Process dynamic control method, device, system and storage medium of dividing wall column

Abstract

The application provides a process dynamic control method, device and system of a dividing wall column and a storage medium, and applies a chemical engineering rectification process. The method comprises the following steps: obtaining first parameter information of the dividing wall column, including a first heat load of a reboiler, a second heat load of a condenser, a temperature difference of a feed section and a temperature difference of a side line production section; performing joint processing on the first parameter information, an adaptive neuro-fuzzy inference system and an MPC model to obtain control information, wherein the control information comprises liquid phase flow allocated to the feed section by a common rectification section, a first production flow of the side line production section and reflux flow of a column top; and controlling the dividing wall column according to the control information. By adopting the joint control of the adaptive neuro-fuzzy inference system and the MPC model and taking the temperature difference as a control point, the control problems in the rectification process, such as difficult real-time measurement of components and strong coupling of variables, are solved, the dynamic control performance is improved, and disturbances of feed flow and feed composition are effectively resisted.

Description

Technical Field

[0001] This application relates to the fields of chemical distillation processes and automation control, and in particular to a method, apparatus, system and storage medium for dynamic process control of a partitioned column. Background Technology

[0002] Currently, many types of trichlorosilane distillation technologies have been developed for practical application, such as traditional two-tower processes, three-tower processes, differential pressure thermal coupling distillation, and multi-effect distillation. The application of these distillation technologies provides a strong guarantee for the preparation of high-purity trichlorosilane. However, the only drawback is that the energy consumption required in the process is huge, and the initial equipment investment is large, which makes the cost of producing high-purity trichlorosilane remain high.

[0003] The introduction of a baffle column reduces the energy consumption of separating chlorosilane systems. For chlorosilanes, a three-component mixture, the boiling points of the various components are not significantly different, making this system suitable for baffle column distillation processes. Because the baffle column has the most ideal thermodynamic system structure, it requires less heat and condensation consumption in the purification of trichlorosilane compared to traditional processes, significantly reducing energy consumption and thus lowering production costs.

[0004] The complex interactions between variables in the adjacent column and the difficulties in process control when feed disturbances occur in the adjacent column hinder the widespread application of adjacent column structures. Therefore, how to restore product purity to its original set value within a short time and with a small overshoot after a feed disturbance has become an important research direction for those skilled in the art. Summary of the Invention

[0005] The technical objective of this application is to provide a method, apparatus, system, and storage medium for dynamic process control of a partition tower, in order to solve the problem that it is difficult to restore the original set value after a feed disturbance occurs in the current partition tower.

[0006] To address the aforementioned technical problems, embodiments of this application provide a method for dynamic process control of a partitioned tower, comprising:

[0007] Obtain the first parameter information of the adjacent tower, which includes: the first heat load of the reboiler, the second heat load of the condenser, the temperature difference of the feed section, and the temperature difference of the side stream outlet section;

[0008] Based on the first parameter information, and through joint processing of the adaptive neural fuzzy inference system and the model predictive control (MPC) model, control information is obtained. The control information includes: the liquid phase flow rate allocated from the common rectification section to the feed section, the first outflow flow rate of the side stream outflow section, and the reflux flow rate at the top of the column.

[0009] The adjacent tower is controlled according to the control information.

[0010] Specifically, as described above, the step of obtaining control information based on the first parameter information, the adaptive neural fuzzy inference system, and the MPC model through joint processing includes:

[0011] The liquid flow rate is obtained by predicting the temperature difference in the feed section based on the MPC model.

[0012] Based on the MPC model, the temperature difference of the side-line extraction section is predicted to obtain the first extraction flow rate.

[0013] The adaptive neural fuzzy inference system performs fuzzy inference on the first heat load and the second heat load to obtain the set temperature of the common rectification section, and combines the set temperature, the first temperature of the common rectification section and the second outflow rate at the top of the column to obtain the reflux flow rate.

[0014] Furthermore, as described above, obtaining the reflux flow rate by combining the set temperature, the first temperature of the common rectification section, and the second drawdown flow rate at the top of the column includes:

[0015] The set reflux ratio is determined based on the set temperature and the first temperature;

[0016] The reflux flow rate is determined based on the set reflux ratio and the second extraction flow rate.

[0017] Specifically, as described above, obtaining the first parameter information includes:

[0018] The second temperature of the feed section sensitive plate and the third temperature of the side-line extraction section sensitive plate are obtained;

[0019] Based on a preset reference temperature, the difference between the second temperature and the corresponding reference temperature is determined as the temperature difference of the feeding section, and the difference between the third temperature and the corresponding reference temperature is determined as the temperature difference of the side-line sampling section. The reference temperature is determined based on the average absolute deviation of the temperature change of the sensitive plate in a closed loop with component disturbance.

[0020] Specifically, the method described above also includes:

[0021] Obtain historical parameter samples of the adjacent tower, the historical parameter samples including: the first heat load and the second heat load corresponding to each detection point within a preset time period;

[0022] The first heat load and the second heat load are normalized.

[0023] Based on the normalized first heat load and second heat load, determine the membership function and number of fuzzy rules of the preset adaptive neural fuzzy inference system;

[0024] The preset adaptive neural fuzzy inference system is trained based on the normalized first thermal load and the second thermal load to obtain the adaptive neural fuzzy inference system.

[0025] Specifically, the method described above also includes:

[0026] Obtain the feed flow rate of the feed section of the partition tower and the fourth temperature of the common stripping section;

[0027] The heat load control information of the reboiler is determined based on the feed flow rate of the feed section and the fourth temperature.

[0028] The first heat load of the reboiler is controlled according to the heat load control information.

[0029] Another embodiment of this application provides a process dynamic control device for a partition tower, comprising:

[0030] The first processing module is used to acquire the first parameter information of the adjacent tower, which includes: the first heat load of the reboiler, the second heat load of the condenser, the temperature difference of the feed section, and the temperature difference of the side stream outlet section.

[0031] The second processing module is used to perform joint processing based on the first parameter information, the adaptive neural fuzzy inference system, and the model predictive control (MPC) model to obtain control information. The control information includes: the liquid phase flow rate allocated from the common rectification section to the feed section, the first outflow flow rate of the side stream outflow section, and the reflux flow rate at the top of the column.

[0032] The third processing module is used to control the adjacent tower according to the control information.

[0033] Another embodiment of this application provides a process dynamic control system for a partitioned column, including: a partitioned column, a reboiler, a condenser, a reflux pump, a reflux tank, a reflux control valve, a side-stream sampling pump, a side-stream sampling control valve, a liquid phase distribution control valve, a side-stream sampling section temperature difference controller, a feed section temperature difference controller, an adaptive neuro-fuzzy inference system controller, and a model prediction controller.

[0034] The partition tower is divided into a common rectification section near the top of the tower, a common stripping section near the bottom of the tower, a feed section with a feed inlet in the middle, and a side-stream extraction section with a side-stream extraction outlet in the middle by a partition plate set in the middle.

[0035] The reboiler is connected to the common stripping section;

[0036] The condenser is connected to the common rectification section via the reflux tank, the reflux pump, and the reflux control valve;

[0037] The sideline extraction control valve and the pipeline containing the sideline extraction pump are connected to the sideline extraction section.

[0038] The temperature difference controller for the side-line extraction section is connected to the side-line extraction section;

[0039] The temperature difference controller for the feeding section is connected to the feeding section;

[0040] The liquid phase distribution control valve is connected to the common distillation section and the feed section via pipelines;

[0041] The adaptive neural fuzzy inference system controller is connected to the reboiler, the condenser and the reflux control valve. It is used to obtain the heat load of the reboiler and the condenser, perform fuzzy inference, and control the reflux flow rate at the top of the column by controlling the reflux control valve.

[0042] The model prediction controller is connected to the side-line extraction section temperature difference controller and the side-line extraction control valve. It is used to perform model prediction based on the side-line extraction section temperature difference obtained by the side-line extraction section temperature difference controller, and to control the first extraction flow rate of the side-line extraction section by controlling the side-line extraction control valve.

[0043] The model prediction controller is also connected to the feed section temperature difference controller and the liquid phase distribution control valve, and is used to perform model prediction based on the feed section temperature difference obtained by the feed section temperature difference controller, and to control the liquid phase flow rate distributed from the common rectification section to the feed section by controlling the liquid phase distribution valve.

[0044] Specifically, the system described above also includes: a common rectification section temperature controller and a reflux ratio controller;

[0045] The common distillation section temperature controller is connected to the common distillation section, and is also connected to the adaptive neural fuzzy inference system controller and the reflux ratio controller. It is used to determine the set reflux ratio of the reflux ratio controller based on the first temperature of the common distillation section and the temperature set value output by the adaptive neural fuzzy inference system.

[0046] The reflux ratio controller is also connected to the reflux control valve and is used to determine the reflux flow rate based on the set reflux ratio and the second outflow rate at the top of the tower, and to control the flow rate through the reflux control valve based on the reflux flow rate.

[0047] Another embodiment of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the process dynamic control method for the partition tower as described above.

[0048] Compared with the prior art, the process dynamic control method, apparatus, system and storage medium for a partition tower provided in this application have at least the following beneficial effects:

[0049] This embodiment solves the control problems of difficult real-time component measurement and strong variable coupling in the distillation process by adopting an adaptive neuro-fuzzy inference system and MPC model joint control, thus improving dynamic control performance. Furthermore, the breakthrough of using temperature difference as a control point better addresses control issues such as variable coupling, effectively resisting disturbances in feed flow rate and feed composition. This satisfies product specifications while resulting in control results with smaller overshoot, transition time, and oscillation. Simultaneously, it reduces the frequency of manual operation, effectively ensuring the safety of the production process. Attached Figure Description

[0050] Figure 1 This is one of the process flow diagrams for the process dynamic control method of the partition tower in this application;

[0051] Figure 2 This is the second flow diagram of the process dynamic control method for the partition tower of this application;

[0052] Figure 3 This is the third flow diagram of the process dynamic control method for the adjacent tower of this application;

[0053] Figure 4 This is the fourth flow diagram of the process dynamic control method for the adjacent tower of this application;

[0054] Figure 5 This is the fifth flow diagram of the process dynamic control method for the partition tower of this application;

[0055] Figure 6 This is the sixth flow diagram of the process dynamic control method for the adjacent tower of this application;

[0056] Figure 7 This is a schematic diagram of the process dynamic control device for the partition tower in this application;

[0057] Figure 8 This is a schematic diagram of the process dynamic control system of the adjacent tower in this application. Detailed Implementation

[0058] To make the technical problems, technical solutions, and advantages of this application clearer, a detailed description will be provided below in conjunction with the accompanying drawings and specific embodiments. In the following description, specific details such as particular configurations and components are provided merely to aid in a comprehensive understanding of the embodiments of this application. Therefore, those skilled in the art should understand that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this application. Furthermore, for clarity and brevity, descriptions of known functions and structures have been omitted.

[0059] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments.

[0060] In the various embodiments of this application, it should be understood that the sequence number of each process described below does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0061] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0062] In the embodiments provided in this application, it should be understood that "B corresponding to A" means that B is associated with A, and B can be determined based on A. However, it should also be understood that determining B based on A does not mean determining B solely based on A; B can also be determined based on A and / or other information.

[0063] See Figure 1 One embodiment of this application provides a method for dynamic process control of a partition tower, comprising:

[0064] Step S101: Obtain the first parameter information of the adjacent tower. The first parameter information includes: the first heat load of the reboiler, the second heat load of the condenser, the temperature difference of the feed section, and the temperature difference of the side stream outlet section.

[0065] Step S102: Based on the first parameter information, and the adaptive neural fuzzy inference system and MPC model, control information is obtained through joint processing. The control information includes: the liquid phase flow rate allocated from the common rectification section to the feed section, the first outflow rate of the side stream outflow section, and the reflux flow rate at the top of the column.

[0066] Step S103: Control the adjacent tower according to the control information.

[0067] In this embodiment, when performing dynamic process control on the partition column (i.e., the partition distillation column), based on the structure of the partition column and the equipment connected to it, first parameter information related to the partition column is obtained. Specifically, this first parameter information includes: the first heat load of the reboiler, the second heat load of the condenser, the feed section temperature difference, and the side-stream outlet temperature difference. It should be noted that the partition column is divided into a common rectification section near the top, a common stripping section near the bottom, a feed section with a feed inlet in the middle, and a side-stream outlet in the middle by a partition plate in the middle. The reboiler is connected to the bottom of the partition column, and the condenser is connected to the top of the partition column. Since the first heat load of the reboiler and the feed section temperature difference are related to feed disturbances, and the second heat load of the condenser and the side-stream outlet temperature difference are related to component extraction, this first parameter information is used as the dependent variable for control of the partition column.

[0068] Specifically, in the control process, the first step involves jointly processing the acquired first parameter information with a pre-built adaptive neuro-fuzzy inference system and MPC model to obtain the control information required for controlling the adjacent column. This control information represents the target state that the adjacent column needs to achieve, specifically including the liquid flow rate allocated from the common rectification section to the feed section, the first outflow rate from the side-stream outflow section, and the reflux flow rate at the top of the column. Furthermore, by controlling the adjacent column according to this control information, adaptive adjustments can be made to address feed disturbances, allowing it to quickly return to its original setpoint. In summary, this embodiment, by employing a combined control system of an adaptive neuro-fuzzy inference system and an MPC model, solves the control problems of difficulty in real-time component measurement and strong variable coupling during the rectification process, thus improving dynamic control performance. Moreover, the breakthrough of using temperature difference as a control point better addresses control problems such as variable coupling, effectively resisting disturbances in feed flow rate and feed composition, thus meeting product specifications while achieving control results with smaller overshoot, transition time, and oscillation. Simultaneously, it reduces the frequency of manual operation, effectively ensuring the safety performance of the production process.

[0069] It should be noted that the goal of control through the MPC model and adaptive neurofuzzy inference system is to minimize: the difference between the controlled variable (e.g., product purity) and the target value of the controlled variable; the difference between the manipulated variable (e.g., temperature difference at critical control points) and the target value of the manipulated variable; and the fluctuation of the manipulated variable.

[0070] It should also be noted that the predictive control of the MPC model in this embodiment is an advanced process control strategy. Its principle includes: data prediction, online optimization, and feedback correction. Data prediction involves predicting the corresponding predicted values, and even the values ​​of other related parameters, based on the historical states and current input values ​​of predetermined key control points, to demonstrate the future dynamic behavior of the system. Online optimization involves optimizing the parameters online for a finite future time period from that time at each sampling moment. At the next sampling moment, the optimal sequence from the previous moment is used for optimization, and this process is repeated. Feedback correction involves detecting the corresponding actual output at the next moment to prevent model mismatch and bias, and using real-time information to correct the model-based predictions before performing new optimizations. The feedback information forms a closed-loop optimization.

[0071] See Figure 2 Specifically, as described above, the step of obtaining control information based on the first parameter information, the adaptive neural fuzzy inference system, and the MPC model through joint processing includes:

[0072] Step S201: Based on the MPC model, predict the temperature difference in the feed section to obtain the liquid flow rate;

[0073] Step S202: Based on the MPC model, perform model prediction on the temperature difference of the side-line extraction section to obtain the first extraction flow rate;

[0074] Step S203: Perform fuzzy reasoning on the first heat load and the second heat load according to the adaptive neural fuzzy reasoning system to obtain the set temperature of the common rectification section, and combine the set temperature, the first temperature of the common rectification section and the second outflow rate at the top of the column to obtain the reflux flow rate.

[0075] In this embodiment, when obtaining control information through joint processing based on the first parameter information, prediction and / or inference are performed separately based on the correlation between each parameter in the first parameter information and each parameter in the control information, as well as the complexity of the correlation. Specifically, the temperature difference in the feed section is predicted using the MPC model to obtain the liquid phase flow rate; the temperature difference in the side stream extraction section is predicted using the MPC model to obtain the first extraction flow rate; the first heat load and the second heat load are fuzzy inferred using the adaptive neural fuzzy inference system to obtain the set temperature of the common rectification section; and the reflux flow rate is obtained by combining the set temperature obtained by fuzzy inference, the first temperature of the common rectification section, and the second extraction flow rate at the top of the column.

[0076] It should be noted that, depending on the different relationships or calculation formulas between the temperature difference in the feed section and the liquid flow rate, and between the temperature difference in the side-stream extraction section and the first extraction flow rate, MPC models can be constructed accordingly. That is, each relationship corresponds to an MPC model, in order to improve calculation efficiency and reduce the error rate by avoiding the influence of irrelevant parameters.

[0077] See Figure 3 Furthermore, as described above, obtaining the reflux flow rate by combining the set temperature, the first temperature of the common rectification section, and the second outflow rate at the top of the column includes:

[0078] Step S301: Determine the set reflux ratio based on the set temperature and the first temperature;

[0079] Step S302: Determine the reflux flow rate based on the set reflux ratio and the second extraction flow rate.

[0080] In this embodiment, when obtaining the reflux flow rate, since the set temperature of the common rectification section is obtained through fuzzy inference based on the first and second heat loads, and the reflux flow rate to be controlled at the top of the column needs to be obtained, the current first temperature of the common rectification section needs to be acquired. Based on the relationship between the first temperature and the set temperature, the set reflux ratio required when the first temperature changes to the set temperature is determined. The required reflux flow rate to be refluxed to the top of the column can then be determined based on the set reflux ratio and the second outflow rate at the top of the column. It should be noted that the sum of the reflux flow rate and the second outflow rate is the pumping flow rate of the reflux pump.

[0081] It should be noted that the set temperature can be the target value of the common rectification section temperature directly, or it can be controlled in segments from the first temperature to the target value of the common rectification section temperature. That is, there are multiple set temperatures with the same temperature difference between the two, and the setting temperature is achieved through gradual adjustment.

[0082] See Figure 4 Specifically, according to the method described above, obtaining the first parameter information includes:

[0083] Step S401: Obtain the second temperature of the feed section sensitive plate and the third temperature of the side-line extraction section sensitive plate;

[0084] Step S402: Based on a preset reference temperature, determine the difference between the second temperature and the corresponding reference temperature as the temperature difference of the feeding section, and determine the difference between the third temperature and the corresponding reference temperature as the temperature difference of the side-line sampling section, wherein the reference temperature is determined based on the average absolute deviation of the temperature change of the sensitive plate in a closed loop with component disturbance.

[0085] In this embodiment, the acquisition of the feed section temperature difference and the side-stream extraction section temperature difference in the first parameter information is illustrated. Since the sensitive plate is the plate in the adjacent tower that is sensitive to external disturbance factors, in this embodiment, the second temperature of the feed section sensitive plate (i.e., the sensitive plate corresponding to the feed section) and the third temperature of the side-stream extraction section sensitive plate (i.e., the sensitive plate corresponding to the side-stream extraction section) are acquired. By subtracting the second and third temperatures from the corresponding pre-obtained reference temperatures, the feed section temperature difference and the side-stream extraction section temperature difference are obtained respectively. The impact of feed disturbance on the feed section and the side-stream extraction section can be determined by the feed section temperature difference and the side-stream extraction section temperature difference, so as to facilitate subsequent model prediction based on the above temperature differences using the MPC model.

[0086] It should be noted that this reference temperature is determined based on the average absolute deviation of the temperature change of the sensitive plate in a closed loop with component disturbances (i.e., product purity remains stable). It should also be noted that the feed section and the side-stream extraction section can each correspond to a separate reference temperature, or they can correspond to the same reference temperature.

[0087] See Figure 5 Specifically, the method described above also includes:

[0088] Step S501: Obtain historical parameter samples of the adjacent tower, the historical parameter samples include: the first heat load and the second heat load corresponding to each detection point within a preset time period;

[0089] Step S502: Normalize the first heat load and the second heat load;

[0090] Step S503: Determine the membership function and number of fuzzy rules of the preset adaptive neural fuzzy inference system based on the normalized first heat load and second heat load;

[0091] Step S504: Train the preset adaptive neural fuzzy inference system based on the normalized first thermal load and the second thermal load to obtain the adaptive neural fuzzy inference system.

[0092] In this embodiment, the construction of the adaptive neurofuzzy inference system is further explained. First, historical parameter samples of the adjacent tower are obtained. These historical parameter samples include the first and second heat loads corresponding to each detection point within a preset time period. This can also be understood as the first and second heat loads detected by each detection point constituting the historical parameter samples corresponding to that detection point. Then, the first and second heat loads are normalized to ensure that the data ultimately used in model training have the same metric scale and improve training efficiency. Next, based on the normalized first and second heat loads, the membership function and the number of fuzzy rules of the preset adaptive neurofuzzy inference system are determined. Specifically, this can be determined using a trial-and-error method in Matlab software. It should be noted that the particle swarm optimization algorithm can also be used to optimize the model parameters of the preset adaptive neurofuzzy inference system. Finally, the preset adaptive neurofuzzy inference system is trained using the normalized first and second heat loads to obtain the desired adaptive neurofuzzy inference system. The training and use of the preset adaptive neurofuzzy inference system can be achieved using the Adaptive Neurofuzzy Inference System (ANFIS) toolbox in Matlab software.

[0093] It should be noted that the set temperature or other output values ​​obtained through the adaptive neurofuzzy inference system are all values ​​after denormalization.

[0094] See Figure 6 Specifically, the method described above also includes:

[0095] Step S601: Obtain the feed flow rate of the feed section of the partition tower and the fourth temperature of the common stripping section;

[0096] Step S602: Determine the heat load control information of the reboiler based on the feed flow rate of the feed section and the fourth temperature;

[0097] Step S603: Control the first heat load of the reboiler according to the heat load control information.

[0098] Since the heat load of the reboiler affects the dynamic control of the process in the adjacent column, this embodiment provides an example of adjusting the heat load of the reboiler. Specifically, this includes: first, obtaining the feed flow rate of the feed section and the fourth temperature of the common stripping section, where the fourth temperature is the sensitive plate temperature of the common stripping section; then, determining the heat load control information of the reboiler based on the preset adjustment relationship between the feed flow rate and the fourth temperature and the first heat load (for example, different feed flow rates correspond to different ratios of the fourth temperature to the first heat load); and finally, controlling the first heat load of the reboiler based on this heat load control information, where the fourth temperature is used for negative feedback.

[0099] See Figure 7 Another embodiment of this application also provides a process dynamic control device for a partition tower, comprising:

[0100] The first processing module 701 is used to obtain the first parameter information of the adjacent tower. The first parameter information includes: the first heat load of the reboiler, the second heat load of the condenser, the temperature difference of the feed section, and the temperature difference of the side stream outlet section.

[0101] The second processing module 702 is used to perform joint processing based on the first parameter information, the adaptive neural fuzzy inference system, and the model predictive control (MPC) model to obtain control information. The control information includes: the liquid phase flow rate allocated from the common rectification section to the feed section, the first outflow flow rate of the side stream outflow section, and the reflux flow rate at the top of the column.

[0102] The third processing module 703 is used to control the adjacent tower according to the control information.

[0103] Specifically, in the apparatus described above, the second processing module includes:

[0104] The first processing unit is used to perform model prediction on the temperature difference of the feed section based on the MPC model to obtain the liquid phase flow rate;

[0105] The second processing unit is used to perform model prediction on the temperature difference of the side-line extraction section according to the MPC model to obtain the first extraction flow rate.

[0106] The third processing unit is used to perform fuzzy reasoning on the first heat load and the second heat load according to the adaptive neural fuzzy reasoning system to obtain the set temperature of the common rectification section, and to obtain the reflux flow rate by combining the set temperature, the first temperature of the common rectification section and the second outflow rate at the top of the column.

[0107] Furthermore, in the apparatus described above, the third processing unit includes:

[0108] The first sub-processing unit is used to determine the set reflux ratio based on the set temperature and the first temperature;

[0109] The second sub-processing unit is used to determine the reflux flow rate based on the set reflux ratio and the second extraction flow rate.

[0110] Specifically, in the apparatus described above, the first processing module includes:

[0111] The fourth processing unit is used to obtain the second temperature of the feed section sensitive plate and the third temperature of the side-line extraction section sensitive plate;

[0112] The fifth processing unit is used to determine, based on a preset reference temperature, the difference between the second temperature and the corresponding reference temperature as the temperature difference of the feeding section, and to determine the difference between the third temperature and the corresponding reference temperature as the temperature difference of the side-line sampling section, wherein the reference temperature is determined based on the average absolute deviation of the temperature change of the sensitive plate in a closed loop with component disturbance.

[0113] Specifically, the device described above further includes:

[0114] The fourth processing module is used to obtain historical parameter samples of the adjacent tower, the historical parameter samples including: the first heat load and the second heat load corresponding to each detection point within a preset time period;

[0115] The fifth processing module is used to normalize the first heat load and the second heat load;

[0116] The sixth processing module is used to determine the membership function and the number of fuzzy rules of the preset adaptive neural fuzzy inference system based on the normalized first heat load and the second heat load;

[0117] The seventh processing module is used to train the preset adaptive neural fuzzy inference system based on the normalized first thermal load and the second thermal load to obtain the adaptive neural fuzzy inference system.

[0118] Specifically, the device described above further includes:

[0119] The eighth processing module is used to obtain the feed flow rate of the feed section of the partition tower and the fourth temperature of the common stripping section;

[0120] The ninth processing module is used to determine the heat load control information of the reboiler based on the feed flow rate of the feed section and the fourth temperature.

[0121] The tenth processing module is used to control the first heat load of the reboiler according to the heat load control information.

[0122] See Figure 8 The system embodiments of this application are systems corresponding to the embodiments of the above methods. All implementation means in the embodiments of the above methods are applicable to the embodiments of this system and can achieve the same technical effect.

[0123] Another embodiment of this application provides a process dynamic control system for a partition column, including: partition column T101, reboiler E101, condenser E102, reflux pump P101, reflux tank V101, reflux control valve A, side-stream pump P102, side-stream control valve B, liquid phase distribution control valve C, side-stream temperature difference controller TDC1, feed temperature difference controller TDC2, adaptive neuro-fuzzy inference system controller ANFISI, and model prediction controller MPC;

[0124] The partition tower T101 is divided into a common rectification section I near the top of the tower, a common stripping section II near the bottom of the tower, a feed section III with a feed inlet in the middle, and a side-stream extraction section IV with a side-stream extraction outlet in the middle, by a partition plate set in the middle.

[0125] The reboiler E101 is connected to the common stripping section II;

[0126] The condenser E102 is connected to the common rectification section I via the reflux tank V101, the reflux pump P101, and the reflux control valve A.

[0127] The pipeline containing the sideline extraction control valve B and the sideline extraction pump P102 is connected to the sideline extraction section IV.

[0128] The side-line sampling section temperature difference controller TDC1 is connected to the side-line sampling section IV;

[0129] The temperature difference controller TDC2 for the feeding section is connected to the feeding section III.

[0130] The liquid phase distribution control valve C is connected to the common rectification section I and the feed section III via pipelines;

[0131] The adaptive neural fuzzy inference system controller ANFISI is connected to the reboiler E101, the condenser E102 and the reflux control valve A. It is used to obtain the heat load of the reboiler E101 and the condenser E102, perform fuzzy inference, and control the reflux flow rate at the top of the column by controlling the reflux control valve A.

[0132] The model prediction controller MPC is connected to the side-line extraction section temperature difference controller TDC1 and the side-line extraction control valve B. It is used to perform model prediction based on the side-line extraction section temperature difference ΔT2 obtained by the side-line extraction section temperature difference controller TDC1, and to control the first extraction flow rate of the side-line extraction section IV by controlling the side-line extraction control valve B.

[0133] The model prediction controller MPC is also connected to the feed section temperature difference controller TDC2 and the liquid phase distribution control valve C. It is used to perform model prediction based on the temperature difference ΔT4 of feed section III obtained by the feed section temperature difference controller TDC2, and to control the liquid phase flow rate distributed from the common rectification section I to the feed section III by controlling the liquid phase distribution.

[0134] In this embodiment, a process dynamic control system for a partition tower is specifically disclosed, including a partition tower T101 and various related devices or modules. The connection relationships between the various devices or modules are shown below.

[0135] The reboiler E101 is connected to the common stripping section II to achieve reboiling treatment of the adjacent column T101;

[0136] The condenser E102 is connected to the common rectification section I, which can be the top of the column, through the reflux tank V101, the reflux pump P101 and the reflux control valve A, so that the gas phase generated in the common rectification section I can be condensed, collected and refluxed to the adjacent column.

[0137] The pipeline containing the sideline extraction control valve B and the sideline extraction pump P102 is connected to the sideline extraction section IV to achieve sideline extraction.

[0138] The liquid phase distribution control valve C is connected to the common rectification section I and the feed section III through pipelines to realize the redistribution of gas and liquid phases between the common rectification section I and the feed section III;

[0139] The Adaptive Neural Fuzzy Inference System Controller (ANFISI) is connected to the reboiler E101, the condenser E102, and the reflux control valve A. After obtaining the heat load of the reboiler E101 and the condenser E102, it performs fuzzy inference and controls the reflux flow rate at the top of the column by controlling the reflux control valve A.

[0140] The model prediction controller MPC is connected to the side-line extraction section temperature difference controller TDC1 and the side-line extraction control valve B. It is used to perform model prediction based on the side-line extraction section temperature difference ΔT2 obtained by the side-line extraction section temperature difference controller TDC1, and to control the first extraction flow rate of the side-line extraction section IV by controlling the side-line extraction control valve B.

[0141] The model prediction controller MPC is also connected to the feed section temperature difference controller TDC2 and the liquid phase distribution control valve C. It is used to perform model prediction based on the temperature difference ΔT4 of feed section III obtained by the feed section temperature difference controller TDC2, and to control the liquid phase flow rate distributed from the common rectification section I to the feed section III by the liquid phase distribution control valve C.

[0142] See Figure 8 Specifically, the system described above also includes: a common distillation section temperature controller TC1 and a reflux ratio controller RR;

[0143] The common distillation section temperature controller TC1 is connected to the common distillation section, and is also connected to the adaptive neural fuzzy inference system controller ANFISI and the reflux ratio controller. It is used to determine the set reflux ratio of the reflux ratio controller based on the first temperature T1 of the common distillation section I and the temperature set value output by the adaptive neural fuzzy inference system controller ANFISI.

[0144] The reflux ratio controller RR is also connected to the reflux control valve A, and is used to determine the reflux flow rate according to the set reflux ratio and the second output flow rate at the top of the tower, and to control the flow rate through the reflux control valve A according to the reflux flow rate.

[0145] This embodiment also provides examples of specific equipment or modules involved in the reflux ratio control process. Since the adaptive neural fuzzy inference system controller ANFISI obtains and outputs a temperature setpoint based on fuzzy inference of the heat load, the common rectification section temperature controller TC1 determines the corresponding set flow rate ratio based on the detected first temperature T1 of the common rectification section and the temperature setpoint. That is, the ratio of the reflux flow rate required to adjust the first temperature T1 to the temperature setpoint to the second product flow rate at the top of the column. Then, the reflux ratio controller RR obtains the required reflux flow rate based on the set flow rate ratio and the detected second product flow rate. Thus, the required reflux flow rate can be obtained by controlling the reflux control valve A to adjust the flow rate.

[0146] Specifically, the system described above further includes: a common stripping section temperature controller TC2 and a reboiler heat load ratio controller Q1 / F, wherein the common stripping section temperature controller TC2 is used to obtain the fourth temperature T3 of the common stripping section II, and the reboiler heat load ratio controller Q1 / F is used to determine the heat load control information of the reboiler E101 based on the obtained feed flow rate of the feed section III of the partition column and the fourth temperature, and to control the first heat load of the reboiler E101 according to the heat load control information.

[0147] exist Figure 8 It should also be noted that the pressure controller PC is used to regulate the operating pressure of the adjacent tower T101;

[0148] The reflux tank level controller LC1 is used to adjust the liquid level of the reflux tank V101 of the partition tower T101 through the first control liquid level control valve F;

[0149] The column bottom level controller LC2 is used to adjust the liquid level in the column bottom tank through the second liquid level control valve E;

[0150] The bottom pump P102 is used to extract the product from the bottom of the tower.

[0151] The flow control valve FC is used to control the feed flow rate.

[0152] Another embodiment of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the process dynamic control method for the partition tower as described above.

[0153] Furthermore, reference numerals and / or letters may be repeated in different examples within this application. Such repetition is for the purpose of simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or settings discussed.

[0154] It should also be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion.

[0155] The above description is the preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principles described in this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for dynamic process control of a partition tower, characterized in that, include: Obtain the first parameter information of the adjacent tower, which includes: the first heat load of the reboiler, the second heat load of the condenser, the temperature difference of the feed section, and the temperature difference of the side stream outlet section; Based on the first parameter information, and through joint processing of the adaptive neural fuzzy inference system and the model predictive control (MPC) model, control information is obtained. The control information includes: the liquid phase flow rate allocated from the common rectification section to the feed section, the first outflow rate of the side stream outflow section, and the reflux flow rate at the top of the column. The adjacent tower is controlled according to the control information; The control information obtained by jointly processing the first parameter information, the adaptive neural fuzzy inference system, and the model predictive control (MPC) model includes: The liquid flow rate is obtained by predicting the temperature difference in the feed section based on the MPC model. Based on the MPC model, the temperature difference of the side-line extraction section is predicted to obtain the first extraction flow rate. The adaptive neural fuzzy inference system performs fuzzy inference on the first heat load and the second heat load to obtain the set temperature of the common rectification section, and combines the set temperature, the first temperature of the common rectification section and the second outflow rate at the top of the column to obtain the reflux flow rate.

2. The method according to claim 1, characterized in that, The reflux flow rate is obtained by combining the set temperature, the first temperature of the common rectification section, and the second outflow rate at the top of the column, including: The set reflux ratio is determined based on the set temperature and the first temperature; The reflux flow rate is determined based on the set reflux ratio and the second extraction flow rate.

3. The method according to claim 1, characterized in that, The step of obtaining the first parameter information includes: The second temperature of the feed section sensitive plate and the third temperature of the side-line extraction section sensitive plate are obtained; Based on a preset reference temperature, the difference between the second temperature and the corresponding reference temperature is determined as the temperature difference of the feeding section, and the difference between the third temperature and the corresponding reference temperature is determined as the temperature difference of the side-line sampling section. The reference temperature is determined based on the average absolute deviation of the temperature change of the sensitive plate in a closed loop with component disturbance.

4. The method according to claim 1, characterized in that, Also includes: Obtain historical parameter samples of the adjacent tower, the historical parameter samples including: the first heat load and the second heat load corresponding to each detection point within a preset time period; The first heat load and the second heat load are normalized. Based on the normalized first heat load and second heat load, determine the membership function and number of fuzzy rules of the preset adaptive neural fuzzy inference system; The preset adaptive neural fuzzy inference system is trained based on the normalized first thermal load and the second thermal load to obtain the adaptive neural fuzzy inference system.

5. The method according to claim 1, characterized in that, Also includes: Obtain the feed flow rate of the feed section of the partition tower and the fourth temperature of the common stripping section; The heat load control information of the reboiler is determined based on the feed flow rate of the feed section and the fourth temperature. The first heat load of the reboiler is controlled according to the heat load control information.

6. A process dynamic control device for a partition tower, characterized in that, include: The first processing module is used to acquire the first parameter information of the adjacent tower, which includes: the first heat load of the reboiler, the second heat load of the condenser, the temperature difference of the feed section, and the temperature difference of the side stream outlet section. The second processing module is used to perform joint processing based on the first parameter information, the adaptive neural fuzzy inference system, and the model predictive control (MPC) model to obtain control information. The control information includes: the liquid phase flow rate allocated from the common rectification section to the feed section, the first outflow flow rate of the side stream outflow section, and the reflux flow rate at the top of the column. The third processing module is used to control the adjacent tower according to the control information; The second processing module includes: The first processing unit is used to perform model prediction on the temperature difference of the feed section based on the MPC model to obtain the liquid phase flow rate; The second processing unit is used to perform model prediction on the temperature difference of the side-line extraction section according to the MPC model to obtain the first extraction flow rate. The third processing unit is used to perform fuzzy reasoning on the first heat load and the second heat load according to the adaptive neural fuzzy reasoning system to obtain the set temperature of the common rectification section, and to obtain the reflux flow rate by combining the set temperature, the first temperature of the common rectification section and the second outflow rate at the top of the column.

7. A process dynamic control system for a partition tower, characterized in that, include: Divider tower, reboiler, condenser, reflux pump, reflux tank, reflux control valve, side-stream sampling pump, side-stream sampling control valve, liquid phase distribution control valve, side-stream sampling section temperature difference controller, feed section temperature difference controller, adaptive neural fuzzy inference system controller and model prediction controller; The partition tower is divided into a common rectification section near the top of the tower, a common stripping section near the bottom of the tower, a feed section with a feed inlet in the middle, and a side-stream extraction section with a side-stream extraction outlet in the middle by a partition plate set in the middle. The reboiler is connected to the common stripping section; The condenser is connected to the common rectification section via the reflux tank, the reflux pump, and the reflux control valve; The sideline extraction control valve and the pipeline containing the sideline extraction pump are connected to the sideline extraction section. The temperature difference controller for the side-line extraction section is connected to the side-line extraction section; The temperature difference controller for the feeding section is connected to the feeding section; The liquid phase distribution control valve is connected to the common distillation section and the feed section via pipelines; The adaptive neural fuzzy inference system controller is connected to the reboiler, the condenser and the reflux control valve. It is used to obtain the heat load of the reboiler and the condenser, perform fuzzy inference, and control the reflux flow rate at the top of the column by controlling the reflux control valve. The model prediction controller is connected to the side-line extraction section temperature difference controller and the side-line extraction control valve. It is used to perform model prediction based on the side-line extraction section temperature difference obtained by the side-line extraction section temperature difference controller, and to control the first extraction flow rate of the side-line extraction section by controlling the side-line extraction control valve. The model prediction controller is also connected to the feed section temperature difference controller and the liquid phase distribution control valve, and is used to perform model prediction based on the feed section temperature difference obtained by the feed section temperature difference controller, and to control the liquid phase flow rate distributed from the common rectification section to the feed section by controlling the liquid phase distribution valve.

8. The system according to claim 7, characterized in that, Also includes: Common rectification section temperature controller and reflux ratio controller; The common distillation section temperature controller is connected to the common distillation section, and is also connected to the adaptive neural fuzzy inference system controller and the reflux ratio controller. It is used to determine the set reflux ratio of the reflux ratio controller based on the first temperature of the common distillation section and the temperature set value output by the adaptive neural fuzzy inference system. The reflux ratio controller is also connected to the reflux control valve and is used to determine the reflux flow rate based on the set reflux ratio and the second outflow rate at the top of the tower, and to control the flow rate through the reflux control valve based on the reflux flow rate.

9. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements the steps of the process dynamic control method for the partition tower as described in any one of claims 1 to 5.

Images