CN115754107B

CN115754107B - Automatic sampling analysis system and method for lithium hexafluorophosphate preparation

Info

Publication number: CN115754107B
Application number: CN202211391469.3A
Authority: CN
Inventors: 张志强; 华杭州; 吴远胜; 陈东林
Original assignee: Fujian Longde New Energy Co ltd
Current assignee: Fujian Longde New Energy Co ltd
Priority date: 2022-11-08
Filing date: 2022-11-08
Publication date: 2023-06-23
Anticipated expiration: 2042-11-08
Also published as: CN115754107A; WO2024098653A1

Abstract

The application relates to the field of sampling analysis of lithium hexafluorophosphate, and particularly discloses an automatic sampling analysis system and a method for preparing lithium hexafluorophosphate.

Description

Automatic sampling analysis system and method for lithium hexafluorophosphate preparation

Technical Field

The present application relates to the field of sample analysis of lithium hexafluorophosphate, and more particularly, to an automated sample analysis system for lithium hexafluorophosphate preparation and a method thereof.

Background

The lithium ion battery has the characteristics of high energy density, high specific power, good cycle performance, no memory effect, no pollution and the like, is widely applied to electronic digital products at present, and is also an ideal choice of energy sources of electric automobiles in the future.

The battery electrolyte consists of an organic solvent and a lithium salt electrolyte, and the lithium salt electrolyte commonly used at present comprises lithium perchlorate, lithium hexafluorophosphate, lithium tetrafluoroborate and the like, wherein the lithium hexafluorophosphate has good conductivity and electrochemical stability, and is the lithium salt electrolyte with the widest application range at present.

Since the quality of lithium hexafluorophosphate directly affects the quality of the electrolyte and the battery performance applied in lithium ion batteries, monitoring of the quality of lithium hexafluorophosphate is critical to the electrolyte. However, most of the existing quality inspection schemes are to perform experimental analysis on the prepared lithium hexafluorophosphate to compare the reaction result with the standard reference mass, so as to determine the quality of the prepared lithium hexafluorophosphate. Therefore, a large amount of raw materials and resources are wasted, more variable factors exist in the experimental process, the reaction result is difficult to control, and further the accuracy of lithium hexafluorophosphate quality inspection is difficult to guarantee.

Accordingly, an optimized automated sampling analysis system for lithium hexafluorophosphate preparation 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 an automatic sampling analysis system and a method thereof for preparing lithium hexafluorophosphate, which adopts an artificial intelligent detection algorithm based on deep learning to extract multi-scale implicit characteristic distribution information with different sizes in a noise-reduced sampling gas chromatogram, and further strengthens the influence on detecting the content of small-scale phosphorus trifluoride oxide in the gas chromatogram by using an attention mechanism and filtering useless interference characteristics so as to improve the accuracy of detecting the content of phosphorus trifluoride oxide. Therefore, the content of the phosphorus trifluoride oxide can be intelligently and accurately detected, so that the quality inspection accuracy of the prepared lithium hexafluorophosphate is improved, and the quality of the electrolyte is improved, and the electrolyte is applied to the battery performance of a lithium ion battery.

According to one aspect of the present application, there is provided an automated sampling analysis system for lithium hexafluorophosphate preparation, comprising: the sampling gas data acquisition module is used for acquiring a gas phase chromatogram of the sampling gas; the noise reduction module is used for enabling the gas chromatogram to pass through the noise reduction module based on the automatic coder-decoder to obtain a noise-reduced gas chromatogram; the chromatographic characteristic extraction module is used for enabling the denoised gas chromatograph to pass through a convolutional neural network model comprising a plurality of mixed convolutional layers so as to obtain a denoised gas chromatograph characteristic diagram; the code compensation module is used for performing feature clustering defocusing fuzzy optimization on the noise-reduced gas chromatography feature map so as to obtain an optimized noise-reduced gas chromatography feature map; the feature distribution enhancement module is used for enabling the optimized noise reduction gas chromatography feature map to pass through the residual double-attention mechanism module to obtain an enhanced gas chromatography feature map as a decoding feature map; and the analysis result generation module is used for enabling the decoding characteristic diagram to pass through a decoder to obtain a decoding value, wherein the decoding value is used for representing the content value of the phosphorus trifluoride oxide in the sampling gas.

According to another aspect of the present application, there is provided an automated sampling analysis method for lithium hexafluorophosphate preparation, comprising: acquiring a gas chromatograph of the sampled gas; the gas chromatogram is passed through a noise reduction module based on an automatic coder-decoder to obtain a noise-reduced gas chromatogram; the denoised gas chromatograph is processed through a convolutional neural network model comprising a plurality of mixed convolutional layers to obtain a denoised gas chromatograph characteristic diagram; performing feature clustering defocusing fuzzy optimization on the noise-reduced gas chromatography feature map to obtain an optimized noise-reduced gas chromatography feature map; the optimized noise reduction gas chromatography feature map is passed through a residual error dual-attention mechanism module to obtain an enhanced gas chromatography feature map as a decoding feature map; and passing the decoded signature through a decoder to obtain a decoded value, wherein the decoded value is used for representing the content value of the phosphorus oxytrifluoride in the sampling gas.

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 an automated sample analysis method for lithium hexafluorophosphate preparation as described above.

According to yet another 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 an automated sample analysis method for lithium hexafluorophosphate preparation as described above.

Compared with the prior art, the automatic sampling analysis system and the method for preparing lithium hexafluorophosphate, provided by the application, adopt an artificial intelligent detection algorithm based on deep learning to extract multi-scale implicit characteristic distribution information with different sizes in a noise-reduced sampling gas chromatogram, and further strengthen the influence on detecting the content of small-scale phosphorus trifluoride oxide in the gas chromatogram by using an attention mechanism to filter useless interference characteristics so as to improve the accuracy of detecting the content of phosphorus trifluoride oxide. Therefore, the content of the phosphorus trifluoride oxide can be intelligently and accurately detected, so that the quality inspection accuracy of the prepared lithium hexafluorophosphate is improved, and the quality of the electrolyte is improved, and the electrolyte is applied to the battery performance of a lithium ion battery.

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 illustrates an application scenario diagram of an automated sampling analysis system for lithium hexafluorophosphate preparation and methods thereof according to embodiments of the present application.

Fig. 2 illustrates a block diagram schematic of an automated sampling analysis system for lithium hexafluorophosphate preparation according to an embodiment of the present application.

Fig. 3 illustrates a block diagram of a noise reduction module in an automated sampling analysis system for lithium hexafluorophosphate preparation according to an embodiment of the present application.

Fig. 4 illustrates a block diagram of a feature enhancement module in an automated sampling analysis system for lithium hexafluorophosphate preparation according to an embodiment of the present application.

Fig. 5 illustrates a block diagram of a spatial attention unit in an automated sampling analysis system for lithium hexafluorophosphate preparation according to an embodiment of the present application.

Fig. 6 illustrates a block diagram of a channel attention unit in an automated sampling analysis system for lithium hexafluorophosphate preparation according to an embodiment of the present application.

Fig. 7 illustrates a flow chart of an automated sampling analysis system for lithium hexafluorophosphate preparation according to an embodiment of the present application.

Fig. 8 illustrates a schematic diagram of a system architecture of an automated sampling analysis system for lithium hexafluorophosphate preparation according to an embodiment of the present application.

Fig. 9 illustrates 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, the battery electrolyte is composed of an organic solvent and a lithium salt electrolyte, and the lithium salt electrolyte commonly used at present is lithium perchlorate, lithium hexafluorophosphate, lithium tetrafluoroborate, and the like, wherein the lithium hexafluorophosphate has good conductivity and electrochemical stability, and is the lithium salt electrolyte with the widest application range at present.

Since the quality of lithium hexafluorophosphate directly affects the quality of the electrolyte and the battery performance applied in lithium ion batteries, monitoring of the quality of lithium hexafluorophosphate is critical to the electrolyte. However, most of the existing quality inspection schemes are to perform experimental analysis on the prepared lithium hexafluorophosphate to compare the reaction result with the standard reference mass, so as to determine the quality of the prepared lithium hexafluorophosphate. Therefore, a large amount of raw materials and resources are wasted, more variable factors exist in the experimental process, the reaction result is difficult to control, and further the accuracy of lithium hexafluorophosphate quality inspection is difficult to guarantee. Accordingly, an optimized automated sampling analysis system for lithium hexafluorophosphate preparation is desired.

According to the technical problem, a lithium hexafluorophosphate solution sample is stored in a sealed container, one fifth to one third of the volume is reserved in the sealed container to contain volatile gas, after the sample is placed for a period of time, the volatile gas above the lithium hexafluorophosphate solution in the container is sampled, and then a gas chromatograph is used for quantitatively analyzing phosphorus trifluoride oxide in the sampled gas component, so that the quality of the lithium hexafluorophosphate solution is judged according to the phosphorus trifluoride oxide content. Therefore, in order to ensure the quality inspection accuracy of the prepared lithium hexafluorophosphate, the content of phosphorus oxytrifluoride needs to be accurately detected.

At present, deep learning and neural networks have been widely used in the fields of computer vision, natural language processing, speech signal processing, and the like. In addition, deep learning and neural networks have also shown levels approaching and even exceeding humans in the fields of image classification, object detection, semantic segmentation, text translation, and the like.

In recent years, deep learning and development of a neural network provide new solutions and schemes for intelligent detection of the content of phosphorus oxytrifluoride.

Specifically, in the technical scheme of the application, the artificial intelligent detection algorithm based on deep learning is adopted to extract multi-scale implicit characteristic distribution information of different sizes in the noise-reduced sampling gas chromatogram, and an attention mechanism is utilized to further strengthen the influence of filtering useless interference characteristics on the detection of the small-scale trifluoro-phosphorus oxide content in the gas chromatogram so as to improve the accuracy of the trifluoro-phosphorus oxide content detection. Therefore, the content of the phosphorus trifluoride oxide can be intelligently and accurately detected, so that the quality inspection accuracy of the prepared lithium hexafluorophosphate is improved, and the quality of the electrolyte is improved, and the electrolyte is applied to the battery performance of a lithium ion battery.

Specifically, in the technical scheme of the application, firstly, a gas chromatograph of a sampling gas is obtained, and the gas chromatograph is subjected to noise reduction processing in a noise reduction module based on an automatic coder-decoder to obtain a noise-reduced gas chromatograph. It should be understood that in the process of collecting the gas chromatogram of the sampled gas by the actual gas chromatograph, the accuracy of determining the content of phosphorus trifluoride oxide in the gas chromatogram is low due to the influence of the environmental gas factors in the air and other gas factors generated by the gas chromatograph, so that the quality detection of lithium hexafluorophosphate is affected. Therefore, in the technical solution of the present application, a noise reduction module of an automatic codec is further required to perform noise reduction processing on the gas chromatogram to obtain a denoised gas chromatogram.

Then, considering that the phosphorus trifluoride oxide is small in size for the sampling gas, the phosphorus trifluoride oxide also belongs to small-scale features in the gas chromatogram, so in the technical scheme of the application, the denoised gas chromatogram is selectively processed through a mixed convolution layer to extract multi-scale implicit correlation features of the denoised gas chromatogram, and thus a denoised gas chromatogram feature map is obtained. That is, in a specific example of the present application, in the hybrid convolution layer (mixed convolution layer, MCL), the design of the module includes four branches connected in parallel, and is composed of a common convolution layer with a convolution kernel size of 3×3 and a cavity convolution layer with a convolution kernel size of 3×3, the input feature images are respectively operated, the expansion rates of the three branches of the cavity convolution are respectively set to 2, 3 and 4, and the image information of different receptive fields can be obtained through the setting of different expansion rates, so that feature images with different scales can be obtained, the receptive fields are enlarged, meanwhile, the downsampling loss information is avoided, and then the 4 branch feature images are fused, so that the sampling is more dense, the high-level features are possessed, and the additional parameter number is not increased.

Further, considering that the phosphorus trifluoride oxide belongs to a small-sized object (i.e., its proportion in the gas chromatogram of the entire sampling gas is small), secondly, the sampling gas has various gas types, which may interfere with the target detection of the phosphorus trifluoride oxide. Therefore, in the technical scheme of the application, the characteristic data of the noise-reduced gas chromatography characteristic diagram is further enhanced to obtain an enhanced gas chromatography characteristic diagram.

Specifically, the noise-reduced gas chromatography feature map is passed through a residual dual-attention mechanism module to obtain an enhanced gas chromatography feature map as a decoding feature map. It should be appreciated that after a series of convolutions, the network may obtain partial feature information, but the difference between the detailed information of high and low frequencies and the features of each category may not be automatically distinguished, the network has limited ability to selectively use the features, the focusing position can be selected in view of the attention mechanism, a more resolved feature representation is generated, and the features added to the attention module may change adaptively with the deepening of the network. Therefore, in the technical scheme of the application, a channel attention mechanism and a spatial attention mechanism are introduced, a residual error structure is introduced and combined with a proposed double-attention network to construct a residual error double-attention model, the model combines the spatial attention and the channel attention in parallel, so that different types of effective information in the gas chromatogram are captured in a large amount, the characteristic distinguishing learning capability can be effectively enhanced, and in the network training process, a task processing system is more focused on finding significant useful information related to current output in input data, thereby improving the quality of output, and an increasing attention module brings continuous performance improvement. It should be understood that the residual dual-attention mechanism module uses a parallel combination of spatial attention and channel attention, so that different types of effective information in the denoised gas chromatographic feature map are captured in a large amount, and feature recognition learning capability can be effectively enhanced, thereby improving content value detection of phosphorus trifluoride oxide in the sampled gas.

And then, the decoding characteristic diagram is passed through a decoder to obtain a decoding value, wherein the decoding value is used for representing the content value of the phosphorus oxyfluoride in the sampling gas. That is, the enhanced gas chromatograph characteristic map is subjected to a decoding regression as the decoding characteristic map to obtain a decoded value representing the content value of phosphorus oxytrifluoride in the sample gas. Therefore, the content of the phosphorus oxyfluoride in the sampling gas can be accurately detected, and the quality inspection accuracy of preparing lithium hexafluorophosphate is improved.

Particularly, in the technical scheme of the application, the mixed convolution layer obtains a plurality of branch characteristic diagrams from the noise-reduced gas chromatograph through each branch with different expansion rates, and then fuses each branch characteristic diagram in a point adding mode to finally obtain the noise-reduced gas chromatograph characteristic diagram. Here, the applicant of the present application considers that since each branch of the hybrid convolution layer has a respective expansion ratio, the fusion of the plurality of branch feature maps by means of point addition will conform to a gaussian distribution in a natural state, i.e. a part having an average correlation degree between feature values added at points so as to have an average fusion effect has the highest probability density, and a part having a relatively high and relatively low correlation degree so as to have a relatively high and relatively low fusion effect has a lower probability density. Therefore, the obtained noise-reduced gas chromatograph feature map may have poor clustering effect on the global correlation among the feature values, and the spatial attention mechanism and the channel attention mechanism of the residual dual-attention mechanism module further reduce the clustering effect on the global correlation by strengthening the attention weight to the local distribution, which results in poor representation of the key global feature distribution affecting the decoding result in the decoding feature map and affects the decoding accuracy of the decoding feature map.

Based on the above, the defocusing fuzzy optimization of feature clustering is performed on the denoised gas chromatograph feature map, and the defocusing fuzzy optimization is expressed as follows:

and->

Feature set +.>

Mean and standard deviation of (2), and->

Is the +.f. of the denoised gas chromatograph feature map>

Characteristic values of the location.

Here, the focus-free fuzzy optimization of the feature clustering compensates the dependency similarity of the high-frequency distribution feature following the gaussian point distribution with respect to the uniform representation of the overall feature distribution by performing the feature clustering index based on the statistical information on the focus stack representation for estimating the clustering metric value, thereby avoiding the focus fuzzy of the overall feature distribution caused by the low dependency similarity, thus improving the feature clustering effect of the denoised gas chromatograph feature map and optimizing the decoding accuracy of the decoded feature map obtained by the residual dual-attention mechanism module of the denoised gas chromatograph feature map. Therefore, the content of the phosphorus trifluoride oxide can be intelligently and accurately detected, so that the quality inspection accuracy of the prepared lithium hexafluorophosphate is improved, and the quality of the electrolyte is improved, and the electrolyte is applied to the battery performance of a lithium ion battery.

Based thereon, the present application provides an automated sampling analysis system for lithium hexafluorophosphate preparation, comprising: the sampling gas data acquisition module is used for acquiring a gas phase chromatogram of the sampling gas; the noise reduction module is used for enabling the gas chromatogram to pass through the noise reduction module based on the automatic coder-decoder to obtain a noise-reduced gas chromatogram; the chromatographic characteristic extraction module is used for enabling the denoised gas chromatograph to pass through a convolutional neural network model comprising a plurality of mixed convolutional layers so as to obtain a denoised gas chromatograph characteristic diagram; the code compensation module is used for performing feature clustering defocusing fuzzy optimization on the noise-reduced gas chromatography feature map so as to obtain an optimized noise-reduced gas chromatography feature map; the feature distribution enhancement module is used for enabling the optimized noise reduction gas chromatography feature map to pass through the residual double-attention mechanism module to obtain an enhanced gas chromatography feature map as a decoding feature map; and the analysis result generation module is used for enabling the decoding characteristic diagram to pass through a decoder to obtain a decoding value, wherein the decoding value is used for representing the content value of the phosphorus oxytrifluoride in the sampling gas.

Fig. 1 illustrates an application scenario diagram of an automated sampling analysis system for lithium hexafluorophosphate preparation and methods thereof according to embodiments of the present application. As shown in fig. 1, in this application scenario, a lithium hexafluorophosphate solution sample (e.g., L as illustrated in fig. 1) is stored in a sealed container (e.g., a as illustrated in fig. 1), and it should be understood that a fifth to third of the volume is reserved in the sealed container to accommodate volatile gas), and after a period of time, the volatile gas above the lithium hexafluorophosphate solution in the container is sampled using a gas tube (e.g., G as illustrated in fig. 1), and then a gas chromatograph (e.g., T as illustrated in fig. 1) is used to collect a gas chromatograph of the sampled gas (e.g., P as illustrated in fig. 1). The acquired gas chromatograph is then input into a server (e.g., S illustrated in fig. 1) deployed with an automated sampling analysis algorithm for lithium hexafluorophosphate production, wherein the server is capable of processing the gas chromatograph using the automated sampling analysis algorithm for lithium hexafluorophosphate production to generate an analysis result representative of the content value of phosphorus trifluoride oxide in the sampled gas.

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 illustrates a block diagram schematic of an automated sampling analysis system for lithium hexafluorophosphate preparation according to an embodiment of the present application. As shown in fig. 2, the automated sampling analysis system 100 for lithium hexafluorophosphate preparation according to an embodiment of the present application comprises: a sampling gas data acquisition module 110 for acquiring a gas chromatogram of the sampling gas; the noise reduction module 120 is configured to pass the gas chromatograph through a noise reduction module based on an automatic codec to obtain a noise-reduced gas chromatograph; the chromatographic feature extraction module 130 is configured to pass the denoised gas chromatograph through a convolutional neural network model that includes a plurality of mixed convolutional layers to obtain a denoised gas chromatograph feature map; the code compensation module 140 is configured to perform feature clustering of the de-focusing fuzzy optimization on the denoised gas chromatograph feature map to obtain an optimized denoised gas chromatograph feature map; the feature distribution enhancing module 150 is configured to pass the optimized noise reduction gas chromatograph feature map through a residual dual-attention mechanism module to obtain an enhanced gas chromatograph feature map as a decoding feature map; and an analysis result generation module 160, configured to pass the decoding feature map through a decoder to obtain a decoded value, where the decoded value is used to represent a content value of phosphorus oxytrifluoride in the sampling gas.

In this embodiment, the sampled gas data acquisition module 110 is configured to acquire a gas chromatograph of the sampled gas. As described above, since the quality of lithium hexafluorophosphate directly affects the quality of an electrolyte and the battery performance applied to a lithium ion battery, monitoring of the quality of lithium hexafluorophosphate is critical to the electrolyte. However, most of the existing quality inspection schemes are to perform experimental analysis on the prepared lithium hexafluorophosphate to compare the reaction result with the standard reference mass, so as to determine the quality of the prepared lithium hexafluorophosphate. Therefore, a large amount of raw materials and resources are wasted, more variable factors exist in the experimental process, the reaction result is difficult to control, and further the accuracy of lithium hexafluorophosphate quality inspection is difficult to guarantee. Meanwhile, considering that volatile impurity phosphorus trifluoride oxide is generated when water exists in the lithium hexafluorophosphate solution, the intervention degree of impurity moisture in the generation process of the lithium hexafluorophosphate solution can be reversely estimated through detecting the content of phosphorus trifluoride oxide in the volatile gas of the lithium hexafluorophosphate solution, so that the quality of the lithium hexafluorophosphate can be detected.

Therefore, the lithium hexafluorophosphate solution is prepared by storing a lithium hexafluorophosphate solution sample in a sealed container, reserving one fifth to one third of the volume in the sealed container to contain volatile gas, sampling the volatile gas above the lithium hexafluorophosphate solution in the container by using a gas sampling tube after a period of time, and quantitatively analyzing phosphorus trifluoride oxide in the sampled gas component by using a gas chromatograph, so as to determine the quality of the lithium hexafluorophosphate solution according to the phosphorus trifluoride oxide content. Therefore, in order to ensure the quality inspection accuracy of the prepared lithium hexafluorophosphate, the content of phosphorus oxytrifluoride needs to be accurately detected. Meanwhile, it is considered that since the chromatogram refers to an image of the distribution of the detection signal of the separated component over time, the content of phosphorus trifluoride oxide can be derived from the gas-phase chromatogram, but since other gases such as hydrogen fluoride are also present in the sampling gas, it is difficult to obtain the content of phosphorus trifluoride oxide accurately by the gas-phase chromatogram.

Based on the above, in the technical scheme of the application, the artificial intelligent detection algorithm based on deep learning is adopted to extract multi-scale implicit characteristic distribution information of different sizes in the noise-reduced sampling gas chromatogram, and the attention mechanism is utilized to further strengthen the influence of filtering useless interference characteristics on the detection of the small-scale trifluoro-phosphorus oxide content in the gas chromatogram so as to improve the accuracy of the trifluoro-phosphorus oxide content detection. Therefore, the content of the phosphorus trifluoride oxide can be intelligently and accurately detected, so that the quality inspection accuracy of the prepared lithium hexafluorophosphate is improved, and the quality of the electrolyte is improved, and the electrolyte is applied to the battery performance of a lithium ion battery.

In this embodiment of the present application, the noise reduction module 120 is configured to pass the gas chromatogram through a noise reduction module based on an automatic codec to obtain a noise reduced gas chromatogram. It should be understood that in the process of collecting the gas chromatogram of the sampled gas by the actual gas chromatograph, the accuracy of determining the content of phosphorus trifluoride oxide in the gas chromatogram is low due to the influence of the environmental gas factors in the air and other gas factors generated by the gas chromatograph, so that the quality detection of lithium hexafluorophosphate is affected. Therefore, in the technical solution of the present application, a noise reduction module of an automatic codec is further required to perform noise reduction processing on the gas chromatogram to obtain a denoised gas chromatogram.

In a specific embodiment of the present application, the noise reduction module 120 includes: a gas chromatograph encoding unit 121, configured to input the gas chromatograph into an encoder of the noise reduction module, where the encoder uses a convolution layer to perform explicit spatial encoding on the gas chromatograph to obtain a gas chromatograph feature map; the gas chromatograph feature decoding unit 122 is configured to input the gas chromatograph feature map to a decoder of the signal noise reduction module, where the decoder uses a deconvolution layer to deconvolute the gas chromatograph feature map to obtain the noise reduced gas chromatograph.

More specifically, in one embodiment of the present application, the gas chromatography feature decoding unit includes: the image method unit is used for carrying out zero padding on the gas chromatography characteristic map according to the step length of the convolution kernel of the encoder so as to obtain a zero padding characteristic map; and a transpose convolution unit configured to perform convolution processing on the zero padding feature map by using a transpose convolution kernel transposed to a convolution kernel of the encoder, so as to generate the denoised gas chromatogram, where the denoised gas chromatogram has the same size as the gas chromatogram.

In this embodiment of the present application, the chromatographic feature extraction module 130 is configured to pass the denoised gas chromatograph through a convolutional neural network model including a plurality of mixed convolutional layers to obtain a denoised gas chromatograph feature map. It should be understood that, although the local features can be extracted by using a single convolution kernel in the conventional convolution neural network, the large-scale periodic features and the small-scale time sequence features cannot be considered, so in the technical scheme of the application, the denoised gas chromatograph is selected to be processed through the mixed convolution layer so as to extract the multi-scale implicit correlation features of the denoised gas chromatograph, thereby obtaining the denoised gas chromatograph feature map.

In a specific embodiment of the present application, the chromatographic feature extraction module 130 is further configured to: each mixed convolution layer using the convolutional neural network model performs respective processing on input data in forward transfer of the layer: performing convolutional encoding on the denoised gas chromatogram by using a first convolutional check with a first size to obtain a first scale feature map; performing convolutional encoding on the denoised gas chromatogram by using a second convolutional check with the first void ratio to obtain a second scale feature map; performing convolutional encoding on the denoised gas chromatogram by using a third convolutional check with a second void ratio to obtain a third scale feature map; performing convolutional encoding on the denoised gas chromatograph by using a fourth convolutional kernel with a third void fraction to obtain a fourth scale feature map, wherein the first convolutional kernel, the second convolutional kernel, the third convolutional kernel and the fourth convolutional kernel have the same size, and the second convolutional kernel, the third convolutional kernel and the fourth convolutional kernel have different void fractions; performing aggregation on the first scale feature map, the second scale feature map, the third scale feature map and the fourth scale feature map along a channel dimension to obtain an aggregated feature map; pooling the aggregate feature map to generate a pooled feature map; and performing activation processing on the pooled feature map to generate an activated feature map; and the output of the last layer of the convolutional neural network model comprising a plurality of mixed convolutional layers is the noise-reduced gas chromatographic characteristic diagram.

In a specific example of the application, in the mixed convolution layer (mixed convolution layer, MCL), the design of the module includes four branches connected in parallel, and the four branches are formed by a common convolution layer with a convolution kernel size of 3×3 and a cavity convolution layer with a convolution kernel size of 3×3, the input feature images are respectively operated, the expansion rates of the three branches of the cavity convolution are respectively set to 2, 3 and 4, the image information of different receptive fields can be obtained through the setting of different expansion rates, the feature images with different scales can be obtained, the receptive fields are enlarged, the downsampling loss information is avoided, and then the 4 branch feature images are fused, so that the sampling is more dense, the high-layer features are possessed, and the additional parameter number is not increased.

In this embodiment of the present application, the code compensation module 140 is configured to perform a focus-removing fuzzy optimization of feature clustering on the denoised gas chromatograph feature map to obtain an optimized denoised gas chromatograph feature map. It should be understood that, in the technical solution of the present application, the mixed convolution layer obtains a plurality of branch feature graphs from the denoised gas chromatograph through each branch with different expansion rates, and then fuses each branch feature graph through a point adding manner to finally obtain the denoised gas chromatograph feature graph. Here, the applicant of the present application considers that since each branch of the hybrid convolution layer has a respective expansion ratio, the fusion of the plurality of branch feature maps by means of point addition will conform to a gaussian distribution in a natural state, i.e. a part having an average correlation degree between feature values added at points so as to have an average fusion effect has the highest probability density, and a part having a relatively high and relatively low correlation degree so as to have a relatively high and relatively low fusion effect has a lower probability density. Therefore, the obtained noise-reduced gas chromatograph feature map may have poor clustering effect on the global correlation among the feature values, and the spatial attention mechanism and the channel attention mechanism of the residual dual-attention mechanism module further reduce the clustering effect on the global correlation by strengthening the attention weight to the local distribution, which results in poor representation of the key global feature distribution affecting the decoding result in the decoding feature map and affects the decoding accuracy of the decoding feature map. Based on the feature clustering, performing defocusing fuzzy optimization on the denoised gas chromatography feature map.

In a specific embodiment of the present application, the code compensation module 140 is further configured to: performing feature clustering defocusing fuzzy optimization on the noise-reduced gas chromatography feature map by using the following formula to obtain the optimized noise-reduced gas chromatography feature map; wherein, the formula is:

wherein the method comprises the steps of

A +.f. representation of the denoised gas chromatograph>

Characteristic value of the location->

And->

Feature value sets respectively representing all positions of the denoised gas chromatograph feature mapMean and standard deviation of (a).

In this embodiment of the present application, the feature distribution enhancing module 150 is configured to pass the optimized noise reduction gas chromatograph feature map through a residual dual-attention mechanism module to obtain an enhanced gas chromatograph feature map as a decoding feature map. It should be appreciated that considering that the phosphorus trifluoride oxide is a small-sized object (i.e., its proportion in the gas chromatogram of the entire sampling gas is small), secondly, the sampling gas has various gas types, which may interfere with the target detection of the phosphorus trifluoride oxide. Therefore, in the technical scheme of the application, the optimized noise reduction gas chromatography characteristic diagram is further subjected to characteristic data enhancement to obtain an enhanced gas chromatography characteristic diagram.

Specifically, the optimized noise reduction gas chromatography characteristic diagram passes through a residual dual-attention mechanism module to obtain an enhanced gas chromatography characteristic diagram as a decoding characteristic diagram. It should be appreciated that after a series of convolutions, the network may obtain partial feature information, but the difference between the detailed information of high and low frequencies and the features of each category may not be automatically distinguished, the network has limited ability to selectively use the features, the focusing position can be selected in view of the attention mechanism, a more resolved feature representation is generated, and the features added to the attention module may change adaptively with the deepening of the network. Therefore, in the technical scheme of the application, a channel attention mechanism and a spatial attention mechanism are introduced, a residual error structure is introduced and combined with a proposed double-attention network to construct a residual error double-attention model, the model combines the spatial attention and the channel attention in parallel, so that different types of effective information in the gas chromatogram are captured in a large amount, the characteristic distinguishing learning capability can be effectively enhanced, and in the network training process, a task processing system is more focused on finding significant useful information related to current output in input data, thereby improving the quality of output, and an increasing attention module brings continuous performance improvement. It should be understood that the residual dual-attention mechanism module uses a parallel combination of spatial attention and channel attention, so that different types of effective information in the denoised gas chromatographic feature map are captured in a large amount, and feature recognition learning capability can be effectively enhanced, thereby improving content value detection of phosphorus trifluoride oxide in the sampled gas.

In a specific embodiment of the present application, the feature distribution enhancing module 150 includes: a spatial attention unit 151, configured to input the optimized noise reduction gas chromatograph feature map into a spatial attention module of the residual dual-attention mechanism module to obtain a spatial attention map; a channel attention unit 152, configured to input the optimized noise reduction gas chromatograph feature map into a channel attention module of the residual dual-attention mechanism module to obtain a channel attention map; an attention fusion unit 153 for fusing the spatial attention profile and the channel attention profile to obtain a fused attention profile; an activating unit 154, configured to activate the fused attention map by inputting a Sigmoid activating function to obtain a fused attention profile; an attention applying unit 155 for calculating a weighted feature map by multiplying the fused attention feature map and the optimized noise reduction gas chromatograph feature map by position points; and a residual fusion unit 156, configured to fuse the weighted feature map and the optimized noise reduction gas chromatograph feature map to obtain the enhanced gas chromatograph feature map.

In a specific embodiment of the present application, the spatial attention unit 151 includes: a convolutional encoding subunit 1511, configured to convolutionally encode the depth-enhanced gas chromatograph feature map using a convolutional layer of a spatial attention module of the residual dual-attention mechanism module to obtain a convolutional feature map; a probabilizing subunit 1512 for passing the spatial attention map through a Softmax function to obtain a spatial attention score map; and a spatial attention imparting subunit 1513 configured to perform a dot-by-dot multiplication on the spatial attention score map and the depth-enhanced gas chromatograph feature map to obtain a spatial attention map.

In a specific embodiment of the present application, the channel attention unit 152 includes: a global mean Chi Huazi unit 1521, configured to pool the global mean along the channel dimension for the depth-enhanced gas chromatography feature map to obtain a channel feature vector; a normalization subunit 1522, configured to pass the channel feature vector through a Softmax function to obtain a normalized channel feature vector; and a channel attention applying subunit 1523, configured to weight the feature matrix along the channel dimension of the depth-enhanced gas chromatograph feature map with the feature value of each position in the normalized channel feature vector as a weight to obtain a channel attention map.

In this embodiment, the analysis result generating module 160 is configured to pass the decoding feature map through a decoder to obtain a decoded value, where the decoded value is used to represent the content value of phosphorus oxynitride in the sampled gas. That is, the enhanced gas chromatograph is imported as the decoding feature map to perform decoding regression so as to obtain a decoding value for representing the content value of the phosphorus trifluoride oxide in the sampling gas, so that the content of the phosphorus trifluoride oxide in the sampling gas can be accurately detected, and the quality inspection accuracy for preparing lithium hexafluorophosphate is improved.

In a specific embodiment of the present application, the analysis result generating module is further configured to: performing decoding regression on the decoding feature map to obtain the decoding value by using the decoder according to the following formula

Wherein->

Is a decoding feature map, < >>

Is the decoded value,/->

Is a weight matrix, < >>

Representing a matrix multiplication.

In summary, according to the automated sampling analysis system for preparing lithium hexafluorophosphate according to the embodiments of the present application, the artificial intelligent detection algorithm based on deep learning is adopted to extract multi-scale implicit feature distribution information of different sizes in the sampled gas chromatogram after noise reduction, and the attention mechanism is utilized to further strengthen the influence of filtering useless interference features on the detection of the small-scale phosphorus trifluoride oxide content in the gas chromatogram, so as to improve the accuracy of the phosphorus trifluoride oxide content detection. Therefore, the content of the phosphorus trifluoride oxide can be intelligently and accurately detected, so that the quality inspection accuracy of the prepared lithium hexafluorophosphate is improved, and the quality of the electrolyte is improved, and the electrolyte is applied to the battery performance of a lithium ion battery.

As described above, the automated sampling analysis system 100 for lithium hexafluorophosphate preparation according to the embodiments of the present application may be implemented in various terminal devices, such as a server or the like, where a quality inspection algorithm of hexafluorobutadiene is deployed. In one example, the automated sample analysis system 100 for lithium hexafluorophosphate preparation may be integrated into the terminal device as a software module and/or hardware module. For example, the automated sample analysis system 100 for lithium hexafluorophosphate preparation 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 automated sampling analysis system 100 for lithium hexafluorophosphate preparation can likewise be one of the numerous hardware modules of the terminal device.

Alternatively, in another example, the automated sample analysis system 100 for lithium hexafluorophosphate preparation and the terminal device may be separate devices, and the automated sample analysis system 100 for lithium hexafluorophosphate preparation may be connected to the terminal device through a wired and/or wireless network and transmit the interaction information in accordance with the agreed data format.

An exemplary method is: fig. 7 illustrates a flow chart of an automated sampling analysis method for lithium hexafluorophosphate preparation according to an embodiment of the present application. As shown in fig. 7, the automated sampling analysis method for lithium hexafluorophosphate preparation according to the embodiment of the present application includes: s110, acquiring a gas chromatogram of the sampled gas; s120, enabling the gas chromatogram to pass through a noise reduction module based on an automatic coder-decoder to obtain a noise-reduced gas chromatogram; s130, enabling the denoised gas chromatograph to pass through a convolutional neural network model comprising a plurality of mixed convolutional layers to obtain a denoised gas chromatograph characteristic diagram; s140, performing feature clustering defocusing fuzzy optimization on the noise-reduced gas chromatography feature map to obtain an optimized noise-reduced gas chromatography feature map; s150, enabling the optimized noise reduction gas chromatography feature map to pass through a residual double-attention mechanism module to obtain an enhanced gas chromatography feature map as a decoding feature map; and S160, passing the decoding characteristic map through a decoder to obtain a decoding value, wherein the decoding value is used for representing the content value of the phosphorus trifluoride oxide in the sampling gas.

Fig. 8 illustrates a schematic diagram of a system architecture for an automated sampling analysis method for lithium hexafluorophosphate preparation according to an embodiment of the present application. As shown in fig. 8, in the system architecture of the automated sampling analysis method for lithium hexafluorophosphate preparation according to the embodiment of the present application, first, a gas chromatograph of a sampling gas is obtained, and the gas chromatograph is passed through a noise reduction module based on an automatic codec to obtain a noise reduced gas chromatograph. And then, the denoised gas chromatograph is subjected to deconvolution neural network model comprising a plurality of mixed convolution layers to obtain denoised gas chromatograph feature images, and the denoised gas chromatograph feature images are subjected to defocusing fuzzy optimization of feature clustering to obtain optimized denoised gas chromatograph feature images. And then, the optimized noise reduction gas chromatography characteristic diagram passes through a residual double-attention mechanism module to obtain an enhanced gas chromatography characteristic diagram as a decoding characteristic diagram. And finally, the decoding characteristic diagram is passed through a decoder to obtain a decoding value, wherein the decoding value is used for representing the content value of the phosphorus oxytrifluoride in the sampling gas.

Here, it will be understood by those skilled in the art that the specific operations of the respective steps in the above-described automated sampling analysis method for lithium hexafluorophosphate preparation have been described in detail in the above description of the automated sampling analysis system for lithium hexafluorophosphate preparation with reference to fig. 1 to 6, and thus, repetitive descriptions thereof will be omitted.

Exemplary electronic device: next, an electronic device according to an embodiment of the present application is described with reference to fig. 9.

As shown in fig. 9, 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 can be executed by the processor 11 to perform the automated sampling analysis and/or other desired functions for lithium hexafluorophosphate preparation of the various embodiments of the present application described above. Various contents such as a gas chromatograph of the sampled gas 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.

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 that, when executed by a processor, cause the processor to perform the steps of an automated sampling analysis method for lithium hexafluorophosphate preparation according to 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 the steps of an automated sampling analysis method for lithium hexafluorophosphate preparation according to various embodiments of the present application described in the above "exemplary methods" 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. An automated sampling analysis system for lithium hexafluorophosphate preparation, comprising: the sampling gas data acquisition module is used for acquiring a gas phase chromatogram of the sampling gas; the noise reduction module is used for enabling the gas chromatogram to pass through the noise reduction module based on the automatic coder-decoder to obtain a noise-reduced gas chromatogram; the chromatographic characteristic extraction module is used for enabling the denoised gas chromatograph to pass through a convolutional neural network model comprising a plurality of mixed convolutional layers so as to obtain a denoised gas chromatograph characteristic diagram; the code compensation module is used for performing feature clustering defocusing fuzzy optimization on the noise-reduced gas chromatography feature map so as to obtain an optimized noise-reduced gas chromatography feature map; the feature distribution enhancement module is used for enabling the optimized noise reduction gas chromatography feature map to pass through the residual double-attention mechanism module to obtain an enhanced gas chromatography feature map as a decoding feature map; the analysis result generation module is used for enabling the decoding characteristic diagram to pass through a decoder to obtain a decoding value, wherein the decoding value is used for representing the content value of the phosphorus trifluoride oxide in the sampling gas;

Wherein, the noise reduction module includes: the gas chromatograph coding unit is used for inputting the gas chromatograph into an encoder of the noise reduction module, wherein the encoder uses a convolution layer to carry out explicit space coding on the gas chromatograph so as to obtain a gas chromatograph characteristic diagram; the gas chromatography feature decoding unit is used for inputting the gas chromatography feature map into a decoder of the noise reduction module, wherein the decoder uses a deconvolution layer to carry out deconvolution processing on the gas chromatography feature map so as to obtain the noise-reduced gas chromatography map;

the chromatographic feature extraction module is further configured to use each mixed convolution layer of the convolutional neural network model to perform, in forward transfer of the layer, input data respectively: performing convolutional encoding on the denoised gas chromatogram by using a first convolutional check with a first size to obtain a first scale feature map; performing convolutional encoding on the denoised gas chromatogram by using a second convolutional check with the first void ratio to obtain a second scale feature map; performing convolutional encoding on the denoised gas chromatogram by using a third convolutional check with a second void ratio to obtain a third scale feature map; performing convolutional encoding on the denoised gas chromatograph by using a fourth convolutional kernel with a third void fraction to obtain a fourth scale feature map, wherein the first convolutional kernel, the second convolutional kernel, the third convolutional kernel and the fourth convolutional kernel have the same size, and the second convolutional kernel, the third convolutional kernel and the fourth convolutional kernel have different void fractions; performing aggregation on the first scale feature map, the second scale feature map, the third scale feature map and the fourth scale feature map along a channel dimension to obtain an aggregated feature map; pooling the aggregate feature map to generate a pooled feature map; activating the pooled feature map to generate an activated feature map; the output of the last layer of the convolutional neural network model comprising a plurality of mixed convolutional layers is the noise-reduced gas chromatographic characteristic diagram;

Wherein the code compensation module is further configured to: performing feature clustering defocusing fuzzy optimization on the noise-reduced gas chromatography feature map by using the following formula to obtain the optimized noise-reduced gas chromatography feature map; wherein, the formula is:

，

wherein the method comprises the steps of

A +.f. representation of the denoised gas chromatograph>

Characteristic value of the location->

And->

Respectively representing the mean value and standard deviation of the characteristic value sets of each position of the denoised gas chromatograph characteristic diagram;

wherein, the characteristic distribution enhancement module includes: the spatial attention unit is used for inputting the optimized noise reduction gas chromatography characteristic diagram into a spatial attention module of the residual double-attention mechanism module to obtain a spatial attention diagram; the channel attention unit is used for inputting the optimized noise reduction gas chromatography characteristic diagram into a channel attention module of the residual double-attention mechanism module to obtain a channel attention diagram; an attention fusion unit for fusing the spatial attention map and the channel attention map to obtain a fused attention map; the activating unit is used for inputting the fusion attention map into a Sigmoid activating function to activate so as to obtain a fusion attention feature map; the attention applying unit is used for calculating the weighted feature map obtained by multiplying the fusion attention feature map and the optimized noise reduction gas chromatography feature map according to the position points; and a residual fusion unit, configured to fuse the weighted feature map and the optimized noise reduction gas chromatograph feature map to obtain the enhanced gas chromatograph feature map.

2. The automated sampling analysis system for lithium hexafluorophosphate preparation of claim 1, wherein the spatial attention unit comprises: the convolution coding subunit is used for carrying out convolution coding on the optimized noise reduction gas chromatography characteristic map by using a convolution layer of a spatial attention module of the residual double-attention mechanism module so as to obtain a convolution characteristic map; a probability subunit, configured to pass the spatial attention map through a Softmax function to obtain a spatial attention score map; and a spatial attention applying subunit, configured to perform a dot-by-dot multiplication on the spatial attention score map and the optimized noise reduction gas chromatograph feature map to obtain a spatial attention map.

3. The automated sampling analysis system for lithium hexafluorophosphate preparation of claim 2, wherein the channel attention unit comprises: the global mean Chi Huazi unit is used for carrying out global mean pooling along the channel dimension on the optimized noise reduction gas chromatography feature map so as to obtain a channel feature vector; the normalization subunit is used for passing the channel characteristic vector through a Softmax function to obtain a normalized channel characteristic vector; and the channel attention applying subunit is used for weighting the characteristic matrix of the optimized noise reduction gas chromatography characteristic diagram along the channel dimension by taking the characteristic value of each position in the normalized channel characteristic vector as a weight so as to obtain a channel attention diagram.

4. The automated sampling analysis system for lithium hexafluorophosphate preparation of claim 3, wherein the analysis result generation module is further to: performing decoding regression on the decoding feature map to obtain the decoding value by using the decoder according to the following formula

Wherein->

Is a decoding feature map of the video signal,

is the decoded value,/->

Is a weight matrix, < >>

Representing a matrix multiplication.

5. An automated sampling analysis method for lithium hexafluorophosphate preparation, comprising: acquiring a gas chromatograph of the sampled gas; the gas chromatogram is passed through a noise reduction module based on an automatic coder-decoder to obtain a noise-reduced gas chromatogram; the denoised gas chromatograph is processed through a convolutional neural network model comprising a plurality of mixed convolutional layers to obtain a denoised gas chromatograph characteristic diagram; performing feature clustering defocusing fuzzy optimization on the noise-reduced gas chromatography feature map to obtain an optimized noise-reduced gas chromatography feature map; the optimized noise reduction gas chromatography feature map is passed through a residual error dual-attention mechanism module to obtain an enhanced gas chromatography feature map as a decoding feature map; the decoding characteristic diagram passes through a decoder to obtain a decoding value, wherein the decoding value is used for representing the content value of the phosphorus oxyfluoride in the sampling gas;

The method for obtaining the gas chromatograph after noise reduction by the noise reduction module based on the automatic coder-decoder comprises the following steps: inputting the gas chromatograph into an encoder of the noise reduction module, wherein the encoder uses a convolution layer to carry out explicit space coding on the gas chromatograph so as to obtain a gas chromatograph characteristic diagram; inputting the gas chromatographic feature map into a decoder of the noise reduction module, wherein the decoder uses a deconvolution layer to deconvolute the gas chromatographic feature map so as to obtain the noise-reduced gas chromatographic map;

the method for obtaining the denoised gas chromatograph feature map through a convolutional neural network model comprising a plurality of mixed convolutional layers comprises the following steps of: each mixed convolution layer using the convolutional neural network model performs respective processing on input data in forward transfer of the layer: performing convolutional encoding on the denoised gas chromatogram by using a first convolutional check with a first size to obtain a first scale feature map; performing convolutional encoding on the denoised gas chromatogram by using a second convolutional check with the first void ratio to obtain a second scale feature map; performing convolutional encoding on the denoised gas chromatogram by using a third convolutional check with a second void ratio to obtain a third scale feature map; performing convolutional encoding on the denoised gas chromatograph by using a fourth convolutional kernel with a third void fraction to obtain a fourth scale feature map, wherein the first convolutional kernel, the second convolutional kernel, the third convolutional kernel and the fourth convolutional kernel have the same size, and the second convolutional kernel, the third convolutional kernel and the fourth convolutional kernel have different void fractions; performing aggregation on the first scale feature map, the second scale feature map, the third scale feature map and the fourth scale feature map along a channel dimension to obtain an aggregated feature map; pooling the aggregate feature map to generate a pooled feature map; activating the pooled feature map to generate an activated feature map; the output of the last layer of the convolutional neural network model comprising a plurality of mixed convolutional layers is the noise-reduced gas chromatographic characteristic diagram;

The performing feature clustering defocusing fuzzy optimization on the denoised gas chromatography feature map to obtain an optimized denoised gas chromatography feature map comprises the following steps: performing feature clustering defocusing fuzzy optimization on the noise-reduced gas chromatography feature map by using the following formula to obtain the optimized noise-reduced gas chromatography feature map; wherein, the formula is:

，

wherein the method comprises the steps of

A +.f. representation of the denoised gas chromatograph>

Characteristic value of the location->

And->

the optimizing noise reduction gas chromatography feature map is passed through a residual dual-attention mechanism module to obtain an enhanced gas chromatography feature map as a decoding feature map, and the optimizing noise reduction gas chromatography feature map comprises: inputting the optimized noise reduction gas chromatography feature map into a spatial attention module of the residual double-attention mechanism module to obtain a spatial attention map; inputting the optimized noise reduction gas chromatography feature map into a channel attention module of the residual double-attention mechanism module to obtain a channel attention map; fusing the spatial attention map and the channel attention map to obtain a fused attention map; activating the fusion attention try to input a Sigmoid activation function to obtain a fusion attention feature map; calculating the position-based point multiplication of the fused attention feature map and the optimized noise reduction gas chromatography feature map to obtain a weighted feature map; and fusing the weighted feature map and the optimized noise reduction gas chromatography feature map to obtain the enhanced gas chromatography feature map.