Industrial control system for preparation of dual modified adsorbents for purification of carbon dioxide-based polyols

Abstract

The invention relates to the technical field of adsorption materials, and particularly discloses an industrial control system for preparing a dual-modified adsorbent for purifying carbon dioxide-based polyol. The method comprises the following steps: carrying out time sequence analysis on adsorbent solution pH, carbon dioxide purging pressure and reaction temperature data in a carbon dioxide gas modification process by utilizing a data processing technology based on deep learning, and carrying out comprehensive characterization on reaction condition state change based on time sequence joint information of the carbon dioxide purging pressure and the reaction temperature data; the synergistic effect of the carbon dioxide purging pressure and the reaction temperature on the pH of the solution is captured by performing sequential correlation interaction analysis on the reaction condition characteristics and the pH characteristics of the solution, so that the dynamic optimization control on the carbon dioxide purging pressure is realized. Therefore, dynamic optimization control of reaction conditions in the preparation process of the dual-modified adsorbent can be realized, so that the preparation efficiency and the product quality of the dual-modified adsorbent are improved, and the production cost is reduced.

Description

Technical Field

[0001] The present application relates to the technical field of adsorption materials, and more specifically, to an industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols. Background Art

[0002] With the rapid development of CO2-based polyols, phenolic compounds (such as bisphenol A) as additives make polymer products have excellent properties in terms of high modulus, acid resistance, and hydrolysis resistance. However, there is also the problem that residual phenolic compounds cannot be removed, resulting in polymer products with transparency, color, and residual pollutants that cannot meet the requirements of downstream applications. At present, the main methods for removing phenolic compounds include adsorption, ultrasonic photocatalytic degradation, biodegradation, electrochemical catalytic degradation, nanofiltration, etc. Among them, the adsorption method is generally considered to be superior to other methods because of its simple operation, low cost, ability to quickly and effectively remove pollutants, and not prone to secondary pollution.

[0003] In the adsorption method, commercial activated carbon is widely used as an excellent adsorbent to remove phenolic compound pollutants. The double-modified adsorbent is modified by modifying the traditional activated carbon adsorbent to improve its adsorption performance for phenolic compounds and heavy metal ions in CO2-based polyols.

[0004] Accordingly, the preparation process of the double-modified adsorbent for purifying CO2-based polyols requires precise control to ensure that the modifier is evenly distributed and effectively bound to the adsorbent matrix. However, the traditional preparation method of double-modified adsorbents often relies on empirical judgment and manual operation, lacks fine control of reaction conditions, and is prone to deviations from the ideal state, thereby affecting the modification effect and final performance of the adsorbent, making it difficult to ensure repeatability and product quality during the preparation process.

[0005] Therefore, an industrial control system for the preparation of double-modified adsorbents for purifying CO2-based polyols is desired. Summary of the invention

[0006] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an industrial control system for the preparation of a double-modified adsorbent for purifying carbon dioxide-based polyols (hereinafter referred to as a double-modified adsorbent), which uses a data processing technology based on deep learning to perform time series analysis on the adsorbent solution pH, carbon dioxide purge pressure and reaction temperature data in the carbon dioxide gas modification process, and comprehensively characterizes the change in the state of the reaction conditions based on the time series joint information of the carbon dioxide purge pressure and the reaction temperature data. Then, by performing a time series correlation interaction analysis on the reaction condition characteristics and the solution pH characteristics, the synergistic effect of the carbon dioxide purge pressure and the reaction temperature on the solution pH is captured, thereby achieving dynamic optimization control of the carbon dioxide purge pressure. In this way, dynamic optimization control of the reaction conditions in the preparation process of the double-modified adsorbent can be achieved, thereby improving the preparation efficiency and product quality of the double-modified adsorbent and reducing production costs.

[0007] Accordingly, according to one aspect of the present application, an industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols is provided, comprising: A modification process monitoring module, used for collecting a time queue of a real-time solution pH value, a time queue of a real-time carbon dioxide purge pressure value, and a time queue of a real-time reaction temperature through a sensor matrix; A data transmission module, used to transmit the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value and the time queue of the real-time reaction temperature to the industrial intelligent control center through a 5G communication module; A reaction state causal time series feature extraction module is used to extract a time series correlation feature vector of carbon dioxide purge pressure and a time series correlation feature vector of solution pH under temperature modulation from a time series of the real-time solution pH value, a time series of the real-time carbon dioxide purge pressure value and a time series correlation feature vector of solution pH under temperature modulation in the industrial intelligent control center; A reaction state causal interaction analysis module, used for performing a reaction state causal interaction analysis on the carbon dioxide purge pressure time series correlation feature vector and the solution pH time series correlation feature vector under the temperature modulation to obtain a reaction state causal-effect time series collaborative representation vector; The purge pressure optimization control module is used to generate an optimization instruction for the carbon dioxide purge pressure based on the reaction state cause-effect time series collaborative representation vector.

[0008] In the above-mentioned industrial control system for the preparation of double-modified adsorbents for purifying carbon dioxide-based polyols, the reaction state causal interaction analysis module includes: a conditional feature construction unit, which is used to extract the implicit correlation features between the carbon dioxide purge pressure time series correlation feature vector under the temperature modulation and the solution pH time series correlation feature vector as a conditional feature vector; a cross-domain interactive encoding unit, which is used to guide the carbon dioxide purge pressure time series correlation feature vector under the temperature modulation and the solution pH time series correlation feature vector to perform cross-domain interactive encoding based on the attention mechanism based on the conditional feature vector to obtain the reaction state causal-effect time series collaborative representation vector.

[0009] In the above-mentioned industrial control system for the preparation of double-modified adsorbents for purifying carbon dioxide-based polyols, the reaction state causal time series feature extraction module includes: a time series encoding unit, which is used to input the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value and the time queue of the real-time reaction temperature into a time series encoder based on a forward LSTM model to obtain the solution pH time series association feature vector, the carbon dioxide purge pressure time series association feature vector and the reaction temperature time series association feature vector; a joint modulation unit, which is used to input the carbon dioxide purge pressure time series association feature vector and the reaction temperature time series association feature vector into a reaction state causal joint modulation network to obtain the carbon dioxide purge pressure time series association feature vector under the temperature modulation.

[0010] In the above-mentioned industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols, the joint modulation unit is used to: calculate the Hadamard product between the carbon dioxide purge pressure time series correlation characteristic vector and the reaction temperature time series correlation characteristic vector to obtain the carbon dioxide purge pressure time series correlation characteristic vector under the temperature modulation.

[0011] In the above-mentioned industrial control system for the preparation of double-modified adsorbents for purifying carbon dioxide-based polyols, the conditional feature construction unit includes: an implicit correlation feature extraction subunit, which is used to input the carbon dioxide purge pressure time series correlation feature vector under temperature modulation and the solution pH time series correlation feature vector into an implicit correlation feature capture network to obtain a reaction state cause-effect time series implicit correlation feature vector; a feature activation subunit, which is used to perform feature activation based on a Sigmoid function on the reaction state cause-effect time series implicit correlation feature vector to obtain the conditional feature vector.

[0012] In the above-mentioned industrial control system for the preparation of double-modified adsorbents for purifying carbon dioxide-based polyols, the implicit correlation feature extraction subunit is used to: fuse the carbon dioxide purge pressure time series correlation feature vector under temperature modulation and the solution pH time series correlation feature vector according to the position points and pass them through a neural network layer based on the tanh function to obtain the reaction state cause-effect time series implicit correlation feature vector.

[0013] In the above-mentioned industrial control system for the preparation of double-modified adsorbents for purifying carbon dioxide-based polyols, the cross-domain interactive coding unit includes: a feature modulation subunit, which is used to perform feature modulation on the carbon dioxide purge pressure time series association feature vector under temperature modulation and the solution pH time series association feature vector based on the semantic contribution of the carbon dioxide purge pressure time series association feature vector under temperature modulation and the solution pH time series association feature vector relative to the conditional feature vector to obtain the carbon dioxide purge pressure time series association feature vector under temperature modulation after modulation and the solution pH time series association feature vector after modulation; a guided interaction subunit, which is used to perform feature-significant guided interaction encoding on the carbon dioxide purge pressure time series association feature vector under temperature modulation after modulation, the solution pH time series association feature vector after modulation and the conditional feature vector to obtain the reaction state cause-effect time series collaborative representation vector.

[0014] In the above-mentioned industrial control system for the preparation of a double-modified adsorbent for purifying carbon dioxide-based polyols, the feature modulation subunit is used to: calculate the first semantic contribution factor of the carbon dioxide purge pressure time series association feature vector under the temperature modulation relative to the conditional feature vector; calculate the second semantic contribution factor of the solution pH time series association feature vector to the conditional feature vector; normalize the first semantic contribution factor and the second semantic contribution factor, and use the normalized first semantic contribution factor and the second semantic contribution factor to weighted modulate the carbon dioxide purge pressure time series association feature vector under temperature modulation and the solution pH time series association feature vector to obtain the modulated carbon dioxide purge pressure time series association feature vector under temperature modulation and the modulated solution pH time series association feature vector.

[0015] In the above-mentioned industrial control system for the preparation of double-modified adsorbents for purifying carbon dioxide-based polyols, the guided interaction subunit is used to: use the carbon dioxide purge pressure time series association feature vector under the modulation temperature modulation as the query vector, the modulated solution pH time series association feature vector as the key vector and the conditional feature vector as the value vector, and perform significant guided interaction between features based on the converter structure on the carbon dioxide purge pressure time series association feature vector under the modulation temperature modulation, the modulated solution pH time series association feature vector and the conditional feature vector to obtain the reaction state cause-effect time series collaborative representation vector.

[0016] In the above-mentioned industrial control system for the preparation of double-modified adsorbents for purifying carbon dioxide-based polyols, the purge pressure optimization control module is used to: input the reaction state cause-effect timing synergistic representation vector into the decoder-based purge pressure dynamic optimizer to obtain the optimization instruction, and the optimization instruction is used to represent the recommended value of the carbon dioxide purge pressure at the next moment.

[0017] Compared with the prior art, the industrial control system for the preparation of double-modified adsorbents for purifying carbon dioxide-based polyols provided in this application uses data processing technology based on deep learning to perform time series analysis on the adsorbent solution pH, carbon dioxide purge pressure and reaction temperature data in the carbon dioxide gas modification process, and performs a comprehensive characterization of the change in the state of the reaction conditions based on the time series joint information of the carbon dioxide purge pressure and reaction temperature data. Then, by performing time series correlation interaction analysis on the reaction condition characteristics and the solution pH characteristics, the synergistic effect of the carbon dioxide purge pressure and the reaction temperature on the solution pH is captured, thereby achieving dynamic optimization control of the carbon dioxide purge pressure. In this way, dynamic optimization control of the reaction conditions in the preparation process of the double-modified adsorbent can be achieved, thereby improving the preparation efficiency and product quality of the double-modified adsorbent and reducing production costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0019] Figure 1 Schematic diagram of a carbon dioxide purge device according to an embodiment of the present application.

[0020] Figure 2 The block diagram is a industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to an embodiment of the present application.

[0021] Figure 3 The data flow diagram of the industrial control system for preparing the double-modified adsorbent for purifying carbon dioxide-based polyols according to an embodiment of the present application is shown.

[0022] Figure 4 The block diagram is a reaction state causal time series feature extraction module in an industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to an embodiment of the present application.

[0023] Figure 5 The block diagram is a reaction state causal interaction analysis module in an industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to an embodiment of the present application.

[0024] Figure 6 The block diagram is a cross-domain interactive coding unit in an industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to an embodiment of the present application. DETAILED DESCRIPTION

[0025] Below, the embodiments of the present application will be described in more detail in conjunction with the accompanying drawings, and the above and other purposes, features and advantages of the present application will become more apparent. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the example embodiments described herein.

[0026] In view of the above technical problems, the present application proposes an optimized preparation method of a six-membered nitrogen heterocycle and carbon dioxide gas dual-modified adsorbent, wherein the adsorbent is prepared through the steps of pre-dispersion, six-membered nitrogen heterocycle modification, high-temperature calcination, and carbon dioxide gas modification, specifically comprising: (1) pre-dispersion: dispersing activated carbon in deionized water to prepare an activated carbon dispersion to form an activated and stable activated carbon precursor; (2) six-membered nitrogen heterocycle modification: adding a six-membered nitrogen heterocycle modifier to the activated carbon dispersion according to a mass ratio of the six-membered nitrogen heterocycle modifier to the activated carbon of 1:0.5-20, and heating and refluxing to prepare a six-membered nitrogen heterocycle modified activated carbon; (3) high-temperature calcination: activating the six-membered nitrogen heterocycle modified activated carbon by high-temperature calcination to obtain an activated adsorbent; (4) carbon dioxide gas modification: preparing a solution of the activated adsorbent using water as a solvent, dispersing and soaking it, and blowing it with carbon dioxide gas until the solution pH is 6-8, filtering, and vacuum drying to obtain the six-membered nitrogen heterocycle and carbon dioxide gas dual-modified adsorbent.

[0027] The rich variety of nitrogen functional groups in the hexagonal nitrogen heterocycle can significantly change the surface properties of activated carbon. Since the nitrogen atoms in the hexagonal nitrogen heterocycle usually carry lone pairs of electrons, the surface of the activated carbon exhibits stronger alkalinity, which is conducive to the adsorption of phenolic compounds in carbon dioxide-based polyols, thereby reducing the adsorption of non-target pollutants and improving the efficiency of carbon dioxide-based polyol treatment.

[0028] However, the enhancement of the alkaline surface brought by the six-membered nitrogen heterocycle forms a competitive relationship with the metal ions, which will reduce the adsorption of the adsorbent to the metal ions to a certain extent. Therefore, the present application is aimed at the characteristics of carbon dioxide-based polyols, and further adopts carbon dioxide gas purging modification after the six-membered nitrogen heterocycle modification and high-temperature calcination activation to enhance the acidic functional groups on the surface of the activated carbon, neutralize the alkalinity brought by the six-membered nitrogen heterocycle, and improve the adsorption capacity for metal ions. At the same time, it will not lead to a decrease in the alkaline functional groups, and make the entire modified activated carbon in a state close to neutral, so that it has a high adsorption capacity for phenolic compounds.

[0029] In one embodiment, the activated carbon is commercial activated carbon, selected from one or more of fruit shell activated carbon, coconut shell activated carbon, coal columnar activated carbon, wood activated carbon, medical activated carbon, apricot shell activated carbon, and medicinal activated carbon, more preferably coconut shell activated carbon.

[0030] In one embodiment, the six-membered nitrogen-containing heterocyclic modifier is one or more of pyridine, pyrazine, pyrimidine, pyridazine, triazine compounds and their derivatives, more preferably pyrazine.

[0031] In one embodiment, the stirring time is 1-5 h to form an azacyclic-ethanol solution with a pH>7; In one embodiment, the stirring temperature is 30-80°C; In one embodiment, the mass of the ethanol solution is 15-30 times the mass of the six-membered nitrogen heterocyclic modifier.

[0032] In one embodiment, in step (1), the activated carbon is soaked and stirred for 2-12 hours to form an activated carbon precursor; In one embodiment, the soaking temperature is 10-50°C; In one embodiment, the mass of the deionized water is 10-20 times the mass of the activated carbon.

[0033] In one embodiment, in step (2), the mass ratio of the six-membered nitrogen heterocyclic modifier to the activated carbon is 1:5-1:15; In one embodiment, the heating reflux temperature is 100-160°C; In one embodiment, the heating reflux time is 3-16 h.

[0034] In one embodiment, in step (3), the high temperature calcination activation time is 2-9 h, and the activation temperature is 300-800°C.

[0035] In one embodiment, in step (4), when the carbon dioxide gas is modified, the carbon dioxide purge pressure is a low pressure of 0.2-1.5 MPa, the temperature is 30-90°C, and the reaction time is 2-10 h.

[0036] In one embodiment, in step (4), the method for modifying the carbon dioxide gas is: using a carbon dioxide purge device, such as Figure 1 As shown, carbon dioxide gas is continuously introduced from the air inlet of the bottom pipe, the activated adsorbent enters from the feed inlet at the top of the reaction tank, and is stirred by a stirring paddle in the reaction tank. Excess carbon dioxide gas circulates into the reaction tank from the top of the reaction tank via a gas booster pump. A sampling port is set at the bottom of the reaction tank. When the pH is detected to be 6-8, the adsorbent is obtained from the discharge port of the bottom pipe.

[0037] In particular, considering that during the carbon dioxide gas purge modification process, the pH value of the adsorbent solution will be affected by the carbon dioxide purge and change dynamically, and the traditional double-modified adsorbent preparation method often relies on empirical judgment and manual operation to adjust the reaction conditions to ensure that the pH value of the solution remains within the optimal range, which is not only inefficient, but also easy to introduce human errors, making it difficult to ensure the consistency of the modification effect, thereby affecting the modification effect and final performance of the adsorbent. In this regard, the present application further proposes an industrial control system for the preparation of double-modified adsorbents for purifying carbon dioxide-based polyols to optimize the control of the carbon dioxide gas modification process. Specifically, the industrial control system for preparing the double-modified adsorbent for purifying carbon dioxide-based polyols uses a data processing technology based on deep learning to perform time series analysis on the adsorbent solution pH, carbon dioxide purge pressure and reaction temperature data in the carbon dioxide gas modification process, and performs a comprehensive characterization of the change in the state of the reaction conditions based on the time series joint information of the carbon dioxide purge pressure and reaction temperature data, and then, through the time series correlation interaction analysis of the reaction condition characteristics and the solution pH characteristics, the synergistic effect of the carbon dioxide purge pressure and the reaction temperature on the solution pH is captured, thereby achieving dynamic optimization control of the carbon dioxide purge pressure. In this way, dynamic optimization control of the reaction conditions in the preparation process of the double-modified adsorbent can be achieved, thereby improving the preparation efficiency and product quality of the double-modified adsorbent and reducing production costs.

[0038] Figure 2 The block diagram is a industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to an embodiment of the present application. Figure 3The data flow diagram of the industrial control system for preparing the double-modified adsorbent for purifying carbon dioxide-based polyols according to the embodiment of the present application is shown in FIG. Figure 2 and Figure 3 As shown, according to an embodiment of the present application, an industrial control system 100 for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols includes: a modification process monitoring module 110, which is used to collect a time queue of a real-time solution pH value, a time queue of a real-time carbon dioxide purge pressure value, and a time queue of a real-time reaction temperature through a sensor matrix; a data transmission module 120, which is used to transmit the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value, and the time queue of the real-time reaction temperature to an industrial intelligent control center through a 5G communication module; a reaction state causal time series feature extraction module 130, which is used in the industrial intelligent control center, A carbon dioxide purge pressure time series correlation feature vector and a solution pH time series correlation feature vector under temperature modulation are extracted from the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value and the time queue of the real-time reaction temperature; a reaction state causal interaction analysis module 140 is used to perform a reaction state causal interaction analysis on the carbon dioxide purge pressure time series correlation feature vector and the solution pH time series correlation feature vector under temperature modulation to obtain a reaction state cause-effect time series collaborative representation vector; a purge pressure optimization control module 150 is used to generate an optimization instruction for the carbon dioxide purge pressure based on the reaction state cause-effect time series collaborative representation vector.

[0039] In the above-mentioned industrial control system 100 for preparing double-modified adsorbents for purifying carbon dioxide-based polyols, the modification process monitoring module 110 is used to collect the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value and the time queue of the real-time reaction temperature through the sensor matrix. It should be understood that in the process of carbon dioxide gas modification, the change of the pH value of the adsorbent solution not only directly affects the chemical properties of the adsorbent surface, but is also closely related to its adsorption efficiency and selectivity; while the carbon dioxide purge pressure and reaction temperature indirectly act on the structure and properties of the adsorbent by affecting the solubility, diffusion rate and chemical reaction rate of carbon dioxide gas in the solution, thereby causing changes in the pH value of the adsorbent solution. Therefore, in the technical solution of the present application, it is expected to utilize the time-series influence mechanism of carbon dioxide purge pressure and reaction temperature on solution pH to realize feedback control of carbon dioxide purge pressure and achieve the purpose of optimizing adsorbent performance.

[0040] In the above-mentioned industrial control system 100 for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols, the data transmission module 120 is used to transmit the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value, and the time queue of the real-time reaction temperature to the industrial intelligent control center through the 5G communication module. It should be understood that, considering the real-time requirements of the industrial control process for the control of reaction conditions, the present application further uses 5G communication technology to transmit the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value, and the time queue of the real-time reaction temperature to the industrial intelligent control center, so as to utilize the rich computing resources and powerful data processing capabilities of the industrial intelligent control center to analyze and process the data in real time, and realize real-time monitoring and dynamic optimization of the preparation process of the double-modified adsorbent. In addition, by using 5G communication technology for data transmission, high speed and low latency of data transmission can be ensured.

[0041] In the above industrial control system 100 for preparing double-modified adsorbents for purifying carbon dioxide-based polyols, the reaction state causal time series feature extraction module 130 is used to extract the carbon dioxide purge pressure time series associated feature vector and the solution bottle pH time series associated feature vector under temperature modulation from the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value and the time queue of the real-time reaction temperature in the industrial intelligent control center. Figure 4 FIG. 1 is a block diagram of a reaction state causal time series feature extraction module in an industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to an embodiment of the present application. Figure 4 As shown, the reaction state causal time series feature extraction module 130 includes: a time series encoding unit 131, which is used to input the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value and the time queue of the real-time reaction temperature into a time series encoder based on a forward LSTM model to obtain the solution pH time series association feature vector, the carbon dioxide purge pressure time series association feature vector and the reaction temperature time series association feature vector; a joint modulation unit 132, which is used to input the carbon dioxide purge pressure time series association feature vector and the reaction temperature time series association feature vector into a reaction state causal joint modulation network to obtain the carbon dioxide purge pressure time series association feature vector under the temperature modulation.

[0042] Specifically, the time series encoding unit 131 is used to input the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value, and the time queue of the real-time reaction temperature into the time series encoder based on the forward LSTM model to obtain the solution pH time series associated feature vector, the carbon dioxide purge pressure time series associated feature vector, and the reaction temperature time series associated feature vector. It should be understood that since the adsorbent solution pH, carbon dioxide purge pressure, and reaction temperature are dynamic parameters that change over time, in order to accurately capture the dynamic change law of each reaction parameter, the present application further uses a time series encoder based on the forward LSTM model to process the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value, and the time queue of the real-time reaction temperature, respectively, so as to use the long-term memory ability of the forward LSTM model to learn and memorize the long-term dependency in the time series data, capture the time series change pattern of the solution pH value, carbon dioxide purge pressure, and reaction temperature, and thus obtain the solution pH time series associated feature vector, the carbon dioxide purge pressure time series associated feature vector, and the reaction temperature time series associated feature vector.

[0043] Specifically, the joint modulation unit 132 is used to input the carbon dioxide purge pressure time series correlation feature vector and the reaction temperature time series correlation feature vector into the reaction state factor joint modulation network to obtain the carbon dioxide purge pressure time series correlation feature vector under the temperature modulation. In a specific example of the present application, the joint modulation unit 132 is used to: calculate the Hadamard product between the carbon dioxide purge pressure time series correlation feature vector and the reaction temperature time series correlation feature vector to obtain the carbon dioxide purge pressure time series correlation feature vector under the temperature modulation. It should be understood that considering that the reaction temperature can affect the solubility and diffusion rate of carbon dioxide gas in the solution, thereby indirectly affecting the performance of the adsorbent. Therefore, in order to capture the synergistic effect between the reaction temperature and the carbon dioxide purge pressure, in the technical solution of the present application, the reaction state factor joint modulation network is further used to jointly analyze the carbon dioxide purge pressure time series correlation feature vector and the reaction temperature time series correlation feature vector. Specifically, the reaction state joint modulation network performs information fusion by calculating the Hadamard product between the carbon dioxide purge pressure time series correlation feature vector and the reaction temperature time series correlation feature vector, and generates the carbon dioxide purge pressure time series correlation feature vector under temperature modulation, so as to comprehensively reflect the joint influence of the carbon dioxide purge pressure and the reaction temperature, thereby providing a more comprehensive characterization of the reaction condition information.

[0044] In the above-mentioned industrial control system 100 for preparing double-modified adsorbents for purifying carbon dioxide-based polyols, the reaction state causal interaction analysis module 140 is used to perform reaction state causal interaction analysis on the carbon dioxide purge pressure time series correlation feature vector and the solution pH time series correlation feature vector under the temperature modulation to obtain the reaction state cause-effect time series synergistic representation vector. It should be understood that in order to fully reveal the causal relationship between the carbon dioxide purge pressure, the reaction temperature and the solution pH value, the present application further performs interactive response analysis on the carbon dioxide purge pressure time series correlation feature vector and the solution pH time series correlation feature vector under the temperature modulation to determine the interaction and influence mechanism between the carbon dioxide purge pressure and the solution pH under the influence of temperature. It is worth mentioning that in order to improve the accuracy of feature interaction analysis, the present application proposes a significant guided interaction method between features, which guides the significant feature interaction between the two by mining the implicit correlation features between the carbon dioxide purge pressure time series correlation feature vector and the solution pH time series correlation feature vector under the temperature modulation, so as to effectively identify and strengthen the important correlation mode between the two, and enhance the distinguishability and representativeness of the features.

[0045] Figure 5 FIG. 1 is a block diagram of a reaction state causal interaction analysis module in an industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to an embodiment of the present application. Figure 5 As shown, the reaction state causal interaction analysis module 140 includes: a conditional feature construction unit 141, which is used to extract the implicit correlation features between the carbon dioxide purge pressure time series correlation feature vector under the temperature modulation and the solution pH time series correlation feature vector as a conditional feature vector; a cross-domain interaction encoding unit 142, which is used to guide the carbon dioxide purge pressure time series correlation feature vector under the temperature modulation and the solution pH time series correlation feature vector to perform cross-domain interaction encoding based on the attention mechanism based on the conditional feature vector to obtain the reaction state cause-effect time series collaborative representation vector.

[0046] Specifically, the conditional feature construction unit 141 includes: an implicit correlation feature extraction subunit, which is used to input the carbon dioxide purge pressure time series correlation feature vector under temperature modulation and the solution pH time series correlation feature vector into an implicit correlation feature capture network to obtain a reaction state cause-effect time series implicit correlation feature vector; a feature activation subunit, which is used to perform feature activation based on the Sigmoid function on the reaction state cause-effect time series implicit correlation feature vector to obtain the conditional feature vector. Among them, the implicit correlation feature extraction subunit is used to: fuse the carbon dioxide purge pressure time series correlation feature vector under temperature modulation and the solution pH time series correlation feature vector according to the position point, and then pass through the neural network layer based on the tanh function to obtain the reaction state cause-effect time series implicit correlation feature vector.

[0047] That is, the implicit correlation feature capture network is used to mine the implicit correlation shared between the carbon dioxide purge pressure time series correlation feature vector under temperature modulation and the solution pH time series correlation feature vector, and the Sigmoid function is used to perform nonlinear transformation on the extracted implicit correlation features, and the smoothness characteristic of the Sigmoid function is used to adjust the feature weights, emphasize key information and suppress irrelevant noise, thereby constructing a conditional feature vector to guide the subsequent feature interaction fusion process.

[0048] Specifically, Figure 6 FIG. 1 is a block diagram of a cross-domain interactive coding unit in an industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to an embodiment of the present application. Figure 6 As shown, the cross-domain interaction encoding unit 142 includes: a feature modulation subunit 1421, which is used to perform feature modulation on the carbon dioxide purge pressure time series association feature vector under temperature modulation and the solution pH time series association feature vector based on the semantic contribution of the carbon dioxide purge pressure time series association feature vector under temperature modulation and the solution pH time series association feature vector relative to the conditional feature vector to obtain the carbon dioxide purge pressure time series association feature vector under temperature modulation after modulation and the solution pH time series association feature vector after modulation; a guided interaction subunit 1422, which is used to perform inter-feature significant guided interaction encoding on the carbon dioxide purge pressure time series association feature vector under temperature modulation after modulation, the solution pH time series association feature vector after modulation and the conditional feature vector to obtain the reaction state cause-effect time series collaborative representation vector.

[0049] More specifically, the feature modulation subunit 1421 is further used to: calculate the first semantic contribution factor of the carbon dioxide purge pressure time series association feature vector under the temperature modulation relative to the conditional feature vector; calculate the second semantic contribution factor of the solution pH time series association feature vector to the conditional feature vector; normalize the first semantic contribution factor and the second semantic contribution factor, and use the normalized first semantic contribution factor and the second semantic contribution factor to weighted modulate the carbon dioxide purge pressure time series association feature vector under temperature modulation and the solution pH time series association feature vector to obtain the carbon dioxide purge pressure time series association feature vector under temperature modulation after modulation and the solution pH time series association feature vector after modulation. Among them, calculating the first semantic contribution factor of the timing-related feature vector of the carbon dioxide purge pressure under temperature modulation relative to the conditional feature vector includes: calculating the dot division vector between the timing-related feature vector of the carbon dioxide purge pressure under temperature modulation and the conditional feature vector, and calculating the base-two logarithmic value of the absolute value of each eigenvalue in the dot division vector to obtain a semantic contribution representation vector; calculating the dot product vector between the semantic contribution representation vector and the timing-related feature vector of the carbon dioxide purge pressure under temperature modulation, and calculating an exponential function value with base e and the sum of the eigenvalues ​​of the dot product vector as an exponent to obtain the first semantic contribution factor.

[0050] That is, the semantic contribution of the carbon dioxide purge pressure time series correlation feature vector under temperature modulation and the solution pH time series correlation feature vector are respectively evaluated with the conditional feature vector to determine the importance of the two for the implicit correlation feature representation, and the original carbon dioxide purge pressure time series correlation feature vector under temperature modulation and the solution pH time series correlation feature vector are weighted modulated thereby effectively emphasizing the key information in the two while suppressing irrelevant information.

[0051] More specifically, the guiding interaction subunit 1422 is further used to: use the time series correlation feature vector of the carbon dioxide purge pressure under the modulation temperature modulation as the query vector, the time series correlation feature vector of the solution pH after modulation as the key vector and the conditional feature vector as the value vector, and perform significant guiding interaction between the features based on the converter structure on the time series correlation feature vector of the carbon dioxide purge pressure under the modulation temperature modulation, the time series correlation feature vector of the solution pH after modulation and the conditional feature vector to obtain the reaction state cause-effect time series collaborative representation vector. In a specific example of the present application, the guiding interaction subunit 1422 is also used to multiply the time series correlation feature vector of the carbon dioxide purge pressure under the modulation temperature modulation by the transposed vector of the time series correlation feature vector of the solution pH after modulation and then divide it by the square root of the length of the solution pH time series correlation feature vector to obtain the reaction state cause-effect time series query attention score matrix; multiply the reaction state cause-effect time series query attention score matrix by the conditional feature vector after passing the softmax function to obtain the reaction state cause-effect time series collaborative representation vector.

[0052] That is, the time series associated feature vector of the carbon dioxide purge pressure under temperature modulation is used as the query vector, the time series associated feature vector of the modulated solution pH is used as the key vector, and the conditional feature vector is used as the value vector. The Transformer architecture is used to guide the interaction between features, and the complex relationship between the reaction condition information and the reaction results is dynamically captured and utilized based on the self-attention mechanism to achieve deep interactive fusion of cross-modal features, so as to generate a reaction state cause-effect time series collaborative representation vector, thereby achieving an in-depth understanding of the reaction mechanism of carbon dioxide gas modification.

[0053] Accordingly, the reaction state causal interaction analysis module 140 is used to process the carbon dioxide purge pressure time series correlation feature vector and the solution pH time series correlation feature vector under temperature modulation using the following feature interaction analysis formula to obtain the reaction state causal-effect time series collaborative representation vector, wherein the feature interaction analysis formula is:

[0054]

[0055]

[0056]

[0057]

[0058]

[0059]

[0060]

[0061]

[0062] in, represents the time series correlation feature vector of the carbon dioxide purge pressure under the temperature modulation, represents the solution pH time series correlation feature vector, Indicates adding by position point. represents the implicit correlation feature vector of the reaction state cause-effect time series, and Represent the weight matrix and bias parameters of the neural network layer, respectively. represents the hyperbolic tangent function, represents the sigmoid activation function, represents the conditional eigenvector, , and respectively represent the first eigenvalue, the first eigenvalue in the solution pH time series correlation eigenvector eigenvalue and the conditional eigenvector eigenvalues, is the absolute value symbol, represents the logarithmic function with base 2, represents the exponential function with base e, represents the length of the time series correlation feature vector of the carbon dioxide purge pressure under the temperature modulation, and the time series correlation feature vector of the carbon dioxide purge pressure under the temperature modulation, the time series correlation feature vector of the solution pH and the conditional feature vector have the same length, and Respectively represent the first semantic contribution factor and the second semantic contribution factor, and Respectively represent the normalized first semantic contribution factor and the second semantic contribution factor, and They represent the time series correlation feature vector of carbon dioxide purge pressure under temperature modulation after modulation and the time series correlation feature vector of solution pH after modulation, represents the transpose of a vector, represents the matrix multiplication operation, is the normalized exponential function, A vector representing the cause-effect temporal synergy of the reaction state.

[0063] In the above-mentioned industrial control system 100 for preparing double-modified adsorbents for purifying carbon dioxide-based polyols, the purge pressure optimization control module 150 is used to generate an optimization instruction for the carbon dioxide purge pressure based on the reaction state cause-effect time series synergistic representation vector. In a specific example of the present application, the purge pressure optimization control module 150 is further used to: input the reaction state cause-effect time series synergistic representation vector into a purge pressure dynamic optimizer based on a decoder to obtain the optimization instruction, and the optimization instruction is used to represent the recommended value of the carbon dioxide purge pressure at the next moment. It should be understood that the decoder is based on a multi-layer neural network structure, which can effectively learn the dynamic influence of the carbon dioxide purge pressure on the pH value of the adsorbent solution by performing multi-layer nonlinear transformation on the reaction state cause-effect time series synergistic representation vector, so as to intelligently predict the recommended value of the carbon dioxide purge pressure under the current reaction state based on this, so that it can keep matching with the pH value of the adsorbent solution, so as to achieve the purpose of improving the adsorbent performance and reaction efficiency.

[0064] In the technical solution of the present application, the carbon dioxide purge pressure time series correlation feature vector under temperature modulation and the solution pH time series correlation feature vector respectively represent the carbon dioxide purge pressure time series correlation implicit feature under temperature modulation and the solution pH value time series correlation implicit feature. In this way, when the reaction state cause-effect analysis network based on the significant guided interaction network between features with joint implicit features as conditional information is input, considering that the correlation between the reaction temperature and the carbon dioxide purge pressure is a nonlinear correlation, there is a significant difference in time series distribution and feature modal expression between the carbon dioxide purge pressure time series correlation feature vector under temperature modulation and the solution pH time series correlation feature vector, which makes the extraction of joint implicit features insufficient and inaccurate, resulting in the expression sparsity and distribution imbalance of the feature manifold of the reaction state cause-effect time series collaborative representation vector obtained by the significant guided interaction between features in the high-dimensional feature space, thereby affecting the accuracy of the optimization instructions obtained by the purge pressure dynamic optimizer based on the decoder.

[0065] Based on this, in a preferred example of the present application, the reaction state cause-effect timing synergistic representation vector is input into a decoder-based purge pressure dynamic optimizer to obtain an optimization instruction, including: performing L random sampling on the reaction state cause-effect timing synergistic representation vector to obtain L reaction state cause-effect association random disturbance eigenvalues; arranging the L reaction state cause-effect association random disturbance eigenvalues ​​in order from large to small to obtain a reaction state cause-effect association random disturbance ordered vector; calculating the self-inner product of the reaction state cause-effect timing synergistic representation vector to obtain a reaction state cause-effect association energy modulation eigenvalue; and arranging the reaction state cause-effect association random disturbance eigenvalues ​​based on the reaction state cause-effect association energy modulation eigenvalue. The machine disturbance ordered vector is semantically scaled to obtain a semantically enhanced reaction state cause-effect association random disturbance ordered vector; the product between the semantically enhanced reaction state cause-effect association random disturbance ordered vector and the transposed vector of the reaction state cause-effect temporal synergistic representation vector is calculated to obtain a reaction state cause-effect association best sample prompt orthogonal modulation matrix; the matrix product between the reaction state cause-effect association best sample prompt orthogonal modulation matrix and the reaction state cause-effect temporal synergistic representation vector is calculated to obtain an optimized reaction state cause-effect temporal synergistic representation vector; the optimized reaction state cause-effect temporal synergistic representation vector is input into the decoder-based purge pressure dynamic optimizer to obtain the optimization instruction.

[0066] Accordingly, the optimization process of the reaction state cause-effect time series synergistic representation vector is expressed by the following optimization formula:

[0067]

[0068]

[0069]

[0070]

[0071]

[0072]

[0073]

[0074] in, is the reaction state cause-effect time series synergistic representation vector, represents a normal distribution, represents a random variable that follows a normal distribution, represents the number of random samplings, and They represent the characteristic mean and characteristic variance of the reaction state cause-effect time series collaborative representation vector, represents the random disturbance eigenvalues ​​of the cause-effect correlation of each reaction state obtained by randomly sampling the reaction state cause-effect time series collaborative representation vector L times, is the ordered vector of random disturbances of the cause-effect association of the reaction state, represents the calculation of the inner product of the eigenvector itself, is the causal-effect energy modulation characteristic value of the reaction state, is the optimal example of the reaction state cause-effect association prompt orthogonal modulation matrix, It means point multiplication by position. represents the matrix multiplication operation, It is the optimized reaction state cause-effect timing synergistic representation vector.

[0075] That is, by randomly sampling the reaction state cause-effect temporal co-representation vector L times to obtain a set of new feature values ​​based on the original features but containing slight changes, randomness is introduced to explore different aspects of the feature space, and by simulating the noise or variation in the actual data, it helps the model to better generalize to unseen data. Then, these reaction state cause-effect association random perturbation feature values ​​are arranged in order from large to small to form a reaction state cause-effect association random perturbation ordered vector to identify which perturbation directions may be the most important for the current problem. Such an ordering not only helps the algorithm focus on the feature differences that can best distinguish different categories, but also helps to reveal hidden data structures or patterns. Subsequently, the self-inner product of the reaction state cause-effect temporal collaborative representation vector is calculated to obtain the reaction state cause-effect correlation energy modulation eigenvalue, which reflects the overall strength or "energy" of the original feature vector and provides a benchmark for measuring the importance of the feature. Using this reaction state cause-effect correlation energy modulation eigenvalue as a reference point, the previously formed reaction state cause-effect correlation random perturbation ordered vector is semantically scaled to construct a semantically enhanced reaction state cause-effect correlation random perturbation ordered vector. In this way, those disturbances that appear to be more important in the original feature space are amplified, thereby enhancing the model's sensitivity to changes in key features, while also retaining relatively small but potentially useful disturbance information.

[0076] Then, the product of the random perturbation ordered vector of semantic reinforcement reaction state causal-effect association and the transpose of the reaction state causal-effect temporal synergistic representation vector is calculated to obtain the reaction state causal-effect association best example prompt orthogonal modulation matrix, the purpose of which is to find the interaction between features from two different perspectives. The formed matrix can be regarded as a conversion tool for capturing and representing the complex linear relationship between features. Finally, the final feature structure modulation is completed by applying the above-constructed reaction state causal-effect association best example prompt orthogonal modulation matrix to the original reaction state causal-effect temporal synergistic representation vector, so that the features are more in line with the requirements of the objective function, reducing redundant information, while maximizing the retention of useful information, so as to improve learning efficiency and prediction performance.

[0077] Therefore, in a preferred example, the feature manifold optimization process of the reaction state cause-effect time series collaborative representation vector is explained, which extracts the best prompt sample based on the fine-grained target retrieval of random perturbations, and then guides adversarial robustness learning based on the best prompt sample to complete the structure of the high-dimensional manifold of the feature vector in the feature space to improve the detail and structure of the feature expression. In this way, the accuracy of the optimization instructions obtained by the decoder-based purge pressure dynamic optimizer is improved.

[0078] In summary, the industrial control system for the preparation of double-modified adsorbents for purifying carbon dioxide-based polyols is explained, which uses data processing technology based on deep learning to perform time series analysis on the adsorbent solution pH, carbon dioxide purge pressure and reaction temperature data in the carbon dioxide gas modification process, and comprehensively characterizes the change of reaction condition state based on the time series joint information of carbon dioxide purge pressure and reaction temperature data. Then, by performing time series correlation interaction analysis on reaction condition characteristics and solution pH characteristics, the synergistic effect of carbon dioxide purge pressure and reaction temperature on solution pH is captured, thereby realizing dynamic optimization control of carbon dioxide purge pressure. In this way, dynamic optimization control of reaction conditions in the preparation process of the double-modified adsorbent can be realized, thereby improving the preparation efficiency and product quality of the double-modified adsorbent and reducing production costs.

[0079] The specific embodiment of the present invention takes pyrazine as a six-membered nitrogen heterocyclic modifier as an example to further illustrate the preparation and effect of the modified adsorbent of the present application.

[0080] Example 1 The adsorbent is prepared according to the double-modified adsorbent preparation method as described above, and the specific steps are as follows: (1) Disperse 5 g of coconut shell activated carbon in 100 g of deionized water and soak for 3 h at 20 °C; (2) Dissolve 0.34 g of pyrazine in 10 g of ethanol solution, stir for 2 h, stir at 40°C, to form a pyrazine-ethanol solution with a pH of 8, then filter and wash to obtain a pyrazine modifier with a purity of 99.996%; then add the modifier to the coconut shell activated carbon solution prepared in the above step (1), heat at 80°C for 4 h, filter and dry to obtain activated carbon containing the pyrazine modifier; the purity of the pyrazine modifier is detected by high performance liquid chromatography.

[0081] (3) calcining the pyrazine-modified activated carbon obtained in step (2) above at 500° C. in a muffle furnace for 3 h to obtain activated activated carbon; (4) The activated carbon was purged with 1.5 MPa carbon dioxide for 8 h at a purging temperature of 90 °C, and then filtered and dried to obtain carbon dioxide-purged nitrogen heterocyclic modified activated carbon.

[0082] This application takes a carbon dioxide-based polyol containing heavy metal ions of zinc and cobalt as an example: the carbon dioxide-based polyol is prepared by copolymerization of carbon dioxide and propylene oxide under phenol initiator conditions using a zinc-cobalt double metal cyanide complex as a catalyst, and the process parameters of the copolymerization reaction can be conventional processes in the art. The carbon dioxide-based polyol has a number average molecular weight of 1000-15000, a molecular weight distribution of less than 2, and a viscosity of more than 7000 mPa·s at 70°C. The adsorption method can be: (1) 1 g of the modified adsorbent was placed in a container containing 100 g of CO2-based polyol and stirred at 300 r / min for 5 h. (2) heating and stirring the mixture in the above step (1) for adsorption; (3) transferring the adsorbed mixture into a pressurized filter tank; (4) The pressurized filter tank uses 0.1 MPa nitrogen pressure to filter; (5) Collecting the filtered carbon dioxide-based polyol to obtain purified carbon dioxide-based polyol.

[0083] Determination method of heavy metal ion Zn and Co content in carbon dioxide-based polyols before and after adsorption: accurately weigh about 1 g of the test object (accurate to 0.0001 g), and accurately record the weight M carbon dioxide-based polyol, place the test object in a 100 mL crucible, seal it and place it in an electric heating jacket for preheating to convert the liquid into a solid. The purpose of sealing is to prevent Zn and Co from being brought into the air with the gas or smoke produced by combustion, causing errors. Usually, the method of covering is adopted. Then, the crucible containing the test object is placed in a muffle furnace, set the temperature to 500℃, heat it for 2 h, and keep it at a constant temperature for 2 h, so that the sample is completely ash, and Zn and Co remain on the cover and wall of the crucible. Then, a 7% hydrochloric acid solution is used to dissolve the Zn and Co remaining in the crucible, and a solution is prepared in a 100 mL volumetric flask for testing.

[0084] Determination of Zn and Co content in the solution in the volumetric flask by ICP i ,ICP , in ppm, calculate the Zn and Co contents in the sample to be tested according to formula (1); Ci=(C i ,ICP ×100) / MCO2-based polyol (1) In the formula, C i Refers to the Zn or Co content in the purified CO2-based polyol sample, ppm; C i ,ICP Refers to the Zn or Co content in the test solution measured by ICP, ppm; MCO2-based polyol refers to the mass in the sample, g.

[0085] Removal rate of phenolic compounds: The absorbance corresponding to the maximum absorption wavelength (Amax) of phenolic compounds was detected by ultraviolet spectrophotometer, and then the standard curve was calculated according to the Lambert-Beer law, and finally the concentration of phenolic compounds in carbon dioxide-based polyols was obtained.

[0086]

[0087] ŋ represents the adsorption rate of modified activated carbon for phenolic compounds (%), C 0 represents the initial concentration of phenolic compounds in CO2-based polyols (mg / L), C t It represents the concentration of phenolic compounds in CO2-based polyols after adsorption by modified activated carbon (mg / L).

[0088] The specific test results are shown in Table 1 below: Table 1 Traditional fixed parameter control strategy and dynamic optimization control strategy Purification conditions and product performance test table of the prepared adsorbent

[0089] It can be seen from the data in Table 1 that Example 1 adopts the traditional fixed parameter control strategy. Compared with Example 1, other parameters and reaction materials in Example 2 have not been changed. The only change is that the dynamic optimization control strategy of carbon dioxide purge pressure as described above in the present invention is adopted in Example 2. As can be seen from Table 1, the dynamic optimization control strategy of carbon dioxide purge pressure provided by the present application is significantly shortened in adsorption time compared with the traditional fixed parameter carbon dioxide purge pressure control strategy, but the adsorption effect on the product is significantly improved. This shows that the dynamic optimization control method of carbon dioxide purge pressure described in the present invention is used in the modification of carbon dioxide gas to optimize the adsorption performance of the adsorbent, and the adsorption effect is significantly improved. In Example 1, the adsorption end time is 6 hours, while in Example 2, the adsorption end time is 4.5 hours. Compared with the traditional fixed parameter control preparation method, the dynamic optimization method can significantly shorten the adsorption time by 25%, thereby reducing energy consumption and production costs. At the same time, on the basis of significantly reducing the adsorption time and improving the adsorption efficiency, the modified adsorbent prepared by the dynamic optimization control strategy has a stronger adsorption capacity for metals in the polymer. The adsorption capacity of the modified adsorbent prepared by the traditional fixed parameter strategy for cobalt is reduced from 260 ppm to 59.7 ppm, and the adsorption capacity for zinc is reduced from 274 ppm to 38.9 ppm. The adsorption capacity after dynamic optimization can reduce both zinc and cobalt to 5. ppm or less, it can be seen that the dynamic optimization control strategy can achieve fine regulation of the activated carbon preparation process by accurately controlling the purge process of carbon dioxide. This precise control helps to optimize the pore structure and surface properties of activated carbon, thereby improving its adsorption performance; the process of preparing activated carbon by the dynamic optimization control strategy is relatively simple and easy to control, which helps to improve the removal rate of phenolic compounds, and can increase the removal rate of phenolic compounds from 69.7% to 99.9%, increasing the phenol removal rate by at least 43.3%. At the same time, the activated carbon prepared by the dynamic optimization control strategy has undergone fine regulation and optimization during the preparation process, so it has better stability. This stability makes it difficult for the activated carbon to undergo structural changes or performance degradation during use, thereby extending its service life. Therefore, the dynamic optimization control strategy for carbon dioxide purge pressure provided by the present application can significantly improve the adsorption efficiency and adsorption effect.

[0090] In the above embodiments, the description of each embodiment has its own emphasis. For the parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed system can be implemented in other ways. For example, the system embodiment described above is only schematic. For example, the unit division is only a logical function division, and there may be other division methods in actual implementation. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0091] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference to a figure in a claim should not be considered as limiting the claim to which it relates.

[0092] Finally, it should be noted that the above description has been given for the purpose of illustration and description. In addition, the above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. An industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols, wherein the double-modified adsorbent is prepared by the steps of pre-dispersion, six-membered nitrogen heterocyclic modification, high-temperature calcination at 300-800°C, and carbon dioxide gas modification, characterized in that: The industrial control system for preparing the double-modified adsorbent for purifying carbon dioxide-based polyols is used to optimize the control of the carbon dioxide gas modification process, including: A modification process monitoring module, used for collecting a time queue of a real-time solution pH value, a time queue of a real-time carbon dioxide purge pressure value, and a time queue of a real-time reaction temperature through a sensor matrix; A data transmission module, used to transmit the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value and the time queue of the real-time reaction temperature to the industrial intelligent control center through a 5G communication module; A reaction state causal time series feature extraction module is used to extract a time series correlation feature vector of carbon dioxide purge pressure and a time series correlation feature vector of solution pH under temperature modulation from a time series of the real-time solution pH value, a time series of the real-time carbon dioxide purge pressure value and a time series correlation feature vector of solution pH under temperature modulation in the industrial intelligent control center; A reaction state causal interaction analysis module, used for performing a reaction state causal interaction analysis on the carbon dioxide purge pressure time series correlation feature vector and the solution pH time series correlation feature vector under the temperature modulation to obtain a reaction state causal-effect time series collaborative representation vector; a purge pressure optimization control module, for generating an optimization instruction for the carbon dioxide purge pressure based on the reaction state cause-effect time series collaborative representation vector; Among them, the reaction state causal interaction analysis module includes: a conditional feature construction unit, which is used to extract the implicit correlation features between the carbon dioxide purge pressure time series correlation feature vector under the temperature modulation and the solution pH time series correlation feature vector as a conditional feature vector; a cross-domain interaction encoding unit, which is used to guide the carbon dioxide purge pressure time series correlation feature vector under the temperature modulation and the solution pH time series correlation feature vector to perform cross-domain interaction encoding based on the attention mechanism based on the conditional feature vector to obtain the reaction state cause-effect time series collaborative representation vector.

2. The industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to claim 1, characterized in that: The reaction state causal time series feature extraction module includes: A time series encoding unit, used for inputting the time queue of the real-time solution pH value, the time queue of the real-time carbon dioxide purge pressure value and the time queue of the real-time reaction temperature into a time series encoder based on a forward LSTM model to obtain the solution pH time series correlation feature vector, the carbon dioxide purge pressure time series correlation feature vector and the reaction temperature time series correlation feature vector; The joint modulation unit is used to input the carbon dioxide purge pressure time series correlation feature vector and the reaction temperature time series correlation feature vector into the reaction state factor joint modulation network to obtain the carbon dioxide purge pressure time series correlation feature vector under the temperature modulation.

3. The industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to claim 2, characterized in that: The joint modulation unit is used for: The Hadamard product between the carbon dioxide purge pressure time series correlation characteristic vector and the reaction temperature time series correlation characteristic vector is calculated to obtain the carbon dioxide purge pressure time series correlation characteristic vector under the temperature modulation.

4. The industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to claim 3, characterized in that: The conditional feature construction unit comprises: An implicit correlation feature extraction subunit is used to input the carbon dioxide purge pressure time series correlation feature vector under temperature modulation and the solution pH time series correlation feature vector into an implicit correlation feature capture network to obtain a reaction state cause-effect time series implicit correlation feature vector; The feature activation subunit is used to perform feature activation based on the Sigmoid function on the reaction state cause-effect time series implicit correlation feature vector to obtain the conditional feature vector.

5. The industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to claim 4, characterized in that: The implicit association feature extraction subunit is used to: The carbon dioxide purge pressure time series correlation feature vector under temperature modulation and the solution pH time series correlation feature vector are fused according to position points and passed through a neural network layer based on a tanh function to obtain the reaction state cause-effect time series implicit correlation feature vector.

6. The industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to claim 5, characterized in that: The cross-domain interaction encoding unit includes: a feature modulation subunit, configured to perform feature modulation on the carbon dioxide purge pressure time series association feature vector under temperature modulation and the solution pH time series association feature vector under temperature modulation based on the semantic contribution of the carbon dioxide purge pressure time series association feature vector under temperature modulation and the solution pH time series association feature vector to the conditional feature vector to obtain a carbon dioxide purge pressure time series association feature vector under temperature modulation after modulation and a solution pH time series association feature vector after modulation; The guided interaction subunit is used to perform inter-feature significant guided interaction encoding on the time series correlation feature vector of the carbon dioxide purge pressure under the post-modulation temperature modulation, the time series correlation feature vector of the post-modulation solution pH and the conditional feature vector to obtain the reaction state cause-effect time series collaborative representation vector.

7. The industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to claim 6, characterized in that: The characteristic modulation subunit is used for: Calculate a first semantic contribution factor of the carbon dioxide purge pressure time series associated feature vector under the temperature modulation relative to the conditional feature vector; Calculate a second semantic contribution factor of the solution pH time series associated feature vector to the conditional feature vector; The first semantic contribution factor and the second semantic contribution factor are normalized, and the normalized first semantic contribution factor and the second semantic contribution factor are used to weighted modulate the carbon dioxide purge pressure time series association feature vector under temperature modulation and the solution pH time series association feature vector to obtain the carbon dioxide purge pressure time series association feature vector under temperature modulation after modulation and the solution pH time series association feature vector after modulation.

8. The industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to claim 7, characterized in that: The guiding interaction subunit is used for: Using the time series associated feature vector of the carbon dioxide purge pressure under the post-modulation temperature modulation as the query vector, the time series associated feature vector of the post-modulation solution pH as the key vector and the conditional feature vector as the value vector, the time series associated feature vector of the carbon dioxide purge pressure under the post-modulation temperature modulation, the time series associated feature vector of the post-modulation solution pH and the conditional feature vector are subjected to significant guided interaction among features based on the converter structure to obtain the cause-effect time series collaborative representation vector of the reaction state.

9. The industrial control system for preparing a double-modified adsorbent for purifying carbon dioxide-based polyols according to claim 8, characterized in that: The purge pressure optimization control module is used to: The reaction state cause-effect time series synergistic representation vector is input into a decoder-based purge pressure dynamic optimizer to obtain the optimization instruction, and the optimization instruction is used to represent a recommended value of the carbon dioxide purge pressure at the next moment.

Images