Tubular reactor and temperature control method thereof

Abstract

The invention relates to the technical field of polycarbonate polyether polyol preparation, and particularly discloses a tubular reactor and a temperature control method thereof, the tubular reactor comprises a solid-liquid mixing unit, a gas-liquid mixing heat conduction unit and a reaction unit, the intelligent industrial control technology is adopted to carry out continuous monitoring and time sequence coding on the material density and the reaction temperature in the reaction pipeline so as to extract the time sequence change characteristics of the material density and the reaction temperature, and then the dynamic influence relation between the temperature and the material density is excavated by carrying out collaborative response analysis on the material density and the reaction temperature; therefore, the temperature of the heating jacket is intelligently adjusted, so that the dynamic intelligent optimal control of the temperature in the reaction process is realized. The tubular reactor can provide sufficient mixing conditions and reasonable temperature control for synthesis of the polycarbonate polyether polyol, overcomes the technical prejudice of a traditional reactor, and is beneficial to expanded industrial production of the polycarbonate polyether polyol.

Description

Technical Field

[0001] The present application relates to the technical field of polycarbonate polyether polyol preparation, and more specifically, to a tubular reactor and a temperature control method thereof. Background Art

[0002] Polycarbonate polyether polyols are a class of polyols with carbonate groups in the molecule and hydroxyl groups at the ends of the molecular chain. In the preparation process of this material, carbon dioxide is used as one of the raw materials. Due to its low cost, easy access, non-toxicity and non-flammability, it is used as an effective means of chemically fixing carbon dioxide.

[0003] The process of synthesizing polycarbonate polyether polyols is mainly synthesized by carbon dioxide and epoxy compounds through zinc-cobalt bimetallic catalysts. Chinese patent CN111378106A mentions that the activation energy of generating polymers is lower than that of generating cyclic carbonates, and the potential energy of cyclic carbonates is lower than that of polymers. Therefore, the reaction of generating cyclic carbonates is a thermodynamically favorable reaction, while the reaction of generating polycarbonates is a kinetically favorable reaction. If the temperature of the reaction process is too high, it will lead to low carbonate chain content, wide molecular weight distribution, and high ratio of by-product cyclic carbonates. Therefore, accurate reaction heat regulation is the key to improving the main product-polycarbonate polyether polyols. In addition, another key factor affecting the yield of the main product is the action of the DMC catalyst. Improper activation or instant activation will also cause explosions, which can easily cause high temperature and high pressure safety accidents.

[0004] Chinese patent CN115785435A divides the catalytic reaction of bimetallic complex (DMC) into an induction activation period and a segment growth period. In the induction activation period, the reaction is exothermic, and the temperature peak needs to be paid attention to, and the initial activation temperature is controlled to avoid overheating, thereby affecting the pressure resistance of the equipment and the activity of the catalyst. In the segment growth period, after DMC is fully activated, the active center combines with the initiator, the epoxy compound is inserted into the polymerization, and the initiator and the active structure of the DMC and epoxy compound undergo a replacement reaction. This stage requires external heat to promote chain growth. The CN106471042A patent defines the activation step of the DMC catalyst, including adding an alkylene oxide compound in the presence of CO2, and observing the temperature peak and / or pressure drop. The activation time starts from the addition of the alkylene oxide compound to the DMC catalyst until the hot spot weakens. The CN105408377A patent describes the reactor setting of the activation step, including the activation step of the DMC catalyst in a stirred vessel or a tubular reactor.

[0005] In order to adapt to the above-mentioned characteristics of the synthetic carbonate polyether polyol reaction and reduce the impact of heat fluctuations on the various components in the reaction, the existing technologies all require special catalyst design, process control and / or independent reactor design for the induced activation stage. The specific existing technologies are as follows: 1. The existing technology uses a special high-performance DMC catalyst, as described in Chinese patent CN115785435A, to achieve a one-step preparation of carbonate polyether polyols by optimizing catalyst performance, and can prepare polyether polyols at room temperature in one go. 2. In terms of process control, Chinese patents CN116710504A and CN107108878A disclose methods for preparing (poly) alcohol block copolymers suitable for different processes, involving catalyst preactivation and batch addition of reactants. 3. In order to solve the DMC catalyst activation problem, Chinese patent CN116874759A proposes to use an independent reactor to activate and maintain the activity of the catalyst, while CN106471042A controls the reaction temperature of the DMC induced activation period by connecting a stirred tank main reactor and a tubular reactor in series; Chinese patent CN103403060A introduces a method for producing polyether polycarbonate polyols without pre-activation, and controls the DMC induced activation period by connecting tubular reactors of different inner diameters in series.

[0006] However, from the perspective of reactor optimization, the technology of realizing a one-time centralized feeding of a reactor structure to prepare carbonate polyether polyols has not been described. At present, there are two types of reactors for industrial production of polycarbonate polyether polyols: kettle type and pipeline type. The surface area / volume ratio of the kettle reactor is low, the heat dissipation effect is poor, and it is not easy to handle high reaction heat. The pipeline reactor has controllable pressure and temperature during the reaction process, and has high monomer conversion efficiency, so it has attracted much attention in this field. Due to the small cross-sectional area and limited production capacity of the tubular reactor, it needs to be arrayed into a tube reactor for use. However, the general-purpose tube reactor has a large aspect ratio, which makes the materials unevenly mixed in the pipeline and the viscosity increases sharply, which is easy to produce explosion points, and it is impossible to dissipate heat and break up in time, which is not suitable for the synthesis of polycarbonate polyether polyols. In addition, in the synthesis process of polycarbonate polyether polyols, the reaction temperature is one of the key factors affecting product quality and yield. However, the temperature control of the existing tube reactor mainly depends on the PID control algorithm. Since the various process parameters in the reactor have nonlinear time-varying characteristics during the reaction process, the PID control algorithm is often difficult to achieve an ideal control effect when dealing with a time-varying reaction environment.

[0007] Therefore, in order to ensure gas-liquid mass transfer of reactants, uniform solid-liquid-gas mixing, and sufficient heating and heat dissipation, an optimized shell-and-tube reactor and its temperature control method are expected. Summary of the invention

[0008] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a tubular reactor and a temperature control method thereof, which includes a solid-liquid mixing unit, a gas-liquid mixing heat conduction unit and a reaction unit, and uses intelligent industrial control technology to continuously monitor and time-series encode the material density and reaction temperature in the reaction pipeline to extract the time-series change characteristics of the material density and the reaction temperature, and then through the coordinated response analysis of the two, the dynamic influence relationship between the temperature and the material density is excavated, so as to perform intelligent adjustment of the heating jacket temperature, thereby realizing the dynamic intelligent optimization control of the temperature of the reaction process. The tubular reactor can provide sufficient scale-up industrial production of mixed polyether polyols for the synthesis of polycarbonate polyether polyols.

[0009] Accordingly, according to one aspect of the present application, a tubular reactor is provided, which comprises: a feed port and a discharge port; a solid-liquid mixing unit, the solid-liquid mixing unit comprising a reaction pipe connected to the feed port, a static mixer for dispersing the catalyst and a circulation device for refluxing the liquid reactant; a gas-liquid mixing heat transfer unit, the gas-liquid mixing heat transfer unit comprising a dispersion module for regulating the droplet size and quantity of the liquid reactant, a gas-liquid contact heat transfer zone and the circulation device; a reaction unit, the reaction unit comprising the reaction pipe, a heating medium, a heating jacket and a thermal insulation layer, the heating jacket being arranged on the outside of the reaction pipe, the heating medium A space is arranged between the reaction pipes, and the insulation layer is arranged on the outside of the heating medium; a gas supply port, which is connected to the reaction pipe; a connecting pipe extending between the reaction pipe and the circulation device; a liquid distributor arranged on the top of the reaction pipe; wherein the connecting pipe, the liquid distributor and the reaction pipe constitute a circulation loop, a temperature sensor is arranged inside the reaction pipe, a density detector is arranged at the outlet of the reaction pipe, and the shell-and-tube reactor also includes a temperature controller, which is used to adjust the temperature of the heating jacket according to the material density and the internal temperature of the reactor to control the reaction rate.

[0010] In the above-mentioned shell-and-tube reactor, the temperature controller includes: a data acquisition module, which is used to acquire the time queue of material density and the time queue of reactor internal temperature; a material density time series feature extraction module, which is used to extract the material density time series feature from the material density time queue; a reactor internal temperature time series feature extraction module, which is used to extract the reactor internal temperature time series feature from the reactor internal temperature time queue; a time series collaborative analysis module, which is used to perform reaction temperature-material density time series collaborative analysis on the material density time series feature and the reactor internal temperature time series feature to obtain the reaction temperature-material density time series collaborative correlation feature; a heating jacket temperature control module, which is used to control the temperature of the heating jacket based on the reaction temperature-material density time series collaborative correlation feature.

[0011] In the above-mentioned tubular reactor, the material density time series feature extraction module is used to: input the time queue of the material density into a time series encoder based on the BiLSTM model to obtain the material density time series feature.

[0012] In the above-mentioned shell-and-tube reactor, the timing collaborative analysis module includes: a feature decomposition unit, which is used to perform feature decomposition on the material density timing characteristics to obtain a sequence of local material density timing characteristics; an interactive response analysis unit, which is used to perform local-global interactive response analysis on the sequence of local material density timing characteristics and the reactor internal temperature timing characteristics based on the feature interaction intensity between the reactor internal temperature timing characteristics and each material density local timing characteristics in the sequence of local material density timing characteristics to obtain the reaction temperature-material density timing collaborative correlation characteristics.

[0013] In the above-mentioned tubular reactor, the interactive response analysis unit includes: a nonlinear transformation subunit, which is used to perform nonlinear transformation on the reactor internal temperature time series characteristics to obtain the reactor internal temperature time series characteristics after nonlinear transformation; an interactive correlation feature calculation subunit, which is used to calculate the interactive correlation features between the reactor internal temperature time series characteristics after nonlinear transformation and each material density local time series feature in the sequence of material density local time series features to obtain a sequence of reaction temperature-material density time series interactive correlation features; a feature interaction strength calculation subunit, which is used to calculate each reaction temperature in the sequence of reaction temperature-material density time series interactive correlation features. The characteristic interaction intensity coefficient of the reaction temperature-material density time series interaction correlation feature is obtained by inputting the sequence of the reaction temperature-material density time series interaction intensity coefficient into the Sigmoid function to obtain the sequence of normalized reaction temperature-material density time series interaction intensity coefficient; the weighted fusion subunit is used to use the sequence of normalized reaction temperature-material density time series interaction intensity coefficient as the sequence of weights to perform weighted fusion on the sequence of reaction temperature-material density time series interaction correlation features to obtain the reaction temperature-material density time series synergistic correlation feature.

[0014] In the above-mentioned shell-and-tube reactor, the interactive correlation feature calculation subunit is used to: calculate the product of the transpose of the internal temperature time series feature of the reactor after the nonlinear transformation and each material density local time series feature in the sequence of material density local time series features to obtain the sequence of reaction temperature-material density time series interactive correlation features.

[0015] In the above-mentioned tubular reactor, the heating jacket temperature control module is used to: input the reaction temperature-material density time series collaborative correlation characteristics into a decoder-based temperature recommendation model to obtain a recommended temperature value for the heating jacket.

[0016] In the above-mentioned shell-and-tube reactor, the volume of the gas-liquid contact heat transfer zone is greater than or equal to 20% of the volume of the reaction pipe.

[0017] In the above-mentioned tubular reactor, the number of the reaction pipes is greater than or equal to 2, and the reaction pipes are connected in parallel and are independent of each other.

[0018] According to another aspect of the present application, a temperature control method for a shell-and-tube reactor is provided, which includes: obtaining a time queue of material density and a time queue of reactor internal temperature; extracting material density time series characteristics from the material density time queue; extracting reactor internal temperature time series characteristics from the reactor internal temperature time queue; performing reaction temperature-material density time series collaborative analysis on the material density time series characteristics and the reactor internal temperature time series characteristics to obtain reaction temperature-material density time series collaborative correlation characteristics; and regulating the temperature of the heating jacket based on the reaction temperature-material density time series collaborative correlation characteristics.

[0019] Compared with the prior art, the tubular reactor and its temperature control method provided by the present application include a solid-liquid mixing unit, a gas-liquid mixing heat conduction unit and a reaction unit, and use intelligent industrial control technology to continuously monitor and time-series encode the material density and reaction temperature in the reaction pipeline to extract the time-series change characteristics of the material density and reaction temperature, and then through the coordinated response analysis of the two, the dynamic influence relationship between temperature and material density is excavated, so as to perform intelligent adjustment of the heating jacket temperature, thereby realizing the dynamic intelligent optimization control of the temperature of the reaction process. The tubular reactor can provide sufficient mixing conditions and reasonable temperature control for the synthesis of polycarbonate polyether polyols, effectively overcome the technical bias of traditional reactors, and is conducive to the scaled-up industrial production of polycarbonate polyether polyols. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] 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.

[0021] Figure 1 It is a side view schematic diagram of a shell and tube reactor according to an embodiment of the present application.

[0022] Figure 2 It is a schematic top view of a shell-and-tube reactor according to an embodiment of the present application.

[0023] Figure 3 Schematic diagram of the structure of the reaction pipe in the tube reactor according to an embodiment of the present application.

[0024] Figure 4 4 is a block diagram of a temperature controller in a shell and tube reactor according to an embodiment of the present application.

[0025] Figure 5 Flow chart of a temperature control method for a shell-and-tube reactor according to an embodiment of the present application. DETAILED DESCRIPTION

[0026] 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.

[0027] Figure 1 It is a side view schematic diagram of a shell and tube reactor according to an embodiment of the present application. Figure 2 It is a schematic top view of a shell-and-tube reactor according to an embodiment of the present application. Figure 3 Schematic diagram of the structure of the reaction pipe in the tube reactor according to the embodiment of the present application. Figure 1 , Figure 2 and Figure 3As shown, the tubular reactor of the embodiment of the present application comprises: a feed inlet (104) and a discharge port (105); a solid-liquid mixing unit, the solid-liquid mixing unit comprising a reaction pipe (101) connected to the feed inlet (104), a static mixer (303) for dispersing the catalyst and a circulation device (103) for refluxing the liquid reactant; a gas-liquid mixing heat transfer unit, the gas-liquid mixing heat transfer unit comprising a dispersion module (301) for regulating the droplet size and number of the liquid reactant, a gas-liquid contact heat transfer zone (107) and the circulation device (103); a reaction unit, the reaction unit comprising the reaction pipe (101), a heating medium (202), a heating jacket (302), The invention relates to a heat-insulating layer (203), wherein the heating jacket (302) is arranged on the outside of the reaction pipe (101), the heating medium (202) is arranged in the space between the reaction pipes (101), and the heat-insulating layer (203) is arranged on the outside of the heating medium (202); a gas replenishing port (106), wherein the gas replenishing port (106) is connected to the reaction pipe (101); a connecting pipe (201) extending between the reaction pipe (101) and the circulation device (103); and a liquid distributor (102) arranged on the top of the reaction pipe; wherein the connecting pipe (201), the liquid distributor (102) and the reaction pipe (101) form a circulation loop.

[0028] In order to ensure sufficient mixing of the raw materials and prevent uneven reactions from causing excessively wide molecular weight distribution or implosion, a reaction preparation stage is required before the solid-liquid reactants are passed into the tubular reactor. In a 0.1-0.5MPa carbon dioxide environment, the raw epoxide is purified by the system and added to the raw material mixing device from the storage tank, and then the chain transfer agent and catalyst are added to fully mix with the raw epoxide. The premixing temperature range is 0-60°C, the pressure range is 0.1-2Mpa, and the premixing time is 1-4 hours. The raw material mixing device includes but is not limited to premixing kettles, venturi tubes and other devices. At this time, only physical mixing is performed in the raw material mixing device to ensure sufficient dissolution and mixing of the reactants, no polymerization reaction occurs, and no catalyst activation reaction occurs.

[0029] The premixed raw materials are fed into a reaction pipe (101) with a temperature of 40°C and a carbon dioxide pressure of 0.1-2 MPa from a feed port (104) through a metering pump. When a fixed amount of raw materials enters the reaction unit, the feeding is stopped, carbon dioxide is introduced through the air supply port (106) and the circulation device is turned on. The flow rate of the circulation device is based on a cycle of 1-20 minutes. The carbon dioxide in the raw materials is provided by the carbon dioxide pipeline. The pressure in the raw material mixing zone is adjusted to 2-6 MPa and maintained according to the pressure sensor data. The temperature of the reactants in the reaction pipe is controlled at 80°C by adjusting the temperature of the heating medium (202) in the reaction zone. The density sensor data is observed and reaches 1.1 g / cm3 The reaction is then considered complete. Figure 1 As shown, the green arrow indicates the direction of carbon dioxide replenishment, the black arrow indicates the direction of reaction raw material replenishment, and the blue arrow indicates the circulation direction of the reaction raw material and product mixture; the circulation device (103), the reaction pipeline (101), and the liquid distributor (102) constitute the circulation system of the tubular reactor, and the reaction raw materials and products move downward in the reaction pipeline (101), and finally gather through the circulation device (103) to circulate to the liquid distributor (102) for distribution and then return to each reaction pipeline to complete the cycle. The reaction pipeline (101) is as shown in FIG. Figure 3 As shown, the dispersion module (301) and the static mixer (303) included therein. The solid-liquid mixing unit allows the reaction raw materials and the catalyst to be evenly mixed during the circulation process, thereby avoiding the explosion caused by excessive catalyst content at a certain point in the reaction process and violent reaction. In addition, the gas-liquid mixing heat conduction unit allows the raw materials with low carbon dioxide content at the bottom of the reaction pipe (101) to return to the gas area at the top of the reaction pipe (101) to fully dissolve the carbon dioxide, thereby ensuring that during the synthesis process, all reaction unit areas will not have more polyether products generated due to low carbon dioxide content. Furthermore, the mixing of the reactants and the release of reaction heat by the solid-liquid mixing unit and the gas-liquid mixing heat conduction unit further reduce the distribution of molecular weight. The arrangement of the reaction pipe (101) inside the shell and tube reactor is shown in the top view Figure 2 As shown, but not limited to this arrangement.

[0030] In one embodiment, the volume of the gas-liquid contact heat transfer zone (107) is greater than or equal to 20% of the volume of the reaction pipe (101).

[0031] In one embodiment, the number of the reaction pipes (101) is two or more, which are connected in parallel, and the center distance between the nearest neighbor pipes is less than or equal to twice the pipe diameter. Preferably, the number of the reaction pipes (101) is 2 to 8.

[0032] In one embodiment, the reaction pipes (101) are independent of each other and can be closed and isolated by valves.

[0033] In one embodiment, the dispersion module (301) is located at a position 10 to 50 mm close to the top of the reaction pipe (101), and the dispersion module (301) is a porous structure.

[0034] In one embodiment, the ratio of the length L to the diameter d of the reaction pipe (101) is 10≤L﹕d≤40. The heating time of the industrial production of polycarbonate polyether polyol has particularity. In the polycarbonate polyether polyol tubular reactor, when the flow rate is constant, the heating time per unit volume of material is short, and the reaction conversion rate per unit time will be greatly reduced. While a longer reaction time is required, a longer heating time will cause the polyester chain segments to decompose, and the product will turn yellow, that is, the length-to-diameter ratio of the reaction pipe cannot be too large. At the same time, the reaction process continuously generates heat, which will increase the temperature of the reaction system. The reactor needs to release the heat generated in the material in time. When the flow rate is constant, the liquid needs to complete the cycle through the gas-liquid heat transfer zone within a specified time to complete heat release and gas-liquid mass transfer, wherein the cross-sectional area of ​​the gas-liquid heat transfer zone, that is, the pipe diameter, affects the heat transfer efficiency. The synthetic polyester polyether polyol reaction pipe (101) should have a special length-to-diameter ratio to ensure that the product has sufficient heating reaction time and meets the requirements of gas-liquid mass transfer and heat release at the same time.

[0035] In one embodiment, the circulation flow rate of the circulation device (103) is such that the material in the reactor completes one circulation in 1-20 minutes. Preferably, the circulation flow rate is such that the material completes one circulation in 10 minutes.

[0036] In one embodiment, a temperature sensor is provided inside the reaction pipe (101), a density detector is provided at the outlet, and the tubular reactor further comprises a temperature controller, which is used to adjust the temperature of the heating jacket according to the material density and the temperature inside the reactor to control the reaction rate. In the synthesis process of polycarbonate polyether polyol, as the copolymerization reaction of epoxy compound and carbon dioxide proceeds, the polymer chain grows and the molecular weight increases, thereby causing the product concentration to rise and the material density to increase. In this reaction process, the temperature inside the reaction pipe is an important factor affecting the reaction rate and the change of material density. Specifically, the increase in the temperature inside the pipe will accelerate the reaction rate, leading to the rapid growth of the polymer chain, and then the material density will increase rapidly; but too high a temperature may also cause side reactions to occur, thereby affecting the density and distribution of the product. Therefore, in order to ensure that the reaction is carried out within the optimal temperature range to avoid side reactions or reaction stagnation caused by excessively high or low temperatures, the shell-and-tube reactor described in the present application is designed with a temperature controller, which uses deep learning technology to perform time series encoding and coordinated response analysis on the time queue data of material density and reactor internal temperature, and adjusts the heating jacket temperature in real time based on the dynamic influence relationship between temperature and material density, thereby realizing dynamic intelligent optimization control of the temperature of the reaction process.

[0037] Figure 4 FIG. 1 is a block diagram of a temperature controller in a tube reactor according to an embodiment of the present application. Figure 4As shown, the temperature controller 400 includes a data acquisition module 410 for acquiring a time queue of material density and a time queue of reactor internal temperature; a material density time series feature extraction module 420 for extracting material density time series features from the material density time queue; a reactor internal temperature time series feature extraction module 430 for extracting reactor internal temperature time series features from the reactor internal temperature time queue; a time series collaborative analysis module 440 for performing a reaction temperature-material density time series collaborative analysis on the material density time series features and the reactor internal temperature time series features to obtain a reaction temperature-material density time series collaborative correlation feature; and a heating jacket temperature control module 450 for controlling the temperature of the heating jacket based on the reaction temperature-material density time series collaborative correlation feature.

[0038] After obtaining the time queue of material density and the time queue of reactor internal temperature through the data acquisition module 410, since both material density and reactor internal temperature are dynamic parameters that change with time, in order to fully understand the dynamic change characteristics during the reaction process, the present application further uses deep learning technology to perform time series encoding on the time queue of material density and the time queue of reactor internal temperature, and extracts the material density time series characteristics and the reactor internal temperature time series characteristics therefrom, so as to capture the key information such as the time change trend, periodic change and mutation point of material density and reactor internal temperature, thereby providing an important basis for subsequent temperature control.

[0039] In a specific example of the present application, the material density time series feature extraction module 420 is used to: input the time queue of the material density into the time series encoder based on the BiLSTM model to obtain the material density time series feature. It should be understood that the BiLSTM model has a bidirectional gating structure, which can simultaneously consider the forward and backward contextual information in the time series data, so as to more accurately capture the long-term time series dependency and dynamic change pattern of the material density. Correspondingly, in the reactor internal temperature time series feature extraction module 430, the BiLSTM model-based time series encoder is also used to extract the time series features of the reactor internal temperature time queue to ensure the consistency and accuracy of feature extraction.

[0040] Then, in order to further dig out the dynamic influence relationship between reaction temperature and material density, in the technical solution of the present application, the material density time series characteristics and the reactor internal temperature time series characteristics are subjected to reaction temperature-material density time series collaborative analysis to reveal the synergistic correlation change law between the two, so as to deeply understand the reaction mechanism and provide a more accurate adjustment basis for the control of reaction temperature. Specifically, in the time series collaborative analysis module 440, first, the material density time series characteristics are decomposed into a sequence of local material density time series characteristics to reveal different dimensions or aspects inside it and enhance the model's ability to capture local characteristics. At the same time, the reactor internal temperature time series characteristics are nonlinearly transformed to adjust their feature dimensions to facilitate subsequent feature interaction operations. Then, by calculating the interactive correlation characteristics between the reactor internal temperature time series characteristics after nonlinear transformation and the decomposed local material density characteristics, the interaction between the two is captured to form a sequence of reaction temperature-material density time series interactive correlation characteristics. Then, the feature interaction intensity coefficient of each reaction temperature-material density time series interactive correlation characteristic is further calculated to reveal the intensity and importance of feature interaction. Then, the feature interaction intensity coefficient sequence is normalized by the Sigmoid function and converted into attention weights, so as to perform weighted summation on the sequence of reaction temperature-material density time series interaction correlation features, thereby strengthening the most critical feature interactions and performing global integration of local interaction information to obtain the reaction temperature-material density time series synergistic correlation features.

[0041] In a specific example of the present application, the timing collaborative analysis module 440 includes: a feature decomposition unit, which is used to perform feature decomposition on the material density timing characteristics to obtain a sequence of local material density timing characteristics; an interactive response analysis unit, which is used to perform local-global interactive response analysis on the sequence of local material density timing characteristics and the reactor internal temperature timing characteristics based on the feature interaction intensity between the reactor internal temperature timing characteristics and each material density local timing characteristics in the sequence of local material density timing characteristics to obtain the reaction temperature-material density timing collaborative correlation characteristics.

[0042] Specifically, the interactive response analysis unit includes: a nonlinear transformation subunit, which is used to perform nonlinear transformation on the reactor internal temperature time series characteristics to obtain the reactor internal temperature time series characteristics after nonlinear transformation; an interactive correlation feature calculation subunit, which is used to calculate the interactive correlation features between the reactor internal temperature time series characteristics after nonlinear transformation and each material density local time series feature in the sequence of material density local time series features to obtain a sequence of reaction temperature-material density time series interactive correlation features; a feature interaction strength calculation subunit, which is used to calculate each reaction temperature-material density time series interactive correlation feature in the sequence of reaction temperature-material density time series interactive correlation features. The characteristic interaction intensity coefficient of the material density time series interactive correlation feature is used to obtain a sequence of reaction temperature-material density time series characteristic interaction intensity coefficients; a normalization subunit is used to input the sequence of reaction temperature-material density time series characteristic interaction intensity coefficients into the Sigmoid function to obtain a sequence of normalized reaction temperature-material density time series characteristic interaction intensity coefficients; a weighted fusion subunit is used to use the sequence of normalized reaction temperature-material density time series characteristic interaction intensity coefficients as a sequence of weights to perform weighted fusion on the sequence of reaction temperature-material density time series interactive correlation features to obtain the reaction temperature-material density time series synergistic correlation feature.

[0043] More specifically, the interactive correlation feature calculation subunit is used to calculate the product of the transpose of the internal temperature time series feature of the reactor after the nonlinear transformation and each material density local time series feature in the sequence of material density local time series features to obtain the sequence of reaction temperature-material density time series interactive correlation features.

[0044] More specifically, the feature interaction strength calculation subunit is used to: calculate the square of the difference between the kurtosis of the reaction temperature-material density time series interaction correlation feature and its feature mean, the feature variance of the reaction temperature-material density time series interaction correlation feature and the weighted sum of the bias terms, and then add the feature variance of the reaction temperature-material density time series interaction correlation feature and the bias term and divide the sum by the weighted sum to obtain the reaction temperature-material density time series feature interaction strength coefficient.

[0045] Accordingly, the time series collaborative analysis module 440 is used to process the material density time series characteristics and the reactor internal temperature time series characteristics by the following correlation interaction formula to obtain the reaction temperature-material density time series collaborative correlation characteristics, wherein the correlation interaction formula is:

[0046]

[0047]

[0048]

[0049]

[0050]

[0051]

[0052] in, is the time series characteristic of the temperature inside the reactor, is the material density time series characteristic, is the weight matrix, and are different bias terms, is the time series characteristic of the internal temperature of the reactor after nonlinear transformation, represents the eigendecomposition process, , , , The first, second, and third sequences of the local time series characteristics of the material density are represented respectively. and The local time series characteristics of material density, represents the transpose of the feature, represents the matrix multiplication operation, Indicates The reaction temperature-material density time series interaction characteristics, that is, the internal temperature time series characteristics of the reactor after the nonlinear transformation and the first The interactive correlation features between the local time series features of material density, Indicates The first reaction temperature-material density time series interaction correlation characteristic eigenvalues, Indicates the The characteristic interaction intensity coefficient of the reaction temperature-material density time series interaction correlation feature, , and Respectively represent the The kurtosis, feature mean and feature variance of the reaction temperature-material density time series interaction correlation feature, It means to calculate the expected value of a set. represents the sigmoid activation function, Indicates The normalized reaction temperature-material density time series characteristic interaction intensity coefficient, It represents the reaction temperature-material density time series synergistic correlation characteristics.

[0053] After the above processing, the reaction temperature-material density time series collaborative correlation feature contains the time series change information of the reaction temperature and material density and the collaborative correlation information between the two. Therefore, the reaction temperature-material density time series collaborative correlation feature can be used to regulate the temperature of the heating jacket to achieve real-time temperature control of the reaction process. In a specific example of the present application, the heating jacket temperature control module 450 is used to: input the reaction temperature-material density time series collaborative correlation feature into a decoder-based temperature recommendation model to obtain the recommended temperature value of the heating jacket. That is, the decoder model is used to process the reaction temperature-material density time series collaborative correlation feature, and the decoder can effectively predict the appropriate heating jacket temperature under the current reaction state by learning the collaborative correlation pattern between the reaction temperature and the material density. Then, the heating jacket is adjusted in real time based on the prediction result to ensure that the reaction process is carried out within the optimal temperature range.

[0054] In the technical solution of the present application, the material density timing characteristics and the reactor internal temperature timing characteristics represent the bidirectional long-range-short-range timing correlation characteristics of material density and the bidirectional long-range-short-range timing correlation characteristics of reactor internal temperature, respectively. In the process of reaction temperature-material density timing collaborative analysis of the material density timing characteristics and the reactor internal temperature timing characteristics, the difference in source time domain distribution of the material density and the reactor internal temperature will cause a fine-grained imbalance in the local-global timing response interaction results between the two, and the overall feature distribution of the reaction temperature-material density timing collaborative correlation characteristics finally obtained will have response interaction diversity, thereby affecting the consistency of decoding mapping iterations and reducing the accuracy of decoding results.

[0055] Preferably, inputting the reaction temperature-material density time series collaborative correlation feature into a temperature recommendation model based on a decoder to obtain a temperature recommendation value of the heating jacket comprises: Respectively calculating the first characteristic and value of all characteristic values ​​of the reaction temperature-material density time series synergistic correlation characteristic, the second characteristic and value of all characteristic values ​​of the material density time series characteristic, and the third characteristic and value of all characteristic values ​​of the reactor internal temperature time series characteristic; After dividing the first characteristic sum value by the second characteristic sum value and the third characteristic sum value respectively, the square root of the sum of the squares of the two quotients is calculated to obtain a reaction temperature-material density time series coordinated distribution probability value; Calculate the product of the first reaction temperature-material density time series synergistic characteristic value of the reaction temperature-material density time series synergistic correlation characteristic and the reaction temperature-material density time series synergistic distribution probability value to obtain a first reaction temperature-material density time series synergistic characteristic distribution probability product; Calculate the product of the second reaction temperature-material density time series synergy characteristic value of the reaction temperature-material density time series synergy correlation characteristic and the absolute value of the difference between one and the reaction temperature-material density time series synergy distribution probability value to obtain a second reaction temperature-material density time series synergy characteristic inverse distribution probability product; Calculate the quotient of the first reaction temperature-material density time series synergistic characteristic value divided by the absolute value of the difference between one and the reaction temperature-material density time series synergistic distribution probability value to obtain a first reaction temperature-material density time series synergistic characteristic inverse distribution probability quotient; Calculating the quotient of the second reaction temperature-material density time series synergistic characteristic value and the reaction temperature-material density time series synergistic distribution probability value to obtain a second reaction temperature-material density time series synergistic characteristic distribution probability quotient; After adding the first reaction temperature-material density time series synergistic characteristic distribution probability product and the second reaction temperature-material density time series synergistic characteristic inverse distribution probability product, subtract the difference between the first reaction temperature-material density time series synergistic characteristic inverse distribution probability quotient and the second reaction temperature-material density time series synergistic characteristic distribution probability quotient to obtain a reaction temperature-material density time series synergistic correction characteristic value; For each first reaction temperature-material density time series synergistic characteristic value and second reaction temperature-material density time series synergistic characteristic value of the reaction temperature-material density time series synergistic correlation characteristic, matrix multiplying a reaction temperature-material density time series synergistic correction characteristic matrix composed of the reaction temperature-material density time series synergistic correction characteristic values ​​by the reaction temperature-material density time series synergistic correlation characteristic to obtain a corrected reaction temperature-material density time series synergistic correlation characteristic; The corrected reaction temperature-material density time series synergistic correlation feature is input into a temperature recommendation model based on a decoder to obtain a temperature recommendation value of the heating jacket.

[0056] Among them, the reaction temperature-material density time series synergistic correlation feature The correction process is expressed as:

[0057]

[0058]

[0059] in, It indicates the reaction temperature-material density time series synergistic correlation characteristics, and respectively represent a first reaction temperature-material density time series synergistic characteristic value of the reaction temperature-material density time series synergistic correlation characteristic and a second reaction temperature-material density time series synergistic characteristic value of the reaction temperature-material density time series synergistic correlation characteristic, It represents the probability value of the time series coordinated distribution of reaction temperature and material density. It represents the reaction temperature-material density time series coordinated correction characteristic value, represents a reaction temperature-material density time series collaborative correction characteristic matrix composed of the reaction temperature-material density time series collaborative correction characteristic values, represents matrix multiplication, It represents the corrected reaction temperature-material density time series synergistic correlation characteristics.

[0060] and,

[0061] in, It indicates the reaction temperature-material density time series synergistic correlation characteristics, a first reaction temperature-material density time series synergistic characteristic value representing the reaction temperature-material density time series synergistic correlation characteristic, is the material density time series characteristic, The first eigenvalues, and is the time series characteristic of the temperature inside the reactor, The first eigenvalues, It represents the probability value of the reaction temperature-material density time series coordinated distribution.

[0062] Therefore, the present application targets the regional-boundary integral relationship of the high-dimensional feature manifold of the reaction temperature-material density time series synergistic correlation feature, and defines the spatial distribution probability representation of the feature set of the reaction temperature-material density time series synergistic correlation feature as a manifold constraint boundary relative to the material density time series feature and the reactor internal temperature time series feature to approximate the high-dimensional manifold of the feature set of the reaction temperature-material density time series synergistic correlation feature, which is composed of eigenvalue pairs, so as to avoid the ambiguity in the mapping of the diversified feature representation of the reaction temperature-material density time series synergistic correlation feature to the local manifold representation in the high-dimensional feature decoding regression convergence space, improve the execution iteration consistency of each local feature distribution in the mapping task, improve the feature decoding regression convergence effect, and improve the image quality of the temperature recommendation value of the heating jacket obtained by the temperature recommendation model based on the decoder.

[0063] The product after the reaction contains polycarbonate polyol, polyether polyol, cyclic carbonate and unreacted raw material epoxide. Next, the liquid product is passed through a catalyst filter to filter and recover the catalyst, and the filtrate flows into a distillation device, which is a falling film tower and a scraper evaporator. The filtrate first passes through the falling film tower to distill the epoxide into a buffer tank for recovery and application, and then flows into a scraper evaporator or a distillation tower to separate the main product polycarbonate polyether polyol and the by-product cyclic carbonate. The separation rate of polyol and cyclic carbonate is greater than 98%. After the two are completely separated, they are canned separately for standby use.

[0064] In summary, the tubular reactor according to the embodiment of the present application is explained, which includes a solid-liquid mixing unit, a gas-liquid mixing heat conduction unit and a reaction unit, and uses intelligent industrial control technology to continuously monitor and time-series encode the material density and reaction temperature in the reaction pipeline to extract the time-series change characteristics of the material density and reaction temperature, and then through the coordinated response analysis of the two, the dynamic influence relationship between temperature and material density is excavated, so as to perform intelligent adjustment of the heating jacket temperature, thereby realizing the dynamic intelligent optimization control of the temperature of the reaction process. The tubular reactor can provide sufficient mixing conditions and reasonable temperature control for the synthesis of polycarbonate polyether polyols, overcome the technical bias of traditional reactors, and is conducive to the scaled-up industrial production of polycarbonate polyether polyols.

[0065] Figure 5 Flow chart of the temperature control method of the tube reactor according to the embodiment of the present application. Figure 5 As shown, the temperature control method of the shell and tube reactor according to the embodiment of the present application includes the steps of: S1, obtaining the time queue of material density and the time queue of the temperature inside the reactor; S2, extracting the material density time series characteristics from the time queue of the material density; S3, extracting the reactor internal temperature time series characteristics from the time queue of the reactor internal temperature; S4, performing reaction temperature-material density time series collaborative analysis on the material density time series characteristics and the reactor internal temperature time series characteristics to obtain reaction temperature-material density time series collaborative correlation characteristics; S5, regulating the temperature of the heating jacket based on the reaction temperature-material density time series collaborative correlation characteristics.

[0066] Here, those skilled in the art will appreciate that the specific operations of each step in the temperature control method of the above-mentioned tubular reactor have been described in detail above. Figures 1 to 4 The description of the shell and tube reactor has been described in detail, and therefore, its repeated description will be omitted.

[0067] The tubular reactor described in this application is further described below in conjunction with specific embodiments.

[0068] Embodiment 1: 8L propylene oxide, 4g DMC catalyst and 224g chain transfer agent were added to the premixing kettle respectively, and 0.4Mpa carbon dioxide was introduced and mixed and stirred at a temperature of 40°C. After stirring for 1h, the mixture was pressed into an 8L*2 reactor array, and the diameter of a single reaction pipe was 100mm and the length of the reaction pipe was 1000mm. After all the mixture was pressed in, carbon dioxide was introduced into the reaction pipe to 3Mpa and the pressure was maintained. The circulation device was turned on and the flow rate was controlled at 600g / min. After reacting at a temperature of 80°C for 3h, the density was uniformly increased to 1.12g / cm 3 And stable, at this time the reaction is over, lower the temperature and take out the product. After the catalyst is removed by the catalyst filter device, a scraper evaporator is used to separate the main and by-products. The DMC catalyst described in this embodiment is a water-soluble metal salt of zinc and cobalt that reacts in a water-soluble solvent to obtain a zinc-cobalt double metal cyanide complex catalyst; the specific preparation is as follows: weigh a certain mass of cobalt salt and zinc salt, specifically, sodium thiocyanate cobaltate and zinc bromide, the molar ratio is 1:4, and the mixture is dissolved in an aqueous solvent and stirred continuously. The aqueous solvent includes water and tert-butyl alcohol, and the total mass of the metal salt (i.e., the cobalt salt and the zinc salt) is 1:5 to the mass ratio of the aqueous solvent. Add inorganic acid and organic acid, the inorganic acid is dilute hydrochloric acid, the pH value is 2, the organic acid is glutaric acid, the molar ratio of inorganic acid to organic acid is 5:1, the total molar number of metal salts and the molar number of acid are 4:1, and stirred at a temperature of 10-100°C for several hours, and precipitation continues to form. The above turbid liquid is filtered and dried to obtain a filter cake. The filter cake is re-slurried and washed with an aqueous solvent at a temperature of 10-100°C, specifically, at a temperature of 100°C, washed for 3 minutes, stirred for several hours, and then filtered and dried to obtain a filter cake. The slurrying, washing, and drying steps are repeated for multiple times at a temperature of 10-100°C until the pH value of the liquid in the system is 6-7, specifically, at a temperature of 60°C, for 6 minutes each time. The solid product is further dried under vacuum conditions at 80-100°C to obtain the final catalyst, which is processed into powder particles by mechanical grinding under anhydrous drying conditions before use.

[0069] The chain transfer agent described in this embodiment is selected from any one or more of ethylene glycol, diethylene glycol, 1,2-propylene glycol, 1,3-propylene glycol, 1,4-butanediol, 1,5-pentanediol, 1,6-hexanediol, 1,4-cyclohexanedimethanol, neopentyl glycol, glycerol, trimethylolpropane, trimethylolethane, 1,2,4-butanetriol, 1,2,6-hexanetriol, pentaerythritol, dipentaerythritol, succinic acid, glutaric acid, adipic acid, pimelic acid, suberic acid, azelaic acid, sebacic acid, dodecanedioic acid, terephthalic acid, isophthalic acid, phthalic acid, trimesic acid, pyromellitic acid, catechol, resorcinol, and hydroquinone.

[0070] The epoxide described in this embodiment is selected from at least one of ethylene oxide, propylene oxide, butylene oxide, and epichlorohydrin. Specifically, the epoxide is propylene oxide (PO).

[0071] The purified product is sampled and analyzed. Specifically, the polymerization reaction products (polycarbonate-polyether polyols and cyclic propylene carbonate and unreacted epoxide) are sampled and collected in a container, and the polymerization reaction product logistics samples are characterized by nuclear magnetic resonance spectroscopy to calculate the ratio of polymer to cyclic small molecules in the crude product. After the polymer is purified, nuclear magnetic resonance spectroscopy is performed again to calculate the ratio of polycarbonate chain segments to polyether chain segments on the polymer main chain (i.e., ester-ether ratio). There are only two structures, polycarbonate chain segments and polyether chain segments, on the polymer main chain, and the sum of the two percentages is 100%.

[0072] With the help of 1 The amount of carbon dioxide introduced in the obtained polycarbonate polyether polyol (carbonate segment content) and the ratio of propylene carbonate (cyclic carbonate) to polycarbonate polyether polyol were determined by H-NMR (Bruker, DPX400, 400 MHz; pulse program zg30, waiting time d1:10 s, 64 scans). In each case, the sample was dissolved in deuterated chloroform. 1 The relevant resonances in H-NMR (based on TMS = 0 ppm) are as follows: Among them, 5.0ppm and 4.2ppm belong to the proton peaks of the methine and methylene on the polycarbonate chain, 4.9ppm, 4.5ppm and 4.1ppm belong to the proton peaks of the methine and methylene in the five-membered ring carbonate, and 3.5-3.8ppm belong to the proton peak of the ether chain. The capital letter A plus a number subscript represents the integrated area of ​​the peak at a certain ppm on the nuclear magnetic resonance hydrogen spectrum. A is the abbreviation of Area in English, for example, A 5.0 represents the integrated area of ​​the peak at 5.0 ppm. 1 The integrated area of ​​the HNMR spectrum and its related proton peaks, the ratio of carbonate chains in the copolymerization reaction (molar ratio) (F CO2 ) and the mass fraction of cyclic carbonate content (W PC ), the amount of carbon dioxide embedded (M CO2 ) is calculated as follows: in, F CO2 =(A 5.0 +A 4.2 -2×A 4.6 ) / [(A 5.0 +A 4.2 -2×A 4.6 )+A 3.5 ]×100%; WPC =102×A 1.5 / [102×(A 5.0 +A 4.2 -2×A 4.6 +A1.5)+58×A 3.5 ]×100%; M CO2 =44×F CO2 / [102×F CO2 +58×(1-F CO2 )]×100%; The coefficient 44 is the molar mass of carbon dioxide, the coefficient 58 is the molar mass of PO, and the coefficient 102 is the sum of the molar masses of carbon dioxide (molar mass 44 g / mol) and the molar masses of PO (molar mass 58 g / mol).

[0073] The product conversion rate is calculated as follows: n = m × (1-M CO2 ) / (Mm×M CO2 ); wherein m is the mass of the product after removing propylene oxide, and M is the total sample containing propylene oxide.

[0074] The polymer number average molecular weight (Mn) and polymer molecular weight polydispersity index (PDI) were determined by gel permeation chromatography (GPC).

[0075] The final product was tested for colorimetry. 5g of the test object was placed in a cuvette and the cuvette was placed in a dark environment to reduce the influence of the light source on the test and cause errors. The NR10QC colorimeter was used for the test, and the test data included three basic indicators: L, a, and b.

[0076] Where L represents color brightness, L=0 indicates black, and L=100 indicates white.

[0077] Where a represents a position between red / magenta and green, with negative values ​​of a indicating green and positive values ​​of a indicating magenta.

[0078] Where b represents the position between yellow and blue, a negative value of b indicates blue, and a positive value of b indicates yellow.

[0079] The smaller the absolute values ​​of a and b, the better; the smaller they are, the closer they are to colorless.

[0080] The data are shown in Table 1: Table 1

[0081] From the data of Example 1, it can be seen that the present device can fully react the raw materials at 80°C, and the prepared polycarbonate polyether polyol has a narrower molecular weight distribution of 1.1; a higher carbonate segment ratio of more than 55%, that is, a high carbon dioxide fixation rate; and a very short reaction time, only 3 hours to reach a given conversion rate, and relatively less energy consumption.

[0082] Embodiment 2: Compared with Example 1, other parameters and reaction materials in Example 2 have not been changed. The only change is that the temperature control method described above is adopted in Example 2. The experiment is shown in Table 2: Table 2

[0083] From the data of Example 2, it can be known that the device can fully react the raw materials in the temperature control method described above. It should be noted that the reaction end time in Example 1 is 3 hours, while the reaction end time in Example 2 is 2.5 hours. After testing, the reaction products have little difference in conversion rate, ester-ether ratio, molecular weight and molecular weight distribution between the two, and the quality is equivalent, but the product in Example 2 is optimized by more than 30% in terms of chromaticity a and b values, and tends to be colorless. Therefore, the temperature control method provided by this application can improve the reaction efficiency and ensure the quality of the reaction product.

[0084] The above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A tubular reactor, characterized in that: include: Feed inlet and discharge outlet; A solid-liquid mixing unit, the solid-liquid mixing unit comprising a reaction pipeline connected to the feed port, a static mixer for dispersing the catalyst, and a circulation device for refluxing the liquid reactant; A gas-liquid mixed heat transfer unit, the gas-liquid mixed heat transfer unit comprising a dispersion module for regulating the droplet size and quantity of the liquid reactant, a gas-liquid contact heat transfer zone and the circulation device; A reaction unit, wherein the reaction unit comprises the reaction pipe, a heating medium, a heating jacket, and a thermal insulation layer, wherein the heating jacket is arranged on the outside of the reaction pipe, the heating medium is arranged in the space between the reaction pipes, and the thermal insulation layer is arranged on the outside of the heating medium; A gas supply port, the gas supply port is connected to the reaction pipeline; A connecting pipe extending between the reaction pipe and the circulation device; A liquid distributor disposed at the top of the reaction pipe; Wherein, the connecting pipe, the liquid distributor and the reaction pipe constitute a circulation loop, a temperature sensor is arranged inside the reaction pipe, a density detector is arranged at the outlet of the reaction pipe, and the shell-and-tube reactor also includes a temperature controller, which is used to adjust the temperature of the heating jacket according to the material density and the internal temperature of the reactor to control the reaction rate.

2. The tubular reactor according to claim 1, characterized in that: The temperature controller comprises: A data acquisition module, used to acquire the time queue of material density and the time queue of the internal temperature of the reactor; A material density time series feature extraction module, used to extract material density time series features from the time queue of the material density; A reactor internal temperature time series feature extraction module, used to extract the reactor internal temperature time series feature from the reactor internal temperature time queue; A time series collaborative analysis module, used for performing a reaction temperature-material density time series collaborative analysis on the material density time series characteristics and the reactor internal temperature time series characteristics to obtain a reaction temperature-material density time series collaborative correlation characteristic; The heating jacket temperature control module is used to control the temperature of the heating jacket based on the reaction temperature-material density time series collaborative correlation characteristics.

3. The tubular reactor according to claim 2, characterized in that: The material density time series feature extraction module is used to: The time queue of the material density is input into a time series encoder based on a BiLSTM model to obtain the time series characteristics of the material density.

4. The tubular reactor according to claim 3, characterized in that: The timing collaborative analysis module includes: A feature decomposition unit, used for performing feature decomposition on the material density time series feature to obtain a sequence of local time series features of the material density; An interactive response analysis unit is used to perform a local-global interactive response analysis on the sequence of the local time series characteristics of the material density and the internal temperature time series characteristics of the reactor based on the characteristic interaction strength between the local time series characteristics of the material density in the sequence of the local time series characteristics of the material density to obtain the reaction temperature-material density time series synergistic correlation characteristics.

5. The tubular reactor according to claim 4, characterized in that: The interactive response analysis unit comprises: A nonlinear transformation subunit, used for performing a nonlinear transformation on the reactor internal temperature time series characteristic to obtain a nonlinearly transformed reactor internal temperature time series characteristic; An interactive correlation feature calculation subunit, used for calculating the interactive correlation features between the reactor internal temperature time series feature after the nonlinear transformation and each material density local time series feature in the sequence of material density local time series features to obtain a sequence of reaction temperature-material density time series interactive correlation features; A feature interaction strength calculation subunit, used for calculating the feature interaction strength coefficient of each reaction temperature-material density time series interaction correlation feature in the sequence of reaction temperature-material density time series interaction correlation features to obtain a sequence of reaction temperature-material density time series feature interaction strength coefficients; A normalization subunit, used for inputting the sequence of interaction intensity coefficients of reaction temperature-material density time series characteristics into a Sigmoid function to obtain a sequence of normalized interaction intensity coefficients of reaction temperature-material density time series characteristics; The weighted fusion subunit is used to use the sequence of interaction intensity coefficients of the normalized reaction temperature-material density time series characteristics as the sequence of weights, and to perform weighted fusion on the sequence of reaction temperature-material density time series interactive correlation characteristics to obtain the reaction temperature-material density time series synergistic correlation characteristics.

6. The tubular reactor according to claim 5, characterized in that: The interactive correlation feature calculation subunit is used to: The product of the transpose of the reactor internal temperature time series characteristic after the nonlinear transformation and each material density local time series characteristic in the sequence of the material density local time series characteristics is calculated to obtain the sequence of the reaction temperature-material density time series interaction correlation characteristics.

7. The tubular reactor according to claim 6, characterized in that: The heating jacket temperature control module is used for: The reaction temperature-material density time series synergistic correlation feature is input into a temperature recommendation model based on a decoder to obtain a recommended temperature value for the heating jacket.

8. The tubular reactor according to claim 1, characterized in that: The volume of the gas-liquid contact heat transfer zone is greater than or equal to 20% of the volume of the reaction pipe.

9. The tubular reactor according to claim 1, characterized in that: The number of the reaction pipes is greater than or equal to 2, and the reaction pipes are connected in parallel and are independent of each other.

10. A temperature control method for a tubular reactor according to any one of claims 2 to 7, characterized in that: include: Obtaining the time sequence of material density and the time sequence of reactor internal temperature; Extracting material density time series characteristics from the time sequence of material density; Extracting the time series characteristics of the internal temperature of the reactor from the time series of the internal temperature of the reactor; Performing a reaction temperature-material density time series synergistic analysis on the material density time series characteristics and the reactor internal temperature time series characteristics to obtain a reaction temperature-material density time series synergistic correlation characteristic; Based on the reaction temperature-material density time series synergistic correlation characteristics, the temperature of the heating jacket is regulated.

Images