CN115342679A

CN115342679A - Intelligent cooling liquid circulation control system for preparing hexafluorobutadiene

Info

Publication number: CN115342679A
Application number: CN202210829857.9A
Authority: CN
Inventors: 胡进军; 张鸿铨; 朱军伟; 吴雪瑛; 张奎
Original assignee: Fujian Hangfu Electronic Materials Co ltd
Current assignee: Fujian Hangfu Electronic Materials Co ltd
Priority date: 2022-07-15
Filing date: 2022-07-15
Publication date: 2022-11-15
Anticipated expiration: 2042-07-15
Also published as: CN115342679B; WO2024011747A1

Abstract

The application relates to the field of intelligent control, and particularly discloses an intelligent cooling liquid circulation control system for preparing hexafluorobutadiene, which extracts dynamic characteristic information of a temperature value of a material subjected to water washing in a condenser, a temperature value of the hexafluorobutadiene at an outlet of the condenser and a flow rate value of cooling liquid by an artificial intelligent control method, and fuses dynamic characteristics of the three in a time sequence to perform intelligent control on the flow rate of the cooling liquid so as to improve the effective utilization rate of energy.

Description

Intelligent cooling liquid circulation control system for preparing hexafluorobutadiene

Technical Field

The invention relates to the field of intelligent control, in particular to an intelligent cooling liquid circulation control system for preparing hexafluorobutadiene and a control method thereof.

Background

The hexafluobutadiene has various industrial applications, not only is a monomer for preparing various fluorine-containing high-molecular elastic materials, namely the polyhexamethylene butadiene, but also is an efficient dry etching gas with extremely low greenhouse effect, environmental protection and higher etching selectivity than the traditional plasma etching gas. In the preparation of hexafluorobutadiene, it is an important step to purify the product obtained to meet the purity requirements of industrial or electronic grade.

Common methods for purifying gases include cryogenic rectification, physical absorption, chemical conversion, selective adsorption, condensation, freezing, and membrane separation, and some manufacturers combine various purification schemes to improve the purification effect. Carry out the condensation through the coolant liquid is the treatment means that all can use in most of hexafluorobutadiene purification schemes, and the condensation effect can influence final purification effect, that is to say, the final purification precision can be influenced in the control of the circulation control system of coolant liquid, and on the other hand, the operation of the circulation control system of coolant liquid itself can consume the energy, adopts appropriate mode to control and is favorable to improving energy effective utilization to energy-concerving and environment-protective.

Therefore, an optimized coolant circulation control system for the production of hexafluorobutadiene is desired.

Disclosure of Invention

The present application is proposed to solve the above-mentioned technical problems. The embodiment of the application provides an intelligent cooling liquid circulation control system for preparing hexafluorobutadiene and a control method thereof, dynamic characteristic information of a temperature value of a material after water washing in a condenser, a temperature value of the hexafluorobutadiene at an outlet of the condenser and a flow rate value of cooling liquid is extracted through an artificial intelligent control method, and dynamic characteristics of the three on a time sequence are fused to carry out intelligent control on the flow rate of the cooling liquid, so that the effective utilization rate of energy is improved.

According to an aspect of the present application, there is provided an intelligent coolant circulation control system for hexafluorobutadiene production, comprising: the temperature data acquisition module is used for acquiring temperature values of the washed material in the condenser at a plurality of preset time points, acquired by a first temperature sensor deployed in the condenser, and temperature values of the hexafluorobutadiene at a plurality of preset time points, acquired by a second temperature sensor deployed at an outlet of the condenser; the cooling liquid circulating flow rate data acquisition module is used for acquiring flow rate values of the cooling liquid at a plurality of preset time points acquired by a flow rate meter arranged in the condenser; the first temperature data coding module is used for enabling temperature values of the washed material at a plurality of preset time points in the condenser to pass through a first time sequence encoder comprising a one-dimensional convolution layer so as to obtain a temperature time sequence characteristic vector of the object to be cooled; the second temperature data encoding module is used for enabling temperature values of the hexafluorobutadiene at a plurality of preset time points at the outlet of the condenser to pass through a second time sequence encoder containing a one-dimensional convolution layer so as to obtain an outlet temperature characteristic vector; the flow rate data coding module is used for enabling the flow rate values of the cooling liquid at the plurality of preset time points to pass through a third time sequence encoder comprising a one-dimensional convolution layer so as to obtain a flow rate characteristic vector; the characteristic distribution correction module is used for respectively performing characteristic distribution correction on the temperature time sequence characteristic vector of the object to be cooled, the outlet temperature characteristic vector and the flow velocity characteristic vector to obtain a corrected temperature time sequence characteristic vector of the object to be cooled, a corrected outlet temperature characteristic vector and a corrected flow velocity characteristic vector; the Bayes-like fusion module is used for fusing the corrected object to be cooled temperature time sequence characteristic vector, the corrected outlet temperature characteristic vector and the corrected flow velocity characteristic vector by using a Bayes-like probability model to obtain a posterior probability vector; and the circulation control result generation module is used for enabling the posterior probability vector to pass through a classifier to obtain a classification result, and the classification result is used for indicating that the flow rate of the cooling liquid at the current time point should be increased or decreased.

According to another aspect of the present application, a control method of an intelligent coolant circulation control system for hexafluorobutadiene production, comprising: acquiring temperature values of a plurality of preset time points of the washed material in the condenser, which are acquired by a first temperature sensor deployed in the condenser, and temperature values of a plurality of preset time points of the hexafluorobutadiene at the outlet of the condenser, which are acquired by a second temperature sensor deployed at the outlet of the condenser; acquiring flow rate values of the cooling liquid at the plurality of preset time points, which are acquired by a flow rate meter arranged in the condenser; enabling temperature values of the washed material at a plurality of preset time points in the condenser to pass through a first time sequence encoder comprising a one-dimensional convolution layer to obtain a temperature time sequence characteristic vector of the object to be cooled; enabling temperature values of the hexafluorobutadiene at a plurality of preset time points at the outlet of the condenser to pass through a second time sequence encoder containing a one-dimensional convolution layer to obtain an outlet temperature characteristic vector; enabling the flow velocity values of the cooling liquid at the plurality of preset time points to pass through a third time sequence encoder comprising a one-dimensional convolution layer to obtain a flow velocity characteristic vector; respectively carrying out characteristic distribution correction on the temperature time sequence characteristic vector of the object to be cooled, the outlet temperature characteristic vector and the flow velocity characteristic vector to obtain a corrected temperature time sequence characteristic vector of the object to be cooled, a corrected outlet temperature characteristic vector and a corrected flow velocity characteristic vector; fusing the corrected temperature time sequence characteristic vector of the object to be cooled, the corrected outlet temperature characteristic vector and the corrected flow velocity characteristic vector by using a Bayes-like probability model to obtain a posterior probability vector; and passing the posterior probability vector through a classifier to obtain a classification result, wherein the classification result is used for indicating that the flow rate of the cooling liquid at the current time point should be increased or decreased.

Compared with the prior art, the intelligent cooling liquid circulation control system and the control method thereof for preparing the hexafluorobutadiene, provided by the application, dynamically extract the temperature correlation characteristics inside the calcinator in real time through the convolutional neural network model based on deep learning, deeply excavate the structural change characteristics of the calcinated product and the internal heat distribution characteristics, and further intelligently adjust the temperature of the calcinator by combining the characteristic information of the three on the time sequence, so that the energy is optimized and the final quality of the finished lithium fluoride is guaranteed.

Drawings

The above and other objects, features and advantages of the present application will become more apparent by describing in more detail embodiments of the present application with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings, like reference numbers generally represent like parts or steps.

Fig. 1 is a diagram of an application scenario of an intelligent cooling liquid circulation control system for preparing hexafluorobutadiene according to an embodiment of the present application.

FIG. 2 is a block diagram of an intelligent coolant circulation control system for hexafluorobutadiene production according to an embodiment of the present application.

FIG. 3 is a block diagram of a first temperature data encoding module in an intelligent coolant circulation control system for hexafluorobutadiene production according to an embodiment of the present application.

Fig. 4 is a flowchart of a control method of an intelligent coolant circulation control system for hexafluorobutadiene production according to an embodiment of the present application.

Fig. 5 is a schematic configuration diagram of a control method of an intelligent cooling liquid circulation control system for preparing hexafluorobutadiene 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 a few embodiments of the present application, and not all embodiments of the present application, and it should be understood that the present application is not limited to the example embodiments described herein.

Overview of a scene

As mentioned above, the hexachlorobutadiene has various industrial applications, not only is a monomer for preparing a plurality of fluorine-containing high polymer elastic materials, namely the polyhexafluorobutadiene, but also is a green and environment-friendly high-efficiency dry etching gas with extremely low greenhouse effect, and the etching selectivity of the dry etching gas is higher than that of the traditional plasma etching gas. In the preparation of hexafluorobutadiene, it is an important step to purify the product obtained to meet the purity requirements of industrial or electronic grade.

Common methods for purifying gases include cryogenic rectification, physical absorption, chemical conversion, selective adsorption, condensation, freezing, and membrane separation, and some manufacturers combine various purification schemes to improve the purification effect. Carry out the condensation through the coolant liquid is the treatment means that all can use in most of hexafluorobutadiene purification schemes, and the condensation effect can influence final purification effect, that is to say, the final purification precision can be influenced in the control of the circulation control system of coolant liquid, and on the other hand, the operation of the circulation control system of coolant liquid itself can consume the energy, adopts appropriate mode to control and is favorable to improving energy effective utilization to energy-concerving and environment-protective. Therefore, an optimized coolant circulation control system for the production of hexafluorobutadiene is desired.

In the prior art, a raw material of the hexafluorobutadiene is sequentially subjected to water washing, condensation, primary adsorption, secondary adsorption, primary rectification and secondary rectification to obtain a purified product of the hexafluorobutadiene. Here, the adsorbents used for the primary adsorption include silica gel and lithium hydroxide and boron oxide supported on the silica gel; the adsorbent used in the second-stage adsorption is A-type zeolite and/or Y-type zeolite.

In the process of preparing the hexafluorobutadiene, firstly, the hexafluorobutadiene raw material enters a water washing tower for water washing, wherein the flow rate of the hexafluorobutadiene entering the water washing tower is 2-10 kg/h, the flow rate of water is (1-3) × 10m3/h, and the temperature of the kettle liquid of the water washing tower is 20-30 ℃. And then, feeding the washed material into a condenser for condensation, wherein the outlet temperature of the hexafluorobutadiene at the condenser is 5-10 ℃.

Based on this, the present inventors found that the purification accuracy and product consistency of the final hexafluorobutadiene can be improved by controlling the flow rate of the cooling liquid circulation system so that the exit temperature of the hexafluorobutadiene at the condenser is maintained within a predetermined range. Therefore, in the application, the temperature value of the material after being washed in the condenser is collected by the first temperature sensor arranged in the condenser, the temperature value of the hexafluobutadiene in the outlet of the condenser is collected by the second temperature sensor arranged at the outlet of the condenser so as to carry out dynamic temperature monitoring, and the flow rate value of the cooling liquid is collected by the flow rate meter arranged in the condenser so as to carry out dynamic flow rate control on the cooling liquid. In this way, the outlet temperature of the hexafluorobutadiene at the condenser can be maintained within a predetermined range based on the dynamic control of the coolant flow rate, thereby improving the purification accuracy and product consistency of the final hexafluorobutadiene.

Specifically, in the technical solution of the present application, first, temperature values of a plurality of predetermined time points in the condenser of a material after washing collected by a first temperature sensor disposed in the condenser and temperature values of a plurality of predetermined time points at an outlet of the condenser of hexafluorobutadiene collected by a second temperature sensor disposed in the outlet of the condenser are obtained, and flow rate values of a cooling liquid at the plurality of predetermined time points collected by a flow rate meter disposed in the condenser are obtained. Then, it should be understood that, considering that the temperature values of the water-washed material in the condenser and the temperature values of the hexafluorobutadiene at the outlet of the condenser both have a dynamic rule in the time dimension, in order to more fully extract the change characteristic information of the dynamic, the temperature values of the water-washed material at a plurality of predetermined time points in the condenser and the temperature values of the hexafluorobutadiene at a plurality of predetermined time points at the outlet of the condenser are respectively encoded by a time sequence encoder comprising a one-dimensional convolution layer, so as to respectively obtain the temperature time sequence characteristic vector of the object to be cooled and the outlet temperature characteristic vector. In particular, in a specific example, the time-series encoder is composed of fully-connected layers and one-dimensional convolutional layers which are alternately arranged, and extracts the dynamic correlation characteristics of the temperature values in the time-series dimension through one-dimensional convolutional coding and extracts the high-dimensional implicit characteristics of the temperature values through fully-connected coding.

In addition, considering that the flow velocity values of the cooling liquid have a regular characteristic of dynamics in the time dimension, the flow velocity values of the cooling liquid at the plurality of predetermined time points are encoded by a third time-series encoder including a one-dimensional convolutional layer to extract local implicit dynamic correlation characteristic information of the flow velocity values of the cooling liquid at the plurality of predetermined time points, so that a flow velocity characteristic vector is obtained.

It should be understood that, considering that the flow velocity feature vector is a prior probability, the technical solution of the present application aims to update the prior probability to obtain a posterior probability on the premise of new evidence, that is, on the premise that the temperature value of the washed material in the condenser is changed. Then, according to the bayesian formula, the posterior probability is the probability of the prior probability multiplied by the probability of the event divided by the probability of the evidence, and therefore, in the technical scheme of the application, the bayesian-like probability model is used for fusing the corrected temperature time sequence feature vector of the object to be cooled, the corrected outlet temperature feature vector and the corrected flow velocity feature vector to obtain the posterior probability vector. In one specific example, the formulation of the bayesian-like probabilistic model is represented as:

=

*

wherein

Is the value for each position in the corrected flow velocity eigenvector,

and

respectively, a value of each position in the corrected outlet temperature characteristic vector and the corrected object-to-be-cooled temperature time series characteristic vector, respectively

Is the value of each position in the a posteriori probability vector.

However, before using the bayesian probability model, the temperature time sequence feature vector of the object to be cooled, the flow velocity feature vector and the outlet temperature feature vector need to be mapped to a probability space, and when mapping is performed by a linear mapping scheme such as maximum normalization, the constraint on the feature distribution expressed by the feature value set of the feature vector to the probabilistic classification target cannot be realized, so that the classification effect of the posterior probability vector calculated by using the bayesian probability model is affected.

Therefore, before using the bayesian probability model, class-condition boundary constraint is firstly performed on the temperature time sequence feature vector of the object to be cooled, the flow velocity feature vector and the outlet temperature feature vector, specifically:

wherein

、

And

respectively, the temperature time sequence characteristic vector of the object to be cooled, the outlet temperature characteristic vector and the flow velocity characteristic vector

The characteristic value of each position is calculated,

、

and

respectively the corrected temperature time sequence characteristic vector of the object to be cooled, the corrected outlet temperature characteristic vector and the corrected flow velocity characteristic vector

Characteristic values of the individual positions.

The conditional boundary constraint carries out the boundary constraint of the features by carrying out the structure understanding based on the information rule on the feature values and the class conditions to which the feature values belong, thereby avoiding the excessive fragmentation of the decision regions in the classification target domain caused by the feature values outside the distribution of the set, leading the feature distribution represented by the feature vector to obtain the steady conditional class boundary, realizing the constraint of each feature distribution to the probabilistic classification target, improving the classification effect of the posterior probability feature distribution obtained by the Bayesian probability model, and further improving the classification accuracy.

Based on this, the present application proposes an intelligent coolant circulation control system for hexafluorobutadiene production, which comprises: the temperature data acquisition module is used for acquiring temperature values of the washed material in the condenser at a plurality of preset time points, acquired by a first temperature sensor deployed in the condenser, and temperature values of the hexafluorobutadiene at a plurality of preset time points, acquired by a second temperature sensor deployed at an outlet of the condenser; the cooling liquid circulating flow rate data acquisition module is used for acquiring flow rate values of the cooling liquid at a plurality of preset time points acquired by a flow rate meter arranged in the condenser; the first temperature data coding module is used for enabling temperature values of the washed material at a plurality of preset time points in the condenser to pass through a first time sequence encoder comprising a one-dimensional convolution layer so as to obtain a temperature time sequence characteristic vector of the object to be cooled; the second temperature data encoding module is used for enabling temperature values of the hexafluorobutadiene at a plurality of preset time points at the outlet of the condenser to pass through a second time sequence encoder containing a one-dimensional convolution layer so as to obtain an outlet temperature characteristic vector; the flow rate data coding module is used for enabling the flow rate values of the cooling liquid at the plurality of preset time points to pass through a third time sequence encoder comprising a one-dimensional convolution layer so as to obtain a flow rate characteristic vector; the characteristic distribution correction module is used for respectively performing characteristic distribution correction on the temperature time sequence characteristic vector of the object to be cooled, the outlet temperature characteristic vector and the flow velocity characteristic vector to obtain a corrected temperature time sequence characteristic vector of the object to be cooled, a corrected outlet temperature characteristic vector and a corrected flow velocity characteristic vector; the quasi-Bayes fusion module is used for fusing the corrected temperature time sequence characteristic vector of the object to be cooled, the corrected outlet temperature characteristic vector and the corrected flow velocity characteristic vector by using a quasi-Bayes probability model to obtain a posterior probability vector; and the circulation control result generation module is used for enabling the posterior probability vector to pass through a classifier to obtain a classification result, and the classification result is used for indicating that the flow rate of the cooling liquid at the current time point should be increased or decreased.

Fig. 1 illustrates an application scenario of an intelligent coolant circulation control system for hexafluorobutadiene production according to an embodiment of the present application. As shown in fig. 1, in this application scenario, first, temperature values of a water-washed material (e.g., M as illustrated in fig. 1) at a plurality of predetermined time points within a condenser (e.g., D as illustrated in fig. 1) are collected by a first temperature sensor (e.g., T1 as illustrated in fig. 1) disposed within the condenser and temperature values of a hexafluorobutadiene (e.g., H as illustrated in fig. 1) at a plurality of predetermined time points at an outlet of the condenser are collected by a second temperature sensor (e.g., T2 as illustrated in fig. 1) disposed at an outlet of the condenser, and flow rate values of a cooling liquid (e.g., L as illustrated in fig. 1) at the plurality of predetermined time points are collected by a flow rate meter (E as illustrated in fig. 1) disposed within the condenser. Then, inputting the obtained temperature values of the water-washed material at the plurality of predetermined time points in the condenser, the temperature values of the hexafluorobutadiene at the outlet of the condenser and the flow rate values of the cooling liquid at the plurality of predetermined time points into a server (for example, a server S as illustrated in fig. 1) deployed with an intelligent cooling liquid circulation control algorithm for preparing the hexafluorobutadiene, wherein the server can process the temperature values of the water-washed material at the plurality of predetermined time points in the condenser, the temperature values of the hexafluorobutadiene at the outlet of the condenser and the flow rate values of the cooling liquid at the plurality of predetermined time points by using the intelligent cooling liquid circulation control algorithm for preparing the hexafluorobutadiene to generate a classification result indicating that the flow rate of the cooling liquid at the current time point should be increased or decreased.

Having described the general principles of the present application, various non-limiting embodiments of the present application will now be described with reference to the accompanying drawings.

Exemplary System

FIG. 2 illustrates a block diagram of an intelligent coolant circulation control system for hexafluorobutadiene production according to an embodiment of the present application. As shown in fig. 2, the intelligent coolant circulation control system 200 for hexafluorobutadiene production according to the embodiment of the present application comprises: the temperature data acquisition module 210 is configured to acquire temperature values of the washed material at a plurality of predetermined time points in the condenser, which are acquired by a first temperature sensor disposed in the condenser, and temperature values of the hexafluorobutadiene at a plurality of predetermined time points in the outlet of the condenser, which are acquired by a second temperature sensor disposed at the outlet of the condenser; a cooling liquid circulation flow rate data acquisition module 220, configured to acquire flow rate values of the cooling liquid at the plurality of predetermined time points acquired by a flow rate meter disposed in the condenser; a first temperature data encoding module 230, configured to pass temperature values of the washed material at multiple predetermined time points in the condenser through a first timing encoder including a one-dimensional convolution layer to obtain a temperature timing characteristic vector of the object to be cooled; a second temperature data encoding module 240, configured to pass temperature values of the hexafluorobutadiene at multiple predetermined time points at the outlet of the condenser through a second time-series encoder including a one-dimensional convolution layer to obtain an outlet temperature feature vector; a flow rate data encoding module 250, configured to pass the flow rate values of the cooling liquid at the plurality of predetermined time points through a third time-series encoder including a one-dimensional convolution layer to obtain a flow rate eigenvector; the characteristic distribution correction module 260 is configured to perform characteristic distribution correction on the to-be-cooled object temperature time sequence characteristic vector, the outlet temperature characteristic vector, and the flow velocity characteristic vector respectively to obtain a corrected to-be-cooled object temperature time sequence characteristic vector, a corrected outlet temperature characteristic vector, and a corrected flow velocity characteristic vector; a bayesian fusion module 270, configured to fuse the corrected temperature time sequence feature vector of the object to be cooled, the corrected outlet temperature feature vector, and the corrected flow velocity feature vector using a bayesian probability model to obtain a posterior probability vector; and a cyclic control result generation module 280 for passing the posterior probability vector through a classifier to obtain a classification result, wherein the classification result is used for indicating that the flow rate of the cooling liquid at the current time point should be increased or decreased.

Specifically, in this embodiment, the temperature data collection module 210 and the cooling liquid circulation flow rate data collection module 220 are configured to obtain temperature values of the material after water washing collected by the first temperature sensor disposed in the condenser at a plurality of predetermined time points in the condenser and temperature values of the hexaflorbutadiene at a plurality of predetermined time points in the outlet of the condenser collected by the second temperature sensor disposed in the outlet of the condenser, and obtain flow rate values of the cooling liquid at the plurality of predetermined time points collected by the flow rate meter disposed in the condenser. As described above, it is understood that it is considered that the purification accuracy and the product uniformity of the final hexafluorobutadiene can be improved if the hexafluorobutadiene is maintained within a predetermined range at the outlet temperature of the condenser by controlling the flow rate of the cooling liquid circulation system. Therefore, in the technical scheme of the application, the temperature values of the washed material at a plurality of preset time points in the condenser are collected through the first temperature sensor arranged in the condenser, the temperature values of the hexafluobutadiene at a plurality of preset time points in the outlet of the condenser are collected through the second temperature sensor arranged at the outlet of the condenser, so that the temperature is dynamically monitored, and the flow rate value of the cooling liquid at a plurality of preset time points is collected through the flow rate meter arranged in the condenser, so that the flow rate of the cooling liquid is dynamically controlled. In this way, the outlet temperature of the hexafluorobutadiene at the condenser can be maintained within a predetermined range based on the dynamic control of the coolant flow rate, thereby improving the purification accuracy and product consistency of the final hexafluorobutadiene.

Specifically, in this embodiment of the present application, the first temperature data encoding module 230 and the second temperature data encoding module 240 are configured to pass temperature values of multiple predetermined time points of the washed material in the condenser through a first time sequence encoder that includes a one-dimensional convolution layer to obtain a temperature time sequence feature vector of the object to be cooled, and pass temperature values of multiple predetermined time points of the hexafluorobutadiene at the outlet of the condenser through a second time sequence encoder that includes a one-dimensional convolution layer to obtain an outlet temperature feature vector. It should be understood that, considering that the temperature values of the material after water washing in the condenser and the temperature values of the hexafluobutadiene at the outlet of the condenser have a dynamic law in the time dimension, in order to more fully exploit the change characteristic information of the dynamic nature, in the technical solution of the present application, the temperature values of the material after water washing at a plurality of predetermined time points in the condenser and the temperature values of the hexafluobutadiene at a plurality of predetermined time points at the outlet of the condenser are further encoded respectively by a first time sequence encoder and a second time sequence encoder which comprise one-dimensional convolution layers, so as to respectively obtain a temperature time sequence characteristic vector of the material to be cooled and an outlet temperature characteristic vector. In particular, the first and second time series encoders have the same network structure. In a specific example, the first time sequence encoder and the second time sequence encoder are both composed of full connection layers and one-dimensional convolution layers which are alternately arranged, and the dynamic association characteristics of the temperature values in the time sequence dimension are extracted through one-dimensional convolution coding, and the high-dimensional implicit characteristics of the temperature values are extracted through full connection coding.

More specifically, in this embodiment of the present application, the first temperature data encoding module includes: arranging temperature values of the washed material at a plurality of preset time points in the condenser into a first temperature input vector according to a time dimension; using a full-connection layer of the time sequence encoder to perform full-connection encoding on the first temperature input vector by using the following formula to extract high-dimensional implicit features of feature values of all positions in the first temperature input vector, wherein the formula is as follows:

wherein

Is the input vector of the said one or more input vectors,

is the output vector of the output vector,

is a matrix of the weights that is,

is a vector of the offset to the offset,

represents a matrix multiplication; performing one-dimensional convolution encoding on the first temperature input vector by using a one-dimensional convolution layer of the time sequence encoder according to the following formula so as to extract high-dimensional implicit correlation features among feature values of all positions in the first temperature input vector, wherein the formula is as follows:

wherein,ais a convolution kernelxA width in the direction,FIs a convolution kernel parameter vector,GIs a local vector matrix that operates with a convolution kernel,wthe size of the convolution kernel.

More specifically, in this embodiment of the present application, the second temperature data encoding module includes: a second temperature input vector construction unit for arranging temperature values of the hexafluorobutadiene at a plurality of predetermined time points at an outlet of the condenser as a second temperature input vector in a time dimension; a second full-connection coding unit, configured to perform full-connection coding on the second temperature input vector by using a full-connection layer of the time sequence encoder according to the following formula to extract a high-dimensional implicit feature of a feature value of each position in the second temperature input vector, where the formula is:

wherein

Is the input vector of the said one or more input vectors,

is the output vector of the output vector,

is a matrix of weights that is a function of,

is a vector of the offset to the offset,

represents a matrix multiplication; a second one-dimensional convolution encoding unit for performing one-dimensional convolution encoding on the second temperature input vector using the one-dimensional convolution layer of the time-series encoder according to the following formula to extract the second temperature inputHigh-dimensional implicit associated features between feature values of each position in the vector, wherein the formula is as follows:

wherein,ais a convolution kernelxWidth in the direction,FIs a convolution kernel parameter vector,GIs a local vector matrix that operates with a convolution kernel,wthe size of the convolution kernel.

FIG. 3 illustrates a block diagram of a first temperature data encoding module in an intelligent coolant loop control system for hexafluorobutadiene production according to an embodiment of the present application. As shown in fig. 3, the first temperature data encoding module 230 includes: a first temperature input vector construction unit 231, configured to arrange temperature values of the washed material at a plurality of predetermined time points in the condenser into a first temperature input vector according to a time dimension; a first full-concatenation encoding unit 232, configured to perform full-concatenation encoding on the first temperature input vector by using a full-concatenation layer of the time sequence encoder according to the following formula to extract a high-dimensional implicit feature of a feature value of each position in the first temperature input vector, where the formula is:

in which

Is the input vector of the said one or more input vectors,

is the output vector of the output vector,

is a matrix of the weights that is,

is a vector of the offset to the offset,

represents a matrix multiplication; a first one-dimensional convolution encoding unit 233, configured to perform one-dimensional convolution encoding on the first temperature input vector by using the one-dimensional convolution layer of the time sequence encoder according to the following formula to extract a high-dimensional implicit correlation feature between feature values of each position in the first temperature input vector, where the formula is:

wherein,ais a convolution kernelxA width in the direction,FIs a convolution kernel parameter vector,GIs a matrix of local vectors operating with a convolution kernel,wthe size of the convolution kernel.

Specifically, in this embodiment, the flow rate data encoding module 250 is configured to pass the flow rate values of the cooling liquid at the predetermined time points through a third time-series encoder including one-dimensional convolution layers to obtain a flow rate eigenvector. It should be understood that, as for the flow velocity value of the cooling liquid, considering that there is a regular characteristic of dynamics also in the time dimension, in the technical solution of the present application, the flow velocity values of the cooling liquid at the multiple predetermined time points are also encoded by a third time-series encoder that includes a one-dimensional convolution layer, so as to extract local implicit dynamic correlation characteristic information of the flow velocity values of the cooling liquid at the multiple predetermined time points, thereby obtaining a flow velocity characteristic vector. In particular, it is worth mentioning that the first time sequence encoder, the second time sequence encoder and the third time sequence encoder have the same network structure, and the first time sequence encoder, the second time sequence encoder and the third time sequence encoder are composed of fully-connected layers and one-dimensional convolutional layers which are alternately arranged.

More specifically, in this embodiment of the present application, the flow rate data encoding module includes: the flow rate input vector construction unit is used for arranging temperature values of a plurality of preset time points of the hexafluorobutadiene at the outlet of the condenser into a flow rate input vector according to a time dimension; total flow rateA connection coding unit, configured to perform full-connection coding on the flow rate input vector by using a full-connection layer of the time-series encoder according to the following formula to extract a high-dimensional implicit feature of a feature value at each position in the flow rate input vector, where the formula is:

in which

Is the input vector of the said one or more input vectors,

is the output vector of the output vector,

is a matrix of the weights that is,

is a vector of the offset to the offset,

represents a matrix multiplication; a flow velocity one-dimensional convolution coding unit, configured to perform one-dimensional convolution coding on the flow velocity input vector by using a one-dimensional convolution layer of the time-series encoder according to the following formula to extract a high-dimensional implicit correlation feature between feature values of each position in the flow velocity input vector, where the formula is:

Specifically, in this embodiment of the application, the characteristic distribution correction module 260 is configured to perform characteristic distribution correction on the to-be-cooled object temperature time sequence characteristic vector, the outlet temperature characteristic vector, and the flow velocity characteristic vector respectively to obtain a corrected to-be-cooled object temperature time sequence characteristic vector, a corrected outlet temperature characteristic vector, and a corrected flow velocity characteristic vector. It should be understood that, before using the bayesian probability model, the temperature time series eigenvector of the object to be cooled, the flow velocity eigenvector and the outlet temperature eigenvector need to be mapped to the probability space first, but when mapping is performed by a linear mapping scheme such as maximum normalization, the constraint on the feature distribution expressed by the eigenvalue set of the eigenvector to the probabilistic classification target cannot be realized, so that the classification effect of the posterior probability vector calculated by using the bayesian probability model can be affected. Therefore, in the technical solution of the present application, before using the bayesian probability model, class condition boundary constraint is performed on the to-be-cooled object temperature time sequence feature vector, the flow velocity feature vector, and the outlet temperature feature vector, so as to obtain a corrected to-be-cooled object temperature time sequence feature vector, a corrected outlet temperature feature vector, and a corrected flow velocity feature vector.

More specifically, in an embodiment of the present application, the feature distribution correction module is further configured to: respectively carrying out characteristic distribution correction on the temperature time sequence characteristic vector of the object to be cooled, the outlet temperature characteristic vector and the flow velocity characteristic vector by the following formulas to obtain the corrected temperature time sequence characteristic vector of the object to be cooled, the corrected outlet temperature characteristic vector and the corrected flow velocity characteristic vector;

wherein the formula is:

wherein

、

And

The characteristic value of each position is calculated,

、

and

Characteristic values of the individual positions. It should be understood that, here, the class conditional boundary constraint performs the boundary constraint of the features by performing the structure understanding based on the information rule on the feature values and the class conditions to which the feature values belong, so as to avoid excessive fragmentation of the decision regions in the classification target domain caused by the feature values outside the distribution of the set of feature values, so that the feature distribution represented by the feature vector obtains a robust conditioned class boundary, thereby implementing the constraint of each feature distribution to the probabilistic classification target, improving the classification effect of the posterior probability feature distribution obtained by the bayesian probability model, and further improving the classification accuracy.

Specifically, in this embodiment of the present application, the bayesian fusion module 270 is configured to fuse the corrected time-series feature vector of the temperature of the object to be cooled, the corrected outlet temperature feature vector, and the corrected flow velocity feature vector by using a bayesian probability-like model to obtain a posterior probability vector. It should be understood that, considering that the flow velocity feature vector is a prior probability, the technical solution of the present application aims to update the prior probability to obtain a posterior probability on the premise of a new evidence, that is, a temperature value of the material after washing in the condenser is changed. Then, according to the bayesian formula, the posterior probability is the probability of the event multiplied by the prior probability divided by the probability of the evidence, and therefore, in the technical scheme of the application, the bayesian-like probability model is used for fusing the corrected temperature time sequence feature vector of the object to be cooled, the corrected outlet temperature feature vector and the corrected flow velocity feature vector to obtain the posterior probability vector.

More specifically, in this embodiment of the present application, the bayesian-like fusion module is further configured to: fusing the corrected temperature time sequence characteristic vector of the object to be cooled, the corrected outlet temperature characteristic vector and the corrected flow velocity characteristic vector by using a Bayesian probability-like model according to the following formula to obtain the posterior probability vector;

wherein the formula is:

=

*

wherein

Is the value of each position in the corrected flow velocity feature vector,

and

Is the value of each position in the a posteriori probability vector.

Specifically, in the embodiment of the present application, the circulation control result generation module 280 is configured to pass the posterior probability vector through a classifier to obtain a classification result, where the classification result is used to indicate that the flow rate of the cooling liquid at the current time point should be increased or decreased. More specifically, in this embodiment of the application, the loop control result generating module is further configured to: processing the posterior probability vector using the classifier to obtain the classification result with the formula:

wherein, in the process,

to

In order to be a weight matrix, the weight matrix,

to

In order to be a vector of the offset,

is the posterior probability vector.

In summary, the intelligent cooling liquid circulation control system 200 for preparing hexafluorobutadiene according to the embodiment of the present application is illustrated, and extracts dynamic characteristic information of the temperature value of the material after water washing in the condenser, the temperature value of the hexafluorobutadiene at the outlet of the condenser, and the flow rate value of the cooling liquid by an artificial intelligent control method, and performs intelligent control of the flow rate of the cooling liquid by fusing dynamic characteristics of the three in time sequence, so as to improve the effective utilization rate of energy.

As described above, the intelligent coolant circulation control system 200 for hexafluorobutadiene production according to the embodiment of the present application can be implemented in various terminal devices, such as a server of an intelligent coolant circulation control algorithm for hexafluorobutadiene production, and the like. In one example, the intelligent coolant circulation control system 200 for hexafluorobutadiene production according to the embodiment of the present application may be integrated into a terminal device as one software module and/or hardware module. For example, the intelligent coolant circulation control system 200 for hexafluorobutadiene production 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 intelligent coolant circulation control system 200 for preparing hexafluorobutadiene can also be one of many hardware modules of the terminal equipment.

Alternatively, in another example, the intelligent coolant circulation control system 200 for preparing hexafluorobutadiene and the terminal device may be separate devices, and the intelligent coolant circulation control system 200 for preparing hexafluorobutadiene may be connected to the terminal device through a wired and/or wireless network and transmit interactive information according to an agreed data format.

Exemplary method

Fig. 4 illustrates a flow chart of a control method of the intelligent coolant circulation control system for hexafluorobutadiene production. As shown in fig. 4, the control method of the intelligent cooling liquid circulation control system for preparing hexafluorobutadiene according to the embodiment of the present application comprises the steps of: s110, acquiring temperature values of the washed material in the condenser at a plurality of preset time points, acquired by a first temperature sensor deployed in the condenser, and temperature values of the hexafluorobutadiene at a plurality of preset time points, acquired by a second temperature sensor deployed at an outlet of the condenser; s120, acquiring flow rate values of the cooling liquid at a plurality of preset time points, which are acquired by a flow rate meter arranged in the condenser; s130, enabling temperature values of the washed material at a plurality of preset time points in the condenser to pass through a first time sequence encoder comprising a one-dimensional convolution layer to obtain a temperature time sequence characteristic vector of the object to be cooled; s140, enabling temperature values of the hexafluorobutadiene at a plurality of preset time points at the outlet of the condenser to pass through a second time sequence encoder containing a one-dimensional convolution layer to obtain an outlet temperature characteristic vector; s150, enabling the flow velocity values of the cooling liquid at the plurality of preset time points to pass through a third time sequence encoder comprising a one-dimensional convolution layer to obtain a flow velocity characteristic vector; s160, respectively carrying out characteristic distribution correction on the temperature time sequence characteristic vector of the object to be cooled, the outlet temperature characteristic vector and the flow velocity characteristic vector to obtain a corrected temperature time sequence characteristic vector of the object to be cooled, a corrected outlet temperature characteristic vector and a corrected flow velocity characteristic vector; s170, fusing the corrected temperature time sequence characteristic vector of the object to be cooled, the corrected outlet temperature characteristic vector and the corrected flow velocity characteristic vector by using a Bayesian-like probability model to obtain a posterior probability vector; and S180, passing the posterior probability vector through a classifier to obtain a classification result, wherein the classification result is used for indicating that the flow rate of the cooling liquid at the current time point should be increased or decreased.

Fig. 5 illustrates an architectural schematic diagram of a control method of an intelligent coolant circulation control system for hexafluorobutadiene production according to an embodiment of the present application. As shown in fig. 5, in the network architecture of the control method of the intelligent cooling liquid circulation control system for preparing hexafluorobutadiene, firstly, the obtained temperature values (for example, P1 as illustrated in fig. 5) of the water-washed material at a plurality of predetermined time points in the condenser are passed through a first timing encoder (for example, E1 as illustrated in fig. 5) containing one-dimensional convolution layers to obtain a temperature time sequence characteristic vector (for example, VF1 as illustrated in fig. 5) of the material to be cooled; then, passing the obtained temperature values (e.g., P2 as illustrated in fig. 5) of the hexafluorobutadiene at a plurality of predetermined time points at the outlet of the condenser through a second time-series encoder (e.g., E2 as illustrated in fig. 5) containing a one-dimensional convolution layer to obtain an outlet temperature characteristic vector (e.g., VF2 as illustrated in fig. 5); then, passing the obtained flow velocity values (e.g., P3 as illustrated in fig. 5) of the cooling liquid at the plurality of predetermined time points through a third time-series encoder (e.g., E3 as illustrated in fig. 5) including one-dimensional convolutional layers to obtain a flow velocity eigenvector (e.g., VF3 as illustrated in fig. 5); then, respectively performing feature distribution correction on the to-be-cooled object temperature time sequence feature vector, the outlet temperature feature vector and the flow velocity feature vector to obtain a corrected to-be-cooled object temperature time sequence feature vector (for example, V1 as illustrated in fig. 5), a corrected outlet temperature feature vector (for example, V2 as illustrated in fig. 5) and a corrected flow velocity feature vector (for example, V3 as illustrated in fig. 5); then, fusing the corrected temperature time sequence feature vector of the object to be cooled, the corrected outlet temperature feature vector and the corrected flow velocity feature vector by using a Bayesian-like probability model to obtain a posterior probability vector (for example, VF as illustrated in FIG. 5); and, finally, passing the posterior probability vector through a classifier (e.g., circle S as illustrated in fig. 5) to obtain a classification result indicating that the flow rate of the cooling liquid at the current time point should be increased or decreased.

More specifically, in steps S110 and S120, temperature values of the water-washed material at a plurality of predetermined time points in the condenser collected by a first temperature sensor disposed in the condenser and temperature values of the hexafluorobutadiene at a plurality of predetermined time points in the outlet of the condenser collected by a second temperature sensor disposed in the outlet of the condenser are obtained, and flow rate values of the cooling liquid at the plurality of predetermined time points collected by a flow rate meter disposed in the condenser are obtained. It should be understood that it is considered that the purification accuracy and product consistency of the final hexafluorobutadiene can be improved if the hexafluorobutadiene is maintained within a predetermined range at the outlet temperature of the condenser by controlling the flow rate of the cooling liquid circulation system. Therefore, in the technical scheme of this application, the temperature value of the material after the material is washed in the condenser at a plurality of predetermined time points is gathered through the first temperature sensor that deploys in the condenser and the temperature value of the second temperature sensor that deploys in the exit of condenser at a plurality of predetermined time points of hexachlorobutadiene at the exit of condenser are gathered to carry out the dynamic monitoring of temperature, and the flow rate meter that deploys in the condenser gathers a plurality of predetermined time points the flow rate value of coolant liquid in order to carry out the flow rate dynamic control of coolant liquid. In this way, the outlet temperature of the hexafiuorobutadiene at the condenser can be maintained within a predetermined range based on the dynamic control of the coolant flow rate, thereby improving the purification accuracy and product consistency of the final hexafiuorobutadiene.

More specifically, in step S130 and step S140, the temperature values of the water-washed material at a plurality of predetermined time points in the condenser are passed through a first time-series encoder containing a one-dimensional convolution layer to obtain a temperature time-series characteristic vector of the object to be cooled, and the temperature values of the hexafluorobutadiene at a plurality of predetermined time points at the outlet of the condenser are passed through a second time-series encoder containing a one-dimensional convolution layer to obtain an outlet temperature characteristic vector. It should be understood that, considering that the temperature value of the water-washed material in the condenser and the temperature value of the hexafluorobutadiene at the outlet of the condenser both have a dynamic law in the time dimension, in order to more fully extract the change characteristic information of such dynamic nature, in the technical solution of the present application, the temperature values of the water-washed material at a plurality of predetermined time points in the condenser and the temperature values of the hexafluorobutadiene at a plurality of predetermined time points at the outlet of the condenser are further encoded respectively by a first time-sequence encoder and a second time-sequence encoder comprising a one-dimensional convolution layer, so as to obtain the temperature time-sequence characteristic vector of the object to be cooled and the outlet temperature characteristic vector respectively. In particular, the first time sequential encoder and the second time sequential encoder have the same network structure. In a specific example, the first time sequence encoder and the second time sequence encoder are both composed of full connection layers and one-dimensional convolution layers which are alternately arranged, and the dynamic association characteristics of the temperature values in the time sequence dimension are extracted through one-dimensional convolution coding, and the high-dimensional implicit characteristics of the temperature values are extracted through full connection coding.

More specifically, in step S150, the flow velocity values of the cooling liquid at the plurality of predetermined time points are passed through a third time-series encoder including one-dimensional convolution layers to obtain a flow velocity eigenvector. It should be understood that, regarding the flow velocity value of the cooling liquid, it is considered that there is a regular characteristic of dynamics also in the time dimension, and therefore, in the technical solution of the present application, the flow velocity values of the cooling liquid at the plurality of predetermined time points are also encoded by a third time-series encoder including a one-dimensional convolutional layer, so as to extract local implicit dynamic correlation characteristic information of the flow velocity values of the cooling liquid at the plurality of predetermined time points, thereby obtaining a flow velocity characteristic vector. In particular, it is worth mentioning that the first time sequence encoder, the second time sequence encoder and the third time sequence encoder have the same network structure, and the first time sequence encoder, the second time sequence encoder and the third time sequence encoder are composed of a fully connected layer and a one-dimensional convolutional layer which are alternately arranged.

More specifically, in step S160, feature distribution correction is performed on the to-be-cooled object temperature time sequence feature vector, the outlet temperature feature vector, and the flow velocity feature vector to obtain a corrected to-be-cooled object temperature time sequence feature vector, a corrected outlet temperature feature vector, and a corrected flow velocity feature vector, respectively. It should be understood that, before using the bayesian probability model, the temperature time series eigenvector of the object to be cooled, the flow velocity eigenvector and the outlet temperature eigenvector need to be mapped to the probability space first, but when mapping is performed by a linear mapping scheme such as maximum normalization, the constraint on the feature distribution expressed by the eigenvalue set of the eigenvector to the probabilistic classification target cannot be realized, so that the classification effect of the posterior probability vector calculated by using the bayesian probability model can be affected. Therefore, in the technical solution of the present application, before using the bayesian probability model, class-conditional boundary constraint is performed on the temperature time sequence feature vector of the object to be cooled, the flow velocity feature vector, and the outlet temperature feature vector to obtain a corrected temperature time sequence feature vector of the object to be cooled, a corrected outlet temperature feature vector, and a corrected flow velocity feature vector.

More specifically, in step S170, the corrected object-to-be-cooled temperature time-series feature vector, the corrected outlet temperature feature vector and the corrected flow velocity feature vector are fused using a bayesian-like probability model to obtain an a posteriori probability vector. It should be understood that, considering that the flow velocity feature vector is a prior probability, the technical solution of the present application aims to update the prior probability to obtain a posterior probability on the premise of a new evidence, that is, a temperature value of the material after washing in the condenser is changed. Then, according to the bayesian formula, the posterior probability is the probability of the prior probability multiplied by the probability of the event divided by the probability of the evidence, and therefore, in the technical scheme of the application, the bayesian-like probability model is used for fusing the corrected temperature time sequence feature vector of the object to be cooled, the corrected outlet temperature feature vector and the corrected flow velocity feature vector to obtain the posterior probability vector.

More specifically, in step S180, the posterior probability vector is passed through a classifier to obtain a classification result, which is used to indicate that the flow rate of the cooling liquid should be increased or decreased at the current time point. Specifically, in the embodiment of the present application, the posterior probability vector is processed by the classifier according to the following formula to obtain the classification result, where the formula is:

wherein, in the process,

to

In order to be a weight matrix, the weight matrix,

to

In order to be a vector of the offset,

is the posterior probability vector.

In summary, the control method of the intelligent cooling liquid circulation control system for preparing hexafluorobutadiene based on the embodiment of the present application is illustrated, and the dynamic characteristic information of the temperature value of the material after water washing in the condenser, the temperature value of the hexafluorobutadiene at the outlet of the condenser and the flow rate value of the cooling liquid is extracted by an artificial intelligent control method, and the dynamic characteristics of the three in time sequence are fused to perform the intelligent control of the flow rate of the cooling liquid, so as to improve the effective utilization rate of energy.

The foregoing describes the general principles of the present application in conjunction with specific embodiments, however, it is noted that the advantages, effects, etc. mentioned in the present application are merely examples and are not limiting, and they should not be considered essential to the various embodiments of the present application. Furthermore, the foregoing disclosure of specific details is for the purpose of illustration and description and is not intended to be limiting, since the foregoing disclosure is not intended to be exhaustive or to limit the disclosure to the precise details disclosed.

The block diagrams of devices, apparatuses, devices, systems referred to in this application are only used as illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. These devices, apparatuses, devices, systems may be connected, arranged, configured in any manner, as will be appreciated by those skilled in the art. Words such as "including," "comprising," "having," and the like are open-ended words that mean "including, but not limited to," and are used interchangeably herein. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".

It should also be noted that in the devices, apparatuses, and methods of the present application, each component or step can be decomposed and/or re-combined. These decompositions and/or recombinations should be considered as equivalents of 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, the description is not intended to limit embodiments of the application to the form disclosed herein. While a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, alterations, additions and sub-combinations thereof.

Claims

1. An intelligent cooling liquid circulation control system for preparing hexafluorobutadiene, which is characterized by comprising: the temperature data acquisition module is used for acquiring temperature values of the washed material in the condenser at a plurality of preset time points, which are acquired by a first temperature sensor deployed in the condenser, and temperature values of the hexafluobutadiene in the condenser at a plurality of preset time points, which are acquired by a second temperature sensor deployed at an outlet of the condenser; the cooling liquid circulating flow rate data acquisition module is used for acquiring flow rate values of the cooling liquid at a plurality of preset time points acquired by a flow rate meter arranged in the condenser; the first temperature data coding module is used for enabling temperature values of the washed material at a plurality of preset time points in the condenser to pass through a first time sequence coder containing a one-dimensional convolution layer so as to obtain a temperature time sequence characteristic vector of the object to be cooled; the second temperature data encoding module is used for enabling temperature values of a plurality of preset time points of the hexafluorobutadiene at the outlet of the condenser to pass through a second time sequence encoder containing a one-dimensional convolution layer so as to obtain an outlet temperature characteristic vector; the flow rate data coding module is used for enabling the flow rate values of the cooling liquid at the plurality of preset time points to pass through a third time sequence encoder comprising a one-dimensional convolution layer so as to obtain a flow rate characteristic vector; the characteristic distribution correction module is used for respectively performing characteristic distribution correction on the temperature time sequence characteristic vector of the object to be cooled, the outlet temperature characteristic vector and the flow velocity characteristic vector to obtain a corrected temperature time sequence characteristic vector of the object to be cooled, a corrected outlet temperature characteristic vector and a corrected flow velocity characteristic vector; the quasi-Bayes fusion module is used for fusing the corrected temperature time sequence characteristic vector of the object to be cooled, the corrected outlet temperature characteristic vector and the corrected flow velocity characteristic vector by using a quasi-Bayes probability model to obtain a posterior probability vector; and the circulation control result generation module is used for enabling the posterior probability vector to pass through a classifier to obtain a classification result, and the classification result is used for indicating that the flow rate of the cooling liquid at the current time point should be increased or decreased.

2. The intelligent coolant loop control system for hexafluorobutadiene production of claim 1, wherein the first temperature data encoding module comprises: the first temperature input vector construction unit is used for arranging temperature values of the washed material at a plurality of preset time points in the condenser into a first temperature input vector according to a time dimension; a first full-connection coding unit, configured to perform full-connection coding on the first temperature input vector by using a full-connection layer of the time sequence encoder according to the following formula to extract a high-dimensional implicit feature of a feature value of each position in the first temperature input vector, where the formula is:

wherein

Is the input vector of the said one or more input vectors,

is the output vector of the digital video signal,

is a matrix of the weights that is,

is a vector of the offset to the offset,

represents a matrix multiplication; a first one-dimensional convolution coding unit, configured to perform one-dimensional convolution coding on the first temperature input vector by using a one-dimensional convolution layer of the time-series encoder according to the following formula to extract high-dimensional implicit correlation features between feature values of each position in the first temperature input vector, where the formula is:

wherein,ais a convolution kernelxWidth in the direction,FIs a convolution kernel parameter vector,GIs a matrix of local vectors operating with a convolution kernel,wthe size of the convolution kernel.

3. The intelligent coolant loop control system for hexafluorobutadiene production of claim 2, wherein the second temperature data encoding module comprises: a second temperature input vector construction unit for arranging temperature values of the hexafluorobutadiene at a plurality of predetermined time points at an outlet of the condenser as a second temperature input vector in a time dimension; a second full-connection encoding unit for full-connection encoding the second temperature input vector using a full-connection layer of the time-series encoder in the following formula to extract the second temperature input vectorHigh-dimensional implicit features of eigenvalues of various positions in the temperature input vector, wherein the formula is:

wherein

Is the input vector of the said one or more input vectors,

is the output vector of the digital video signal,

is a matrix of weights that is a function of,

is a vector of the offset to the offset,

represents a matrix multiplication; a second one-dimensional convolution coding unit, configured to perform one-dimensional convolution coding on the second temperature input vector by using the one-dimensional convolution layer of the time sequence encoder according to the following formula to extract high-dimensional implicit correlation features between feature values of each position in the second temperature input vector, where the formula is:

wherein,ais a convolution kernel inxWidth in the direction,FIs a convolution kernel parameter vector,GIs a local vector matrix that operates with a convolution kernel,wthe size of the convolution kernel.

4. The intelligent coolant circulation control system for hexafluorobutadiene production of claim 3, wherein the flow rate data encoding module comprises: flow velocity input vector structureA manufacturing unit, which is used for arranging temperature values of a plurality of preset time points of the hexafluorobutadiene at the outlet of the condenser into flow speed input vectors according to a time dimension; a flow velocity full-connection coding unit, configured to perform full-connection coding on the flow velocity input vector by using a full-connection layer of the time sequence encoder according to the following formula to extract a high-dimensional implicit feature of a feature value at each position in the flow velocity input vector, where the formula is:

in which

Is the input vector of the said one or more input vectors,

is the output vector of the output vector,

is a matrix of the weights that is,

is a vector of the offset to the offset,

wherein,ais a convolution kernelxWidth in the direction,FIs a convolution kernel parameter vector,GAs part of a convolution kernel operationThe matrix of vectors is then used to generate,wthe size of the convolution kernel.

5. The intelligent coolant circulation control system for hexafluorobutadiene production according to claim 4, wherein the first, second and third time-series encoders have the same network structure, and are composed of fully-connected layers and one-dimensional convolutional layers which are alternately arranged.

6. The intelligent cooling liquid circulation control system for preparing hexafluorobutadiene according to claim 5, wherein the characteristic distribution correction module is further configured to perform characteristic distribution correction on the to-be-cooled object temperature time sequence characteristic vector, the outlet temperature characteristic vector and the flow rate characteristic vector respectively to obtain the corrected to-be-cooled object temperature time sequence characteristic vector, the corrected outlet temperature characteristic vector and the corrected flow rate characteristic vector according to the following formulas; wherein the formula is:

wherein

、

And

The value of the characteristic of each of the positions,

、

and

Characteristic values of the individual positions.

7. The intelligent coolant circulation control system for hexafluorobutadiene production of claim 6, wherein the Bayesian fusion-like module is further configured to: fusing the corrected temperature time sequence characteristic vector of the object to be cooled, the corrected outlet temperature characteristic vector and the corrected flow velocity characteristic vector by using a Bayes-like probability model according to the following formula to obtain the posterior probability vector; wherein the formula is:

=

*

wherein

Is the value of each position in the corrected flow velocity feature vector,

and

Is the value of each position in the a posteriori probability vector.

8. The intelligent coolant loop control system for hexafluorobutadiene production of claim 7, wherein the loop control result generation module is further to: processing the posterior probability vector using the classifier to obtain the classification result with the formula:

wherein

to

In order to be a weight matrix, the weight matrix,

to

In order to be a vector of the offset,

is the posterior probability vector.

9. A control method of an intelligent cooling liquid circulation control system for preparing hexafluorobutadiene is characterized by comprising the following steps: acquiring temperature values of a plurality of preset time points of a washed material in a condenser, which are acquired by a first temperature sensor deployed in the condenser, and temperature values of a plurality of preset time points of hexaflorobutadiene at an outlet of the condenser, which are acquired by a second temperature sensor deployed at the outlet of the condenser; acquiring flow rate values of the cooling liquid at the plurality of preset time points, which are acquired by a flow rate meter arranged in the condenser; enabling temperature values of the washed material at a plurality of preset time points in the condenser to pass through a first time sequence encoder comprising a one-dimensional convolution layer to obtain a temperature time sequence characteristic vector of the object to be cooled; enabling temperature values of the hexafluorobutadiene at a plurality of preset time points of the outlet of the condenser to pass through a second time sequence encoder containing a one-dimensional convolution layer to obtain an outlet temperature characteristic vector; enabling the flow velocity values of the cooling liquid at the plurality of preset time points to pass through a third time sequence encoder comprising a one-dimensional convolution layer to obtain a flow velocity characteristic vector; respectively carrying out characteristic distribution correction on the temperature time sequence characteristic vector of the object to be cooled, the outlet temperature characteristic vector and the flow velocity characteristic vector to obtain a corrected temperature time sequence characteristic vector of the object to be cooled, a corrected outlet temperature characteristic vector and a corrected flow velocity characteristic vector; fusing the corrected temperature time sequence characteristic vector of the object to be cooled, the corrected outlet temperature characteristic vector and the corrected flow velocity characteristic vector by using a Bayes-like probability model to obtain a posterior probability vector; and passing the posterior probability vector through a classifier to obtain a classification result, wherein the classification result is used for indicating that the flow rate of the cooling liquid at the current time point should be increased or decreased.

10. The control method of the intelligent cooling liquid circulation control system for hexafluorobutadiene production according to claim 9, wherein the performing feature distribution correction on the to-be-cooled object temperature time series feature vector, the outlet temperature feature vector and the flow rate feature vector to obtain a corrected to-be-cooled object temperature time series feature vector, a corrected outlet temperature feature vector and a corrected flow rate feature vector respectively comprises:

respectively carrying out characteristic distribution correction on the temperature time sequence characteristic vector of the object to be cooled, the outlet temperature characteristic vector and the flow velocity characteristic vector by the following formulas to obtain the corrected temperature time sequence characteristic vector of the object to be cooled, the corrected outlet temperature characteristic vector and the corrected flow velocity characteristic vector; wherein the formula is:

wherein

、

And

The characteristic value of each position is calculated,

、

and

The characteristic value of each position.