pressure values and the flow velocity values of the seawater at a plurality of preset time points into an air pressure value time sequence input vector, an air flow velocity value time sequence input vector, a seawater pressure value time sequence vector and a seawater flow velocity value time sequence input vector according to the time dimension respectively;
The pressure time sequence change feature extraction unit is used for respectively passing the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector through a multi-scale neighborhood feature extraction module to obtain a gas phase pressure time sequence feature vector and a liquid phase pressure time sequence feature vector;
a pressure difference calculation unit configured to calculate a pressure difference feature vector between the gas phase pressure time series feature vector and the liquid phase pressure time series feature vector;
the flow rate association coding unit is used for carrying out association coding on the gas flow rate value time sequence input vector and the seawater flow rate value time sequence input vector so as to obtain a gas-liquid flow rate synergy matrix;
the flow velocity time sequence change feature extraction unit is used for enabling the gas-liquid flow velocity coordination matrix to pass through a convolutional neural network model serving as a filter so as to obtain a gas-liquid flow velocity coordination feature vector;
a responsiveness correlation unit, configured to calculate responsiveness estimation of the gas-liquid flow rate cooperative feature vector relative to the differential pressure feature vector to obtain a classification feature matrix; and
and the differential pressure control unit is used for passing the classification characteristic matrix through a classifier to obtain a classification result, wherein the classification result is used for indicating that the differential pressure between the liquid phase pressure on the sea water side and the gas phase pressure on the tail gas side of the ship is increased or reduced.
3. The marine vessel exhaust gas desulfurization system of claim 2, wherein the multi-scale neighborhood feature extraction module comprises: the device comprises a first convolution layer, a second convolution layer parallel to the first convolution layer and a multi-scale feature fusion layer connected with the first convolution layer and the second convolution layer, wherein the first convolution layer uses a one-dimensional convolution kernel with a first length, and the second convolution layer uses a one-dimensional convolution kernel with a second length.
4. A marine vessel exhaust gas desulfurization system according to claim 3, wherein the pressure time series variation feature extraction unit comprises:
a first neighborhood scale feature extraction subunit, configured to input the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector into a first convolution layer of the multi-scale neighborhood feature extraction module respectively to obtain the first neighborhood scale gas phase pressure time sequence feature vector and the first neighborhood scale liquid phase pressure time sequence feature vector, where the first convolution layer has a first one-dimensional convolution kernel with a first length;
a second neighborhood scale feature extraction subunit, configured to input the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector into a second convolution layer of the multi-scale neighborhood feature extraction module respectively to obtain the second neighborhood scale gas phase pressure time sequence feature vector and the second neighborhood scale liquid phase pressure time sequence feature vector, where the second convolution layer has a second one-dimensional convolution kernel with a second length, and the first length is different from the second length; and
And the multi-scale cascading subunit is used for cascading the first neighborhood scale gas phase pressure time sequence feature vector and the first neighborhood scale liquid phase pressure time sequence feature vector with the second neighborhood scale gas phase pressure time sequence feature vector and the second neighborhood scale liquid phase pressure time sequence feature vector respectively to obtain the gas phase pressure time sequence feature vector and the liquid phase pressure time sequence feature vector.
Wherein the first neighborhood scale feature extraction subunit is configured to: using a first convolution layer of the multi-scale neighborhood feature extraction module to respectively perform one-dimensional convolution coding on the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector according to the following one-dimensional convolution formula so as to obtain a first neighborhood scale gas phase pressure time sequence feature vector and a first neighborhood scale liquid phase pressure time sequence feature vector;
wherein, the formula is:
Figure QLYQS_1
wherein,,ais the first convolution kernelxWidth in the direction,
Figure QLYQS_2
For the first convolution kernel parameter vector, +.>
Figure QLYQS_3
For a local vector matrix that operates with a convolution kernel,wfor the size of the first one-dimensional convolution kernel，XRepresenting the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector, +. >
Figure QLYQS_4
Representing that one-dimensional convolution coding is respectively carried out on the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector;
the second neighborhood scale feature extraction subunit is configured to: using a second convolution layer of the multi-scale neighborhood feature extraction module to respectively perform one-dimensional convolution coding on the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector according to the following one-dimensional convolution formula so as to obtain a second neighborhood scale gas phase pressure time sequence feature vector and a second neighborhood scale liquid phase pressure time sequence feature vector;
wherein, the formula is:
Figure QLYQS_5
wherein b is the second convolution kernelxWidth in the direction,
Figure QLYQS_6
For a second convolution kernel parameter vector, +.>
Figure QLYQS_7
For the local vector matrix to operate with the convolution kernel function, m is the size of the second one-dimensional convolution kernel,Xrepresenting the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector, +.>
Figure QLYQS_8
And respectively carrying out one-dimensional convolution coding on the gas pressure value time sequence input vector and the seawater pressure value time sequence input vector.
5. The marine vessel exhaust gas desulfurization system according to claim 4, wherein the flow velocity time series variation feature extraction unit is configured to: each layer of the convolutional neural network model used as the filter performs the following steps on input data in forward transfer of the layer:
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; and
non-linear activation is carried out on the pooled feature map so as to obtain an activated feature map;
wherein the output of the last layer of the convolutional neural network as a filter is the gas-liquid flow velocity cooperative eigenvector, and the input of the first layer of the convolutional neural network as a filter is the gas-liquid flow velocity cooperative matrix.
6. The marine vessel exhaust gas desulfurization system of claim 5, wherein the responsiveness correlation unit is configured to: calculating a responsiveness estimate of the gas-liquid flow rate synergy eigenvector relative to the pressure differential eigenvector to obtain a classification eigenvector by the formula;
wherein, the formula is:
Figure QLYQS_9
wherein the method comprises the steps of
Figure QLYQS_10
Representing the gas-liquid flow rate synergy eigenvector, < >>
Figure QLYQS_11
Representing the differential pressure characteristic vector,/->
Figure QLYQS_12
Representing the classification feature matrix.
7. The marine vessel exhaust gas desulfurization system of claim 6, further comprising a training unit for training the multi-scale neighborhood feature extraction module, the convolutional neural network model as a filter, and the classifier.
8. The marine vessel exhaust gas desulfurization system of claim 7, wherein the training unit comprises:
the system comprises a training data acquisition unit, a control unit and a control unit, wherein the training data acquisition unit is used for acquiring training data, the training data comprises training air pressure values and training flow rate values of ship tail gas at a plurality of preset time points in a preset time period, training hydraulic pressure values and training flow rate values of seawater at the preset time points, and a true value that the pressure difference between the seawater side liquid phase pressure and the ship tail gas side gas phase pressure is required to be increased or reduced;
the training data time sequence arrangement unit is used for arranging the training air pressure values and the training flow velocity values of the ship tail gas at a plurality of preset time points and arranging the training hydraulic pressure values and the training flow velocity values of the seawater at a plurality of preset time points into a training air pressure value time sequence input vector, a training air flow velocity value time sequence input vector, a training seawater pressure value time sequence vector and a training seawater flow velocity value time sequence input vector according to time dimensions respectively;
the training pressure time sequence change feature extraction unit is used for enabling the training gas pressure value time sequence input vector and the training seawater pressure value time sequence vector to respectively pass through the multi-scale neighborhood feature extraction module so as to obtain a training gas phase pressure time sequence feature vector and a training liquid phase pressure time sequence feature vector;
The training pressure difference calculation unit is used for calculating a training pressure difference characteristic vector between the training gas phase pressure time sequence characteristic vector and the training liquid phase pressure time sequence characteristic vector;
the training flow rate association coding unit is used for carrying out association coding on the training gas flow rate value time sequence input vector and the training seawater flow rate value time sequence input vector so as to obtain a training gas-liquid flow rate cooperative matrix;
the training flow velocity time sequence change feature extraction unit is used for enabling the training gas-liquid flow velocity cooperative matrix to pass through the convolutional neural network model serving as a filter so as to obtain a training gas-liquid flow velocity cooperative feature vector;
the training response association unit is used for calculating the response estimation of the training gas-liquid flow rate cooperative feature vector relative to the training differential pressure feature vector so as to obtain a training classification feature matrix;
the feature optimization unit is used for performing feature redundancy optimization on the training classification feature matrix based on the stacking of the low-cost bottleneck mechanism so as to obtain an optimized training classification feature matrix; and
the classification loss unit is used for passing the optimized training classification characteristic matrix through the classifier to obtain a classification loss function value;
And the model training unit is used for training the multi-scale neighborhood feature extraction module, the convolution neural network model serving as the filter and the classifier based on the classification loss function value and through the gradient descending direction ship.
9. The marine vessel exhaust gas desulfurization system of claim 8, wherein the feature optimization unit is configured to: performing feature redundancy optimization on the training classification feature matrix based on low-cost bottleneck mechanism stacking by using the following optimization formula to obtain the optimized training classification feature matrix;
wherein, the optimization formula is:
Figure QLYQS_13
Figure QLYQS_14
Figure QLYQS_15
wherein,,
Figure QLYQS_17
classifying a feature matrix for said training, +.>
Figure QLYQS_19
Representing a single layer convolution operation,/->
Figure QLYQS_20
、/>
Figure QLYQS_18
And->
Figure QLYQS_21
Respectively represent the position-by-position addition, subtraction and multiplication of the feature matrix, and +.>
Figure QLYQS_22
And->
Figure QLYQS_23
For biasing the feature matrix +.>
Figure QLYQS_16
And (5) optimizing the multi-scale associated feature map.
10. A marine vessel exhaust gas desulfurization process, comprising:
collecting tail gas and seawater of a sulfur-containing ship;
pretreating the tail gas of the sulfur-containing ship to obtain tail gas after dust removal and oil removal;
pretreating the seawater to obtain pretreated seawater;
introducing the tail gas after dust removal and oil removal and the pretreated seawater into a membrane contactor to carry out marine tail gas desulfurization so as to obtain seawater after sulfur dioxide absorption;
And (3) oxidizing the seawater after absorbing sulfur dioxide to obtain oxidized sulfate.
CN202310445481.6A 2023-04-24 2023-04-24 Ship tail gas desulfurization process and system thereof Active CN116392930B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310445481.6A CN116392930B (en) 2023-04-24 2023-04-24 Ship tail gas desulfurization process and system thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310445481.6A CN116392930B (en) 2023-04-24 2023-04-24 Ship tail gas desulfurization process and system thereof

Publications (2)

Publication Number Publication Date
CN116392930A true CN116392930A (en) 2023-07-07
CN116392930B CN116392930B (en) 2023-08-29

Family

ID=87012298

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310445481.6A Active CN116392930B (en) 2023-04-24 2023-04-24 Ship tail gas desulfurization process and system thereof

Country Status (1)

Country Link
CN (1) CN116392930B (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116832588A (en) * 2023-08-25 2023-10-03 湖北鼎信成套设备有限公司 Acid regeneration flue gas purifying device and method thereof
CN116870654A (en) * 2023-07-10 2023-10-13 滁州锡安环保科技有限责任公司 Industrial waste gas treatment system and method thereof
CN117018858A (en) * 2023-08-11 2023-11-10 滁州锡安环保科技有限责任公司 Industrial waste gas purifying apparatus and control method thereof
CN117138589A (en) * 2023-10-27 2023-12-01 克拉玛依曜诚石油科技有限公司 Pollution monitoring system and method for reverse osmosis membrane
CN117965214A (en) * 2024-04-01 2024-05-03 新疆凯龙清洁能源股份有限公司 Method and system for preparing synthesis gas by removing carbon dioxide from natural gas

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060045801A1 (en) * 2004-08-27 2006-03-02 Alstom Technology Ltd. Model predictive control of air pollution control processes
CN101091873A (en) * 2007-04-21 2007-12-26 大连理工大学 Technical method by using seawater to remove sulfur dioxide in smoke through membrane absorption method
EP3346423A1 (en) * 2017-01-04 2018-07-11 STMicroelectronics Srl Deep convolutional network heterogeneous architecture system and device
KR102271069B1 (en) * 2020-03-31 2021-06-29 조선대학교산학협력단 Method and apparatus for generating state predecting model of coal-fired power plant boiler
CN113694696A (en) * 2021-08-25 2021-11-26 中海油能源发展股份有限公司北京安全环保工程技术研究院 Ship tail gas desulfurization device and desulfurization method
CN114191953A (en) * 2021-12-07 2022-03-18 国网河北能源技术服务有限公司 Flue gas desulfurization and denitrification control method based on convolutional neural network and XGboost
CN115456053A (en) * 2022-08-27 2022-12-09 华能新能源股份有限公司 Wind wheel heat dissipation system and method of offshore wind generating set
CN115688592A (en) * 2022-11-09 2023-02-03 福建德尔科技股份有限公司 Rectification control system and method for preparing electronic-grade carbon tetrafluoride
CN115853173A (en) * 2022-12-16 2023-03-28 吴国尧 Building curtain wall for construction and installation

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060045801A1 (en) * 2004-08-27 2006-03-02 Alstom Technology Ltd. Model predictive control of air pollution control processes
CN101091873A (en) * 2007-04-21 2007-12-26 大连理工大学 Technical method by using seawater to remove sulfur dioxide in smoke through membrane absorption method
EP3346423A1 (en) * 2017-01-04 2018-07-11 STMicroelectronics Srl Deep convolutional network heterogeneous architecture system and device
KR102271069B1 (en) * 2020-03-31 2021-06-29 조선대학교산학협력단 Method and apparatus for generating state predecting model of coal-fired power plant boiler
CN113694696A (en) * 2021-08-25 2021-11-26 中海油能源发展股份有限公司北京安全环保工程技术研究院 Ship tail gas desulfurization device and desulfurization method
CN114191953A (en) * 2021-12-07 2022-03-18 国网河北能源技术服务有限公司 Flue gas desulfurization and denitrification control method based on convolutional neural network and XGboost
CN115456053A (en) * 2022-08-27 2022-12-09 华能新能源股份有限公司 Wind wheel heat dissipation system and method of offshore wind generating set
CN115688592A (en) * 2022-11-09 2023-02-03 福建德尔科技股份有限公司 Rectification control system and method for preparing electronic-grade carbon tetrafluoride
CN115853173A (en) * 2022-12-16 2023-03-28 吴国尧 Building curtain wall for construction and installation

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116870654A (en) * 2023-07-10 2023-10-13 滁州锡安环保科技有限责任公司 Industrial waste gas treatment system and method thereof
CN116870654B (en) * 2023-07-10 2024-06-07 广东源锦自动化科技有限公司 Industrial waste gas treatment system and method thereof
CN117018858A (en) * 2023-08-11 2023-11-10 滁州锡安环保科技有限责任公司 Industrial waste gas purifying apparatus and control method thereof
CN116832588A (en) * 2023-08-25 2023-10-03 湖北鼎信成套设备有限公司 Acid regeneration flue gas purifying device and method thereof
CN116832588B (en) * 2023-08-25 2024-02-02 湖北鼎信成套设备有限公司 Acid regeneration flue gas purifying device and method thereof
CN117138589A (en) * 2023-10-27 2023-12-01 克拉玛依曜诚石油科技有限公司 Pollution monitoring system and method for reverse osmosis membrane
CN117138589B (en) * 2023-10-27 2024-02-13 克拉玛依曜诚石油科技有限公司 Pollution monitoring system and method for reverse osmosis membrane
CN117965214A (en) * 2024-04-01 2024-05-03 新疆凯龙清洁能源股份有限公司 Method and system for preparing synthesis gas by removing carbon dioxide from natural gas

Also Published As

Publication number Publication date
CN116392930B (en) 2023-08-29

Similar Documents

Publication Publication Date Title
CN116392930B (en) Ship tail gas desulfurization process and system thereof
CN110288665B (en) Image description method based on convolutional neural network, computer-readable storage medium and electronic device
CN109559315B (en) Water surface segmentation method based on multipath deep neural network
CN112183061B (en) Multi-intention spoken language understanding method, electronic equipment and storage medium
CN112348911B (en) Semantic constraint-based method and system for generating fine-grained image by stacking texts
CN106919557A (en) A kind of document vector generation method of combination topic model
CN116135797B (en) Intelligent control system for sewage treatment
CN116146309B (en) Ship tail gas treatment system and method thereof
CN109933773A (en) A kind of multiple semantic sentence analysis system and method
CN112835585A (en) Program understanding method and system based on abstract syntax tree
US20230031733A1 (en) Method for training a speech recognition model and method for speech recognition
CN114579699A (en) Training method and device for pre-training language model
CN116603390B (en) Ship tail gas washing system and method thereof
CN110858307B (en) Character recognition model training method and device and character recognition method and device
WO1996010799A1 (en) Method and system for extracting features from handwritten text
Zhou et al. Nonparametric modeling of ship maneuvering motions in calm water and regular waves based on R-LSTM hybrid method
CN105426912A (en) Blind separation method for replacement aliasing image
CN116663501A (en) Chinese variant text conversion method based on multi-modal sharing weight
CN116534700A (en) Control system and method for stair climbing machine
CN115849519A (en) Organic modularization electrocatalytic oxidation treatment device
CN112131879A (en) Relationship extraction system, method and device
Miao et al. Low‐latency transformer model for streaming automatic speech recognition
CN114927128A (en) Voice keyword detection method and device, electronic equipment and readable storage medium
CN111581911B (en) Method for automatically adding punctuation to real-time text, model construction method and device
CN114564962A (en) Semantic communication code rate control method based on Transformer

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
GR01 Patent grant
GR01 Patent grant
CP03 Change of name, title or address
CP03 Change of name, title or address

Address after: 310000 Room 601, building 3, no.371, Mingxing Road, Xiaoshan Economic and Technological Development Zone, Xiaoshan District, Hangzhou City, Zhejiang Province

Patentee after: Zhejiang Zheneng Mailing Environmental Technology Co.,Ltd.

Country or region after: China

Address before: 310000 Room 601, building 3, no.371, Mingxing Road, Xiaoshan Economic and Technological Development Zone, Xiaoshan District, Hangzhou City, Zhejiang Province

Patentee before: Zhejiang zheneng mailing Environmental Technology Co.,Ltd.

Country or region before: China