CN116146309B

CN116146309B - Ship tail gas treatment system and method thereof

Info

Publication number: CN116146309B
Application number: CN202310438910.7A
Authority: CN
Inventors: 沈敏强; 姚盛翔; 汪成; 赵媛媛; 龚良丰; 王德智; 叶慷; 王汝能; 方丰
Original assignee: Zhejiang Zheneng Mailing Environmental Technology Co ltd
Current assignee: Zhejiang Zheneng Mailing Environmental Technology Co ltd
Priority date: 2023-04-23
Filing date: 2023-04-23
Publication date: 2023-07-07
Anticipated expiration: 2043-04-23
Also published as: CN116146309A

Abstract

The utility model relates to the field of waste gas treatment, and specifically discloses a ship tail gas treatment system and a method thereof, which excavates time sequence dynamic change characteristics of a flow velocity value of ship tail gas by adopting a neural network model based on deep learning, so as to accurately carry out self-adaptive control of a current value of plasma discharge equipment in real time based on actual change conditions of the flow velocity of the ship tail gas, thereby improving ionization discharge efficiency and effect and effectively degrading harmful components in the ship tail gas.

Description

Ship tail gas treatment system and method thereof

Technical Field

The present application relates to the field of exhaust gas treatment, and more particularly, to a marine exhaust gas treatment system and method thereof.

Background

Maritime is a globally recognized major source of atmospheric pollutants. In recent years, with the rapid development of international maritime trade, more and more people are concerned about the global influence of atmospheric pollutants, and since marine exhaust emissions are easily transmitted in the atmosphere for a long distance, from sea to land, even from one continent to another, marine exhaust emissions can have a significant influence on the air quality in local and regional areas. In addition, part of ship emissions occur in coastal areas, and exhaust pollutants can be directly spread to the continents, causing environmental problems affecting human health and the ecosystem.

In the field of exhaust gas treatment, exhaust gas (mainly referred to as engine exhaust gas) is purified by adopting a plasma discharge technology, and is a relatively efficient exhaust gas treatment mode in recent years. The principle of plasma discharge treatment is that the pollutants in the tail gas are decomposed by means of plasma generated by ionization, so that the aim of degrading the pollutants is fulfilled. However, most of the existing exhaust gas treatment devices are used for treating automobile exhaust gas, and the treatment efficiency is difficult to meet the requirements of ship exhaust gas treatment.

Accordingly, an optimized marine exhaust gas treatment scheme 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 treatment system and a method thereof, which excavates time sequence dynamic change characteristics of a flow velocity value of ship tail gas by adopting a neural network model based on deep learning, so as to accurately carry out self-adaptive control of a current value of plasma discharge equipment in real time based on actual change conditions of the flow velocity of the ship tail gas, thereby improving ionization discharge efficiency and effect and effectively degrading harmful components in the ship tail gas.

According to one aspect of the present application, there is provided a ship tail gas treatment method comprising:

Receiving marine tail gas through a gas supply pipeline; and

and carrying out ionization discharge on the ship tail gas so as to realize degradation of harmful components in the ship tail gas by means of plasma ionization particles generated by the ionization discharge.

In the above ship tail gas treatment method, the performing ionization discharge on the ship tail gas includes: acquiring flow velocity values of ship tail gas at a plurality of preset time points in a preset time period; arranging the flow velocity values of the ship tail gas at a plurality of preset time points into flow velocity value time sequence input vectors according to the time dimension; calculating the difference value between the flow velocity values of every two adjacent positions in the flow velocity value time sequence input vector to obtain a flow velocity change time sequence input vector; cascading the flow velocity value time sequence input vector and the flow velocity change time sequence input vector to obtain a flow velocity value dynamic-static multidimensional input vector; the flow velocity value dynamic-static multidimensional input vector is passed through a double-flow model comprising a first convolution neural network model and a second convolution neural network model to obtain a flow velocity value dynamic-static characteristic vector; performing feature enhancement on the dynamic-static feature vector of the flow velocity value by using a Gaussian density chart to obtain a decoding feature matrix; performing feature distribution optimization on the decoding feature matrix to obtain an optimized decoding feature matrix; and carrying out decoding regression on the optimized decoding characteristic matrix through a decoder to obtain a decoding value, wherein the decoding value is used for representing the recommended current value of the plasma discharge equipment.

In the above ship tail gas treatment method, passing the flow velocity value dynamic-static multidimensional input vector through a dual-flow model including a first convolutional neural network model and a second convolutional neural network model to obtain a flow velocity value dynamic-static feature vector, including: performing convolution processing, pooling processing and nonlinear activation processing based on a one-dimensional convolution kernel of a first scale on the flow velocity value dynamic-static multidimensional input vector in forward transfer of layers by using each layer of a first convolution neural network of the double-flow model to output the first scale flow velocity value dynamic-static characteristic vector by the last layer of the first convolution neural network; performing convolution processing, pooling processing and nonlinear activation processing based on a one-dimensional convolution kernel of a second scale on the flow velocity value dynamic-static multidimensional input vector in forward transfer of layers by using layers of a second convolution neural network of the double-flow model to output a second scale flow velocity value dynamic-static feature vector by a last layer of the second convolution neural network, wherein the first scale is not equal to the second scale; and concatenating the first scale flow velocity value dynamic-static feature vector and the second scale flow velocity value dynamic-static feature vector to obtain the flow velocity value dynamic-static feature vector.

In the above ship tail gas treatment method, the feature enhancement is performed on the flow velocity value dynamic-static feature vector by using a gaussian density chart to obtain a decoding feature matrix, including: and constructing a Gaussian density map of the dynamic-static characteristic vector of the flow velocity value to obtain the Gaussian density map. The mean value vector of the Gaussian density map is the flow velocity value dynamic-static characteristic vector, and the covariance matrix of the Gaussian density map is the variance between the characteristic values of the corresponding two positions in the flow velocity value dynamic-static characteristic vector; and performing Gaussian discretization on the Gaussian distribution of each position in the Gaussian density map to obtain the decoding characteristic matrix.

In the above ship tail gas treatment method, performing feature distribution optimization on the decoding feature matrix to obtain an optimized decoding feature matrix, including: performing feature affinity space affine learning optimization on the decoding feature matrix by using the following optimization formula to obtain the optimized decoding feature matrix; wherein, the optimization formula is:

wherein,,

diagonal matrix representing the decoding feature matrix by linear transformation,/a>

Representing the two norms of the decoding feature matrix, < >>

Represents the core norms of the decoding feature matrix, and +. >

Is the scale of the decoding feature matrix, +.>

Represents a logarithmic function value based on 2, < +.>

An exponential operation representing a matrix representing the calculation of a natural exponential function value raised to a power by a characteristic value at each position in the matrix,/v>

Representing the multiplication by the position point,

representing the optimized decoding feature matrix.

In the above ship tail gas treatment method, performing decoding regression on the optimized decoding feature matrix through a decoder to obtain a decoded value, where the decoded value is used to represent a current value of a recommended plasma discharge device, and the method includes: performing decoding regression on the optimized decoding feature matrix using the decoder in the following formula to obtain a decoded value representing a recommended current value of the plasma discharge apparatus; wherein, the formula is:

wherein->

Representing the optimized decoding feature matrix,/a>

Is the decoded value,/->

Is a weight matrix, < >>

Representing matrix multiplication.

According to another aspect of the present application, there is provided a marine vessel exhaust gas treatment system comprising:

the data acquisition unit is used for acquiring flow velocity values of the ship tail gas at a plurality of preset time points in a preset time period;

the arrangement unit is used for arranging the flow velocity values of the ship tail gas at a plurality of preset time points into flow velocity value time sequence input vectors according to the time dimension;

A difference value calculating unit, configured to calculate a difference value between flow velocity values of every two adjacent positions in the flow velocity value time sequence input vector to obtain a flow velocity change time sequence input vector;

the cascade unit is used for cascading the flow velocity value time sequence input vector and the flow velocity change time sequence input vector to obtain a flow velocity value dynamic-static multidimensional input vector;

the multi-scale convolution unit is used for enabling the flow velocity value dynamic-static multi-dimensional input vector to pass through a double-flow model comprising a first convolution neural network model and a second convolution neural network model to obtain a flow velocity value dynamic-static characteristic vector;

the Gaussian enhancement unit is used for carrying out characteristic enhancement on the dynamic-static characteristic vector of the flow velocity value by using a Gaussian density chart so as to obtain a decoding characteristic matrix;

the feature distribution optimizing unit is used for optimizing the feature distribution of the decoding feature matrix to obtain an optimized decoding feature matrix; and

and the decoding unit is used for carrying out decoding regression on the optimized decoding characteristic matrix through a decoder to obtain a decoding value, wherein the decoding value is used for representing the recommended current value of the plasma discharge equipment.

According to still another aspect of the present application, there is provided an electronic apparatus including: a processor; and a memory in which computer program instructions are stored which, when executed by the processor, cause the processor to perform a method for marine vessel exhaust gas treatment 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 treatment method as described above.

Compared with the prior art, the ship tail gas treatment system and the ship tail gas treatment method provided by the application excavate the time sequence dynamic change characteristic of the flow velocity value of the ship tail gas by adopting the neural network model based on deep learning, so that the self-adaptive control of the current value of the plasma discharge equipment is accurately performed in real time based on the actual change condition of the flow velocity of the ship tail gas, the ionization discharge efficiency and effect are improved, and harmful components in the ship tail gas are effectively degraded.

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 an application scenario diagram of a ship tail gas treatment method according to an embodiment of the present application;

FIG. 2 is a flow chart of a method of treating marine exhaust gas according to an embodiment of the present application;

FIG. 3 is a flow chart of an ionization discharge process for marine exhaust gas in a marine exhaust gas treatment method according to an embodiment of the present application;

FIG. 4 is a schematic diagram of a method for treating marine exhaust gas according to an embodiment of the present application;

FIG. 5 is a flow chart of dual stream model encoding in a marine tail gas treatment method according to an embodiment of the present application;

FIG. 6 is a flow chart of Gaussian intensity in a marine exhaust treatment method according to an embodiment of the application;

FIG. 7 is a block diagram of a marine vessel exhaust treatment system according to an embodiment of the present application;

fig. 8 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 field of exhaust gas treatment, exhaust gas (mainly engine exhaust gas) is purified by a plasma discharge technique, and this is a relatively efficient exhaust gas treatment system in recent years. However, most of the existing exhaust gas treatment devices are used for treating automobile exhaust gas, and the treatment efficiency is difficult to meet the requirements of ship exhaust gas treatment. Accordingly, an optimized marine exhaust gas treatment scheme is desired.

Specifically, in the technical scheme of the application, a ship tail gas treatment method is provided, which comprises the following steps: receiving marine tail gas through a gas supply pipeline; and carrying out ionization discharge on the ship tail gas so as to realize degradation of harmful components in the ship tail gas by means of plasma ionization particles generated by the ionization discharge.

Accordingly, in the process of actually performing the treatment of the ship tail gas, the ionization discharge is performed on the ship tail gas, and the degradation of harmful components in the ship tail gas can be realized by means of plasma ionization particles generated by the ionization discharge, so that the treatment requirement of the ship tail gas can be met, but the treatment effect and the treatment efficiency are difficult to meet the requirement. The plasma discharge device is characterized in that in the process of actually carrying out ionization discharge on the ship tail gas, only the current value of the plasma discharge device is fixed in a certain range, the suitability relation between the plasma discharge device and the flow velocity value of the ship tail gas is ignored, so that the current of the actual plasma discharge device can not effectively treat the ship tail gas, and the treatment efficiency and effect are reduced.

Based on this, in the technical scheme of this application, it is expected to regulate and control the current value of plasma discharge equipment adaptively based on the time sequence change condition of the flow velocity of actual ship tail gas to improve the efficiency and the effect of ionization discharge to the ship tail gas. However, since the change information of the flow velocity value of the ship tail gas in the time dimension is hidden information of a small scale, the change rule is difficult to capture in a traditional mode, so that the accuracy of current control of the plasma discharge equipment is low. Therefore, in this process, the difficulty lies in how to fully express the characteristic information of the time sequence dynamic change of the flow velocity value of the ship tail gas, so as to accurately perform the self-adaptive control of the current value of the plasma discharge device in real time based on the actual change condition of the flow velocity of the ship tail gas, thereby improving the ionization discharge efficiency and effect and effectively degrading the harmful components in the ship tail gas.

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 development of deep learning and neural networks provides new solutions and schemes for mining time sequence dynamic change characteristic information of the flow velocity value of the ship tail gas.

Specifically, in the technical scheme of the application, the flow velocity values of the ship tail gas at a plurality of preset time points in a preset time period are obtained. Next, considering that the flow velocity value of the ship exhaust gas has a dynamic change rule in the time dimension, in order to extract the change characteristic information of the flow velocity value of the ship exhaust gas in the time dimension, in the technical scheme of the application, the flow velocity values of the ship exhaust gas at a plurality of preset time points are arranged as flow velocity value time sequence input vectors according to the time dimension, so that the distribution information of the flow velocity values of the ship exhaust gas in the time sequence is integrated.

Further, in order to adaptively and accurately control the current value of the plasma discharge device, it is necessary to extract the dynamic change feature of the flow velocity value of the ship tail gas in the time dimension, and considering that the change information of the flow velocity value of the ship tail gas in the time dimension is weak, the weak change feature is small-scale change feature information relative to the whole flow velocity value of the ship tail gas, if the time-series dynamic change feature extraction of the flow velocity value of the ship tail gas is performed by using absolute static change information, the calculated amount is larger, the overfitting is caused, and the small-scale weak change feature of the flow velocity value of the ship tail gas in the time dimension is difficult to be perceived, so that the accuracy of subsequent decoding is affected.

Based on the above, in the technical scheme of the application, the dynamic characteristic extraction of the flow velocity value of the ship tail gas is comprehensively performed by adopting the time sequence relative dynamic change characteristic and the absolute static change characteristic of the flow velocity value of the ship tail gas. Specifically, first, the difference between the flow velocity values of every adjacent two positions in the flow velocity value timing input vector is calculated to obtain a flow velocity variation timing input vector. Then, it is considered that there is a correlation between the time series relative dynamic change characteristic and the time series absolute static change characteristic of the flow velocity value of the ship exhaust gas with respect to the time series change of the flow velocity of the ship exhaust gas. Therefore, in order to fully explore the time sequence change rule of the flow velocity value of the ship tail gas in the time dimension so as to accurately control the current of the plasma discharge equipment, in the technical scheme of the application, the time sequence input vector of the flow velocity value and the time sequence input vector of the flow velocity change are cascaded to obtain the multi-dimensional input vector of the flow velocity value, so that the association relation between the time sequence relative dynamic change information and the absolute static change information of the flow velocity value of the ship tail gas is established.

Then, it is considered that the flow velocity value of the ship exhaust gas exhibits different pattern change characteristic information at different time period spans within the predetermined period due to the fluctuation and uncertainty in the time dimension. Therefore, in order to fully express the time sequence dynamic change characteristics of the flow velocity value of the ship tail gas, in the technical scheme of the application, the flow velocity value dynamic-static multi-dimensional input vector is further passed through a double-flow model comprising a first convolution neural network model and a second convolution neural network model to obtain a flow velocity value dynamic-static characteristic vector. In particular, here, the first convolutional layer and the second convolutional layer have different feature receptive fields, that is, the first convolutional neural network model uses a one-dimensional convolutional kernel having a first scale, the second convolutional neural network model uses a one-dimensional convolutional kernel having a second scale, and the first scale is not equal to the second scale. In this way, the time sequence multi-scale associated characteristic information of the flow velocity value of the ship tail gas under different time spans can be extracted.

Further, it is considered that since the time-series change of the flow velocity value of the marine exhaust gas is not obvious in the actual monitoring process, it is desirable to perform the characteristic expression enhancement of the time-series multi-scale dynamic change characteristic of the flow velocity of the marine exhaust gas after it is obtained. It should be understood that, as a learning target of the neural network model, the gaussian density map may represent a joint distribution in the case where a plurality of feature values constitute an overall distribution due to probability density thereof, that is, a feature distribution is taken as a priori distribution, and probability density under the influence of correlation with other a priori distribution positions at each a priori distribution position is obtained as a posterior distribution, thereby describing the feature distribution more accurately in a higher dimension. Therefore, in the technical scheme of the application, the time sequence multi-scale correlation characteristic of the flow velocity of the ship tail gas can be enhanced through the prior distribution, namely the Gaussian distribution, of the flow velocity value of the ship tail gas, namely, the dynamic-static characteristic vector of the flow velocity value is enhanced in characteristic expression based on the Gaussian density diagram to obtain the decoding characteristic matrix.

And then, further carrying out decoding regression on the decoding characteristic matrix through a decoder to obtain a current value decoding value for representing the recommended plasma discharge equipment. That is, the decoding regression is performed by using the time sequence multi-scale associated characteristics of the flow velocity of the ship tail gas after the characteristic enhancement, so that the time sequence change condition of the flow velocity value of the ship tail gas is fully described, the self-adaptive control of the current value of the plasma discharge equipment is performed, and the ionization discharge efficiency and effect are improved.

In particular, in the technical solution of the present application, when the gaussian density map is used to perform feature enhancement on the flow velocity value dynamic-static feature vector to obtain the decoding feature matrix, although the gaussian density map uses the variance between every two corresponding feature values of the feature vector as the variance matrix, it is considered that the flow velocity value dynamic-static multi-dimensional input vector includes both the flow velocity value and the flow velocity transformation value, and in the cascading process, the correlation between the flow velocity value and the flow velocity transformation value with remote distribution in the time domain is low, and in the gaussian discretization process, randomness is also introduced, which may cause a problem that there is insufficient correlation between local distributions of the decoding feature matrix, and affects the accuracy of the decoded value obtained by the decoder of the decoding feature matrix.

Based on this, in the technical solution of the present application, the decoded feature matrix is preferably subjected to feature affinity space affine learning to perform optimization, expressed as:

wherein,,

is a diagonal matrix obtained by linear transformation of the decoding feature matrix, or the sampling frequency of Gaussian discretization can be controlled so that the decoding feature matrix is directly a diagonal matrix,/a- >

Representing the two norms of the matrix, i.e. +.>

Maximum eigenvalue of>

Represents the core norms of the matrix, i.e. the sum of the eigenvalues of the matrix, and +.>

Is the dimension of the matrix, i.e. width multiplied by height,/->

Is an optimized diagonal matrix.

Here, the feature affinity spatial affine learning performs affine migration based on spatial transformations with relatively low resolution information characterizations by detailed structured information representation in a low-dimensional eigensubspace of high resolution information characterizations in a feature distribution space of the decoding feature matrix, thereby enabling individual local (i.e., feature value-by-feature value or feature fragment-by-feature fragment) super-resolution activation of feature distributions at global matrix feature distributions based on affinity (affinity) dense simulations between gaussian probability density enhanced characterizations. In this way, the optimized diagonal matrix is further transformed by an inverse transformation corresponding to the linear transformation

And restoring the decoding characteristic matrix to the decoding characteristic matrix, so that the relevance among the local characteristic distribution of the optimized decoding characteristic matrix can be improved, and the accuracy of a decoding value obtained by the decoding characteristic matrix through a decoder is improved. Therefore, the self-adaptive control of the current value of the plasma discharging equipment can be accurately performed in real time based on the actual change condition of the flow velocity of the tail gas of the ship, so that the ionization discharging efficiency and effect are improved, and harmful components in the tail gas of the ship are effectively degraded.

Based on this, the application proposes a ship tail gas treatment method, which comprises the following steps: receiving marine tail gas through a gas supply pipeline; and carrying out ionization discharge on the ship tail gas so as to realize degradation of harmful components in the ship tail gas by means of plasma ionization particles generated by the ionization discharge.

Fig. 1 is an application scenario diagram of a ship tail gas treatment method according to an embodiment of the present application. As shown in fig. 1, in this application scenario, the marine exhaust gas is transmitted to a plasma discharge apparatus (e.g., E as illustrated in fig. 1) through a pipe (e.g., P as illustrated in fig. 1) for exhaust gas treatment, wherein flow velocity values of the marine exhaust gas at a plurality of predetermined time points within a predetermined period of time are acquired by a flow velocity sensor (e.g., V as illustrated in fig. 1). Next, the flow velocity values of the ship exhaust gas at the above-described plurality of predetermined time points are input to a server (e.g., S in fig. 1) in which a ship exhaust gas treatment algorithm is deployed, wherein the server is capable of processing the above-described input information with the ship exhaust gas treatment algorithm to generate a decoded value for representing a recommended current value of the plasma discharge apparatus.

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 method

Fig. 2 is a flow chart of a ship exhaust gas treatment method according to an embodiment of the present application. As shown in fig. 2, the ship tail gas treatment method according to the embodiment of the application includes the steps of: s110, receiving ship tail gas through a gas supply pipeline; and carrying out ionization discharge on the ship tail gas so as to realize degradation of harmful components in the ship tail gas by means of plasma ionization particles generated by the ionization discharge.

Specifically, in step S110, the marine exhaust gas is received through the gas supply pipe. It will be appreciated that in practice the treatment of the marine exhaust gas is performed by using an ionisation discharge of said marine exhaust gas, and before that it is necessary to transfer the marine exhaust gas via a pipeline to a plasma discharge device.

Specifically, in step S120, the marine exhaust gas is subjected to ionization discharge, so that degradation of harmful components in the marine exhaust gas is realized by means of plasma ionized particles generated by the ionization discharge. Particularly, the plasma discharge principle is to decompose pollutants in the ship tail gas by means of plasma generated by ionization, so that the aim of degrading harmful components in the tail gas is fulfilled. The mode can be suitable for various harmful gases, and the degradation rate is high.

Fig. 3 is a flowchart of an ionization discharge process for ship exhaust in the ship exhaust treatment method according to an embodiment of the present application. As shown in fig. 3, during the ionization discharge, it includes: s210, acquiring flow velocity values of ship tail gas at a plurality of preset time points in a preset time period; s220, arranging the flow velocity values of the ship tail gas at a plurality of preset time points into flow velocity value time sequence input vectors according to the time dimension; s230, calculating the difference value between the flow velocity values of every two adjacent positions in the flow velocity value time sequence input vector to obtain a flow velocity change time sequence input vector; s240, cascading the flow velocity value time sequence input vector and the flow velocity change time sequence input vector to obtain a flow velocity value dynamic-static multidimensional input vector; s250, the flow velocity value dynamic-static multidimensional input vector is passed through a double-flow model comprising a first convolution neural network model and a second convolution neural network model to obtain a flow velocity value dynamic-static characteristic vector; s260, performing feature enhancement on the flow velocity value dynamic-static feature vector by using a Gaussian density chart to obtain a decoding feature matrix; s270, performing feature distribution optimization on the decoding feature matrix to obtain an optimized decoding feature matrix; and S280, carrying out decoding regression on the optimized decoding characteristic matrix through a decoder to obtain a decoding value, wherein the decoding value is used for representing the recommended current value of the plasma discharge equipment.

Fig. 4 is a schematic diagram of a ship tail gas treatment method according to an embodiment of the present application. As shown in fig. 4, in the network structure, first, flow velocity values of the ship exhaust gas at a plurality of predetermined time points within a predetermined period of time are acquired; arranging the flow velocity values of the ship tail gas at a plurality of preset time points into flow velocity value time sequence input vectors according to the time dimension; then, calculating the difference value between the flow velocity values of every two adjacent positions in the flow velocity value time sequence input vector to obtain a flow velocity change time sequence input vector; cascading the flow velocity value time sequence input vector and the flow velocity change time sequence input vector to obtain a flow velocity value dynamic-static multidimensional input vector; the flow velocity value dynamic-static multidimensional input vector is passed through a double-flow model comprising a first convolution neural network model and a second convolution neural network model to obtain a flow velocity value dynamic-static characteristic vector; then, carrying out characteristic enhancement on the dynamic-static characteristic vector of the flow velocity value by using a Gaussian density chart to obtain a decoding characteristic matrix; performing feature distribution optimization on the decoding feature matrix to obtain an optimized decoding feature matrix; and further, carrying out decoding regression on the optimized decoding characteristic matrix through a decoder to obtain a decoding value, wherein the decoding value is used for representing the recommended current value of the plasma discharge equipment.

More specifically, in step S210, flow velocity values of the marine exhaust gas at a plurality of predetermined time points within a predetermined period of time are acquired. In the technical scheme of the application, the current value of the plasma discharge device is expected to be adaptively regulated and controlled based on the time sequence change condition of the flow velocity of the actual ship tail gas, so that the efficiency and the effect of ionization discharge on the ship tail gas are improved. Thus, in one specific example of the present application, first, the marine exhaust gas is transferred to the plasma discharge apparatus through the pipe for exhaust gas treatment, wherein the flow rate values of the marine exhaust gas at a plurality of predetermined time points within a predetermined period of time are acquired by the flow rate sensor.

More specifically, in step S220, the flow velocity values of the ship exhaust gas at the plurality of predetermined time points are arranged in a time dimension as a flow velocity value timing input vector. In the technical scheme of the application, the flow velocity values of the ship tail gas at a plurality of preset time points are arranged into flow velocity value time sequence input vectors according to the time dimension so as to integrate the distribution information of the flow velocity values of the ship tail gas in time sequence.

More specifically, in step S230 and step S240, the difference between the flow velocity values of each adjacent two positions in the flow velocity value timing input vector is calculated to obtain a flow velocity variation timing input vector, and then the flow velocity value dynamic-static multi-dimensional input vector is passed through a dual-flow model including a first convolutional neural network model and a second convolutional neural network model to obtain a flow velocity value dynamic-static feature vector. In consideration of the fact that the flow velocity value of the ship tail gas has different change characteristics in the time dimension, the dynamic change characteristics of the flow velocity value of the ship tail gas in the time dimension are required to be extracted, but because the change information of the flow velocity value of the ship tail gas in the time sequence is weak and the weak change characteristics are small-scale change characteristic information relative to the whole flow velocity value of the ship tail gas, in the technical scheme of the application, the dynamic characteristic extraction of the flow velocity value of the ship tail gas can be comprehensively carried out by adopting the time sequence relative dynamic change characteristics and the absolute static change characteristics of the flow velocity value of the ship tail gas. Specifically, first, the flow rate change timing input vector may be obtained by calculating the difference between the flow rate values of every adjacent two positions in the flow rate value timing input vector. Then, it is considered that there is a correlation between the time series relative dynamic change characteristic and the time series absolute static change characteristic of the flow velocity value of the ship exhaust gas with respect to the time series change of the flow velocity of the ship exhaust gas. Therefore, in the technical scheme of the application, the flow velocity value time sequence input vector and the flow velocity change time sequence input vector are cascaded to obtain the flow velocity value dynamic-static multidimensional input vector so as to extract the time sequence change rule of the flow velocity value of the ship tail gas in the time dimension, and further accurately control the current of the plasma discharge equipment.

More specifically, in step S250, the flow velocity value dynamic-static multi-dimensional input vector is passed through a dual-flow model including a first convolutional neural network model and a second convolutional neural network model to obtain a flow velocity value dynamic-static feature vector. Considering that the flow velocity value of the ship exhaust gas has fluctuation and uncertainty in the time dimension, the flow velocity value of the ship exhaust gas presents different pattern change characteristic information in different time period spans in the preset time period. Therefore, in order to fully express the time sequence dynamic change characteristics of the flow velocity value of the ship tail gas, in the technical scheme of the application, the flow velocity value dynamic-static multi-dimensional input vector is further passed through a double-flow model comprising a first convolution neural network model and a second convolution neural network model to obtain a flow velocity value dynamic-static characteristic vector. In particular, here, the first convolutional layer and the second convolutional layer have different feature receptive fields, that is, the first convolutional neural network model uses a one-dimensional convolutional kernel having a first scale, the second convolutional neural network model uses a one-dimensional convolutional kernel having a second scale, and the first scale is not equal to the second scale. In one example, the convolution kernel determines a local feature receptive field of the convolutional neural network model. In this way, the time sequence multi-scale associated characteristic information of the flow velocity value of the ship tail gas under different time spans can be extracted. In one example, the first convolutional neural network and the second convolutional neural network comprise a plurality of neural network layers that are cascaded with each other, wherein each neural network layer comprises a convolutional layer, a pooling layer, and an activation layer. In the encoding process of the first convolutional neural network and the second convolutional neural network, each layer of the first convolutional neural network and the second convolutional neural network performs convolutional processing based on a convolutional kernel on input data by using the convolutional layer, performs pooling processing on a convolutional feature map output by the convolutional layer by using the pooling layer and performs activation processing on the pooled feature map output by the pooling layer by using the activation layer in the forward transfer process of the layers.

Fig. 5 is a flow chart of dual-flow model coding in a ship tail gas treatment method according to an embodiment of the present application. As shown in fig. 5, in the process of the dual stream model encoding, it includes: s310, performing convolution processing, pooling processing and nonlinear activation processing based on a one-dimensional convolution kernel of a first scale on the flow velocity value dynamic-static multidimensional input vector in forward transfer of layers by using each layer of a first convolution neural network of the double-flow model to output the first-scale flow velocity value dynamic-static feature vector by the last layer of the first convolution neural network; s320, respectively performing convolution processing, pooling processing and nonlinear activation processing based on a one-dimensional convolution kernel of a second scale on the flow velocity value dynamic-static multi-dimensional input vector in forward transmission of layers by using each layer of a second convolution neural network of the double-flow model to output a flow velocity value dynamic-static characteristic vector of the second scale by the last layer of the second convolution neural network, wherein the first scale is not equal to the second scale; and S330, cascading the first scale flow velocity value dynamic-static characteristic vector and the second scale flow velocity value dynamic-static characteristic vector to obtain the flow velocity value dynamic-static characteristic vector.

More specifically, in step S260, the flow velocity value dynamic-static feature vector is feature enhanced using a gaussian density map to obtain a decoded feature matrix. In consideration of the fact that the time sequence change of the flow velocity value of the ship tail gas is not obvious in the actual monitoring process, the time sequence multi-scale dynamic change characteristic of the flow velocity of the ship tail gas is obtained, and then the characteristic expression is expected to be enhanced. It should be understood that, as a learning target of the neural network model, the gaussian density map may represent a joint distribution of individual feature values of the feature distribution in the case where a plurality of feature values constitute an overall distribution due to probability density thereof, that is, the feature distribution is taken as a priori distribution, and probability density under the effect of correlation due to other a priori distribution positions at each a priori distribution position is obtained as a posterior distribution, thereby being more in a higher dimensionThe feature distribution is accurately described. Therefore, in the technical scheme of the application, the time sequence multi-scale correlation characteristic of the flow velocity of the ship tail gas can be enhanced through the prior distribution, namely the Gaussian distribution, of the flow velocity value of the ship tail gas, namely, the dynamic-static characteristic vector of the flow velocity value is enhanced in characteristic expression based on the Gaussian density diagram to obtain the decoding characteristic matrix. Specifically, first, a gaussian density map of the flow velocity value dynamic-static feature vector is constructed, specifically, a gaussian density map of the flow velocity value dynamic-static feature vector is constructed in the following formula; wherein, the formula is:

Wherein->

Represents the dynamic-static characteristic vector of the flow velocity value, and +.>

Representing the variance between the eigenvalues of the corresponding two locations in the flow velocity value dynamic-static eigenvector; and then, carrying out Gaussian discretization on the Gaussian distribution of each position in the Gaussian density map to obtain the decoding characteristic matrix.

Fig. 6 is a flow chart of gaussian intensity in a ship exhaust gas treatment method according to an embodiment of the present application. As shown in fig. 6, in the gaussian enhancement process, it includes: s410, constructing a Gaussian density map of the dynamic-static characteristic vector of the flow velocity value to obtain the Gaussian density map. The mean value vector of the Gaussian density map is the flow velocity value dynamic-static characteristic vector, and the covariance matrix of the Gaussian density map is the variance between the characteristic values of the corresponding two positions in the flow velocity value dynamic-static characteristic vector; and S420, performing Gaussian discretization on the Gaussian distribution of each position in the Gaussian density map to obtain the decoding feature matrix.

More specifically, in step S270, the feature distribution optimization is performed on the decoding feature matrix to obtain an optimized decoding feature matrix. In particular, in the technical solution of the present application, when the gaussian density map is used to perform feature enhancement on the flow velocity value dynamic-static feature vector to obtain the decoding feature matrix, although the gaussian density map uses the variance between every two corresponding feature values of the feature vector as the variance matrix, it is considered that the flow velocity value dynamic-static multi-dimensional input vector includes both the flow velocity value and the flow velocity transformation value, and in the cascading process, the correlation between the flow velocity value and the flow velocity transformation value with remote distribution in the time domain is low, and in the gaussian discretization process, randomness is also introduced, which may cause a problem that there is insufficient correlation between local distributions of the decoding feature matrix, and affects the accuracy of the decoded value obtained by the decoder of the decoding feature matrix. Based on this, in the technical solution of the present application, the decoded feature matrix is preferably subjected to feature affinity space affine learning to perform optimization, expressed as:

Wherein,,

Representing the two norms of the decoding feature matrix, < >>

Represents the core norms of the decoding feature matrix, and +.>

Is the scale of the decoding feature matrix, +.>

Represents a logarithmic function value based on 2, < +.>

An exponential operation representing a matrix, the exponential operation representing a calculationNatural exponential function value raised to power by eigenvalues at each position in matrix, < >>

Representing the multiplication by the position point,

representing the optimized decoding feature matrix. Here, the feature affinity spatial affine learning performs affine migration based on spatial transformations with relatively low resolution information characterizations by detailed structured information representation in a low-dimensional eigensubspace of high resolution information characterizations in a feature distribution space of the decoding feature matrix, thereby enabling individual local (i.e., feature value-by-feature value or feature fragment-by-feature fragment) super-resolution activation of feature distributions at global matrix feature distributions based on affinity (affinity) dense simulations between gaussian probability density enhanced characterizations. Thus, the optimized diagonal matrix is ++again by the inverse transformation corresponding to the linear transformation >

More specifically, in step S280, the optimized decoding feature matrix is subjected to decoding regression by a decoder to obtain a decoded value, which is used to represent a recommended current value of the plasma discharge apparatus. That is, after obtaining the optimized decoding feature matrix, performing decoding regression on the optimized decoding feature matrix as a decoding feature matrix by a decoder to obtain a decoded value representing a current value of the recommended plasma discharge apparatus, particularly, the feature-enhanced information on the ship tailThe time sequence multi-scale correlation characteristic of the flow velocity of the gas is used for decoding regression, so that the time sequence change condition of the flow velocity value of the ship tail gas is fully described, the self-adaptive control of the current value of the plasma discharge equipment is performed, and the ionization discharge efficiency and effect are improved. Specifically, the decoder is used to perform decoding regression on the optimized decoding feature matrix in the following formula to obtain a decoding value for representing the recommended current value of the plasma discharge device; wherein, the formula is:

Wherein->

Representing the optimized decoding feature matrix,/a>

Is the decoded value,/->

Is a weight matrix, < >>

Representing matrix multiplication.

In summary, the ship tail gas treatment method according to the embodiment of the application is clarified, by adopting a neural network model based on deep learning to excavate the time sequence dynamic change characteristic of the flow velocity value of the ship tail gas, the self-adaptive control of the current value of the plasma discharge equipment is accurately performed in real time based on the actual change condition of the flow velocity of the ship tail gas, and therefore the ionization discharge efficiency and effect are improved, and harmful components in the ship tail gas are effectively degraded.

Exemplary System

Fig. 7 is a block diagram of a marine vessel exhaust treatment system according to an embodiment of the present application. As shown in fig. 7, a marine vessel exhaust treatment system 300 according to an embodiment of the present application includes: a data acquisition unit 310; an arrangement unit 320; a difference value calculation unit 330; a cascade unit 340; a multi-scale convolution unit 350; a gaussian enhancement unit 360; a feature distribution optimizing unit 370; and a decoding unit 380.

The data acquisition unit 310 is configured to acquire flow velocity values of the ship tail gas at a plurality of predetermined time points in a predetermined time period; the arrangement unit 320 is configured to arrange the flow velocity values of the ship tail gas at the plurality of predetermined time points into flow velocity value time sequence input vectors according to a time dimension; the difference calculating unit 330 is configured to calculate a difference between the flow velocity values of each two adjacent positions in the flow velocity value time sequence input vector to obtain a flow velocity variation time sequence input vector; the cascade unit 340 is configured to cascade the flow velocity value time sequence input vector and the flow velocity variation time sequence input vector to obtain a flow velocity value dynamic-static multidimensional input vector; the multi-scale convolution unit 350 is configured to pass the flow velocity value dynamic-static multi-dimensional input vector through a dual-flow model including a first convolutional neural network model and a second convolutional neural network model to obtain a flow velocity value dynamic-static feature vector; the gaussian enhancement unit 360 is configured to perform feature enhancement on the flow velocity value dynamic-static feature vector by using a gaussian density map to obtain a decoded feature matrix; the feature distribution optimizing unit 370 is configured to perform feature distribution optimization on the decoding feature matrix to obtain an optimized decoding feature matrix; and the decoding unit 380 is configured to perform decoding regression on the optimized decoding feature matrix through a decoder to obtain a decoded value, where the decoded value is used to represent a recommended current value of the plasma discharge device.

In one example, in the marine exhaust gas treatment system 300 described above, the multi-scale convolution unit 350 is configured to: performing convolution processing, pooling processing and nonlinear activation processing based on a one-dimensional convolution kernel of a first scale on the flow velocity value dynamic-static multidimensional input vector in forward transfer of layers by using each layer of a first convolution neural network of the double-flow model to output the first scale flow velocity value dynamic-static characteristic vector by the last layer of the first convolution neural network; performing convolution processing, pooling processing and nonlinear activation processing based on a one-dimensional convolution kernel of a second scale on the flow velocity value dynamic-static multidimensional input vector in forward transfer of layers by using layers of a second convolution neural network of the double-flow model to output a second scale flow velocity value dynamic-static feature vector by a last layer of the second convolution neural network, wherein the first scale is not equal to the second scale; and concatenating the first scale flow velocity value dynamic-static feature vector and the second scale flow velocity value dynamic-static feature vector to obtain the flow velocity value dynamic-static feature vector.

In one example, in the above marine exhaust gas treatment system 300, the gaussian enhancement unit 360 is configured to: and constructing a Gaussian density map of the dynamic-static characteristic vector of the flow velocity value to obtain the Gaussian density map. The mean value vector of the Gaussian density map is the flow velocity value dynamic-static characteristic vector, and the covariance matrix of the Gaussian density map is the variance between the characteristic values of the corresponding two positions in the flow velocity value dynamic-static characteristic vector; and performing Gaussian discretization on the Gaussian distribution of each position in the Gaussian density map to obtain the decoding characteristic matrix.

In one example, in the above-mentioned marine exhaust gas treatment system 300, the feature distribution optimizing unit 370 is configured to: performing feature affinity space affine learning optimization on the decoding feature matrix by using the following optimization formula to obtain the optimized decoding feature matrix; wherein, the optimization formula is:

wherein,,

Representing the two norms of the decoding feature matrix, < >>

Represents the core norms of the decoding feature matrix, and +.>

Is the decoding featureDimension of matrix->

Represents a logarithmic function value based on 2, < +.>

Representing the multiplication by the position point,

representing the optimized decoding feature matrix.

In one example, in the marine exhaust gas treatment system 300, the decoding unit 380 is configured to: performing decoding regression on the optimized decoding feature matrix using the decoder in the following formula to obtain a decoded value representing a recommended current value of the plasma discharge apparatus; wherein, the formula is:

Wherein->

Representing the optimized decoding feature matrix,/a>

Is the decoded value,/->

Is a weight matrix, < >>

Representing matrix multiplication.

In summary, the ship tail gas treatment system 300 according to the embodiment of the present application is illustrated, which excavates the time sequence dynamic variation characteristic of the flow velocity value of the ship tail gas by adopting the neural network model based on deep learning, so as to accurately perform the adaptive control of the current value of the plasma discharge device in real time based on the actual change condition of the flow velocity of the ship tail gas, thereby improving the ionization discharge efficiency and effect, and effectively degrading the harmful components in the ship tail gas.

As described above, the ship exhaust gas treatment system according to the embodiment of the present application may be implemented in various terminal devices. In one example, the marine vessel exhaust treatment system 300 according to embodiments of the present application may be integrated into the terminal equipment as one software module and/or hardware module. For example, the marine exhaust gas treatment system 300 may be a software module in the operating system of the terminal device, or may be an application developed for the terminal device; of course, the marine exhaust gas treatment system 300 may equally be one of a number of hardware modules of the terminal equipment.

Alternatively, in another example, the marine exhaust gas treatment system 300 and the terminal device may be separate devices, and the marine exhaust gas treatment system 300 may be connected to the terminal device via a wired and/or wireless network and communicate the interaction information in accordance with a agreed data format.

Exemplary electronic device

Next, an electronic device according to an embodiment of the present application is described with reference to fig. 8.

Fig. 8 illustrates a block diagram of an electronic device according to an embodiment of the present application.

As shown in fig. 8, 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 treatment methods of the various embodiments of the present application described above and/or other desired functions. Various contents such as a classification feature 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 can output various information including a decoded value 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. 8 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 of the marine exhaust gas treatment method according to the various embodiments of the present application described in the "exemplary methods" 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 the flow velocity value dynamic-static feature vector according to various embodiments of the present application described in the above-mentioned "exemplary method" 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

1. A method for treating marine exhaust gas, comprising:

receiving marine tail gas through a gas supply pipeline; and

Carrying out ionization discharge on the ship tail gas so as to realize degradation of harmful components in the ship tail gas by means of plasma ionized particles generated by the ionization discharge;

wherein, carry out ionization discharge to boats and ships tail gas includes:

acquiring flow velocity values of ship tail gas at a plurality of preset time points in a preset time period;

arranging the flow velocity values of the ship tail gas at a plurality of preset time points into flow velocity value time sequence input vectors according to the time dimension;

calculating the difference value between the flow velocity values of every two adjacent positions in the flow velocity value time sequence input vector to obtain a flow velocity change time sequence input vector;

cascading the flow velocity value time sequence input vector and the flow velocity change time sequence input vector to obtain a flow velocity value dynamic-static multidimensional input vector;

the flow velocity value dynamic-static multidimensional input vector is passed through a double-flow model comprising a first convolution neural network model and a second convolution neural network model to obtain a flow velocity value dynamic-static characteristic vector;

performing feature enhancement on the dynamic-static feature vector of the flow velocity value by using a Gaussian density chart to obtain a decoding feature matrix;

performing feature distribution optimization on the decoding feature matrix to obtain an optimized decoding feature matrix; and

Performing decoding regression on the optimized decoding feature matrix through a decoder to obtain a decoding value, wherein the decoding value is used for representing a current value of recommended plasma discharge equipment;

wherein the flow velocity value dynamic-static multidimensional input vector is passed through a double-flow model comprising a first convolution neural network model and a second convolution neural network model to obtain a flow velocity value dynamic-static characteristic vector, comprising:

performing convolution processing, pooling processing and nonlinear activation processing based on a one-dimensional convolution kernel of a first scale on the flow velocity value dynamic-static multidimensional input vector in forward transfer of layers by using each layer of a first convolution neural network of the double-flow model to output the first scale flow velocity value dynamic-static characteristic vector by the last layer of the first convolution neural network;

performing convolution processing, pooling processing and nonlinear activation processing based on a one-dimensional convolution kernel of a second scale on the flow velocity value dynamic-static multidimensional input vector in forward transfer of layers by using layers of a second convolution neural network of the double-flow model to output a second scale flow velocity value dynamic-static feature vector by a last layer of the second convolution neural network, wherein the first scale is not equal to the second scale; and

And cascading the first scale flow velocity value dynamic-static characteristic vector and the second scale flow velocity value dynamic-static characteristic vector to obtain the flow velocity value dynamic-static characteristic vector.

2. The ship exhaust gas treatment method according to claim 1, wherein the feature enhancement of the flow velocity value dynamic-static feature vector using a gaussian density map to obtain a decoded feature matrix comprises:

constructing a Gaussian density map of the flow velocity value dynamic-static characteristic vector to obtain a Gaussian density map, wherein the mean value vector of the Gaussian density map is the flow velocity value dynamic-static characteristic vector, and the covariance matrix of the Gaussian density map is the variance between the characteristic values of two corresponding positions in the flow velocity value dynamic-static characteristic vector; and

and carrying out Gaussian discretization on the Gaussian distribution of each position in the Gaussian density map to obtain the decoding characteristic matrix.

3. The ship exhaust gas treatment method according to claim 2, wherein optimizing the feature distribution of the decoding feature matrix to obtain an optimized decoding feature matrix comprises:

performing feature affinity space affine learning optimization on the decoding feature matrix by using the following optimization formula to obtain the optimized decoding feature matrix;

Wherein, the optimization formula is:

wherein,,

Representing the two norms of the decoding feature matrix, < >>

Represents the core norms of the decoding feature matrix, and +.>

Is the scale of the decoding feature matrix, +.>

Represents a logarithmic function value based on 2, < +.>

Representing multiplication by location +.>

Representing the optimized decoding feature matrix.

4. A ship exhaust gas treatment method according to claim 3, wherein the optimizing decoding feature matrix is subjected to decoding regression by a decoder to obtain a decoded value, the decoded value being used to represent a recommended current value of the plasma discharge apparatus, comprising: performing decoding regression on the optimized decoding feature matrix using the decoder in the following formula to obtain a decoded value representing a recommended current value of the plasma discharge apparatus;

wherein,,the formula is:

wherein->

Representing the optimized decoding feature matrix,/a>

Is the decoded value,/->

Is a weight matrix, < > >

Representing matrix multiplication.

5. A marine vessel exhaust gas treatment system, comprising:

The decoding unit is used for carrying out decoding regression on the optimized decoding characteristic matrix through a decoder to obtain a decoding value, wherein the decoding value is used for representing a current value of the recommended plasma discharge equipment;

wherein, the plasma discharge equipment is used for carrying out ionization discharge on the ship tail gas so as to realize degradation of harmful components in the ship tail gas by means of plasma ionization particles generated by the ionization discharge;

wherein the multi-scale convolution unit is configured to:

6. The marine exhaust gas treatment system according to claim 5, wherein the gaussian enhancement unit is configured to:

7. The marine vessel exhaust gas treatment system according to claim 6, wherein the feature distribution optimizing unit is configured to:

wherein, the optimization formula is:

wherein,,

diagonal matrix representing the decoding feature matrix by linear transformation,/a >

Representing the two norms of the decoding feature matrix, < >>

Represents the core norms of the decoding feature matrix, and +.>

Is the scale of the decoding feature matrix, +.>

Represents a logarithmic function value based on 2, < +.>

Representation matrixThe exponential operation of the matrix representing the calculation of a natural exponential function value raised to the eigenvalue of each position in the matrix, ">

Representing multiplication by location +.>

Representing the optimized decoding feature matrix.