Intelligent control system and intelligent control method for catalytic-free tert-butane production

Abstract

This invention provides an intelligent control system and method for the production of tert-butane without catalytic reduction, belonging to the field of tert-butane production control. The intelligent control system includes a host computer and sensing and execution mechanisms electrically connected to the host computer. The sensing mechanisms include a TDLAS analyzer installed at the top outlet of the reactive distillation column, NIR analyzers installed at the tert-butanol inlet of the mixing preheater and the side outlet of the reactive distillation column, temperature sensors installed at the top, bottom, and side outlet of the reactive distillation column, flow meters installed at the feed inlet, top outlet, and reflux inlet of the reactive distillation column, and multiple pressure sensors spaced at equal intervals along the height of the reaction section. The execution mechanisms include electrically operated regulating valves installed at the feed inlet and reflux inlet of the reactive distillation column. The control system integrates multiple sensors and execution mechanisms, enabling closed-loop intelligent control of the entire process.

Description

Technical Field

[0001] This invention belongs to the field of tert-chlorobutane production control technology, specifically relating to an intelligent control system and intelligent control method for catalytic-free tert-chlorobutane production. Background Technology

[0002] tert-butane chloride is an important organic chemical raw material widely used in pharmaceuticals, pesticides, and other fields. The applicant proposes a catalyst-free method for producing tert-butane chloride. Dilute hydrochloric acid is added as the starting liquid to the bottom of a reactive distillation column, and the column is heated to boiling until reflux occurs at the top. Tert-butanol and concentrated hydrochloric acid are added to a mixing preheater for preheating, and then continuously fed into the reactive distillation column. After the reaction, the product is discharged from the side stream, with reflux at the top. This method achieves efficient production by coupling the reaction and separation of tert-butanol and concentrated hydrochloric acid in the reactive distillation column. As the core equipment integrating reaction and separation, intelligent process control of the reactive distillation column is crucial for achieving continuous, high-quality production.

[0003] Currently, the continuous production of tert-butane chloride relies on manual adjustment of key parameters such as feed rate based on temperature data. However, this method has problems such as the inability to effectively monitor process parameters, lag in feedback control, and inability to effectively respond to disturbances such as feed concentration and flow rate. Therefore, an intelligent control method for the production of tert-butane chloride without catalysis is needed.

[0004] A search revealed a lack of intelligent control solutions for the reactive distillation of tert-butane chloride without catalytic cracking. Existing intelligent control technologies for distillation processes are unable to meet the control requirements of multi-factor disturbances and deep process-equipment coordination in this process. For example, patent document CN121411368A proposes an intelligent anti-interference cascade control method for distillation columns, comprising: S1: real-time monitoring of the main controlled variable, the secondary controlled variable, and at least one feedforward disturbance variable in the distillation process; S2: calculating the feedforward compensation amount based on the change of the feedforward disturbance variable using a dynamic feedforward compensation model; S3: calculating the feedback control amount based on the deviation of the main controlled variable and the secondary controlled variable from the setpoint using a cascade control loop; S4: adding the feedforward compensation amount and the feedback control amount to obtain the final control command; S5: performing amplitude constraint processing and rate of change constraint processing on the final control command, and then outputting it to the actuator controlling the manipulated variable of the distillation column.

[0005] The existing technology has the following shortcomings and cannot meet the special control requirements of non-catalytic reactive distillation of tert-butane chloride: First, the existing feedforward compensation model only addresses the disturbance variable of feed flow rate. The feedforward gain is a fixed parameter, and the compensation amount is calculated using a transfer function in the form of lead-lag. The weight of the feedforward compensation amount in the final control command is fixed, making it impossible to dynamically adjust the strength of the feedforward action according to changes in operating conditions. In actual production, the production process of catalytically untreated tert-butane chloride is affected by multiple disturbances such as feed flow rate and raw material concentration. A fixed feedforward gain cannot adequately account for the influence characteristics of different disturbance sources, leading to a decrease in control accuracy under complex disturbance conditions.

[0006] Second, the existing technology only outputs the final control command to a single manipulated variable (such as reboiler steam flow rate). For a non-catalytic tert-butane reactive distillation column, the feed flow rate directly determines the reaction residence time and production capacity, while the reflux ratio affects the gas-liquid balance and separation efficiency within the column. Both need to work together to optimize production efficiency while ensuring product quality. The adjustment capability of a single manipulated variable control strategy is limited, making it difficult to simultaneously meet the requirements of product quality and production efficiency under wide operating conditions. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides an intelligent control system and intelligent control method for the production of catalytically non-catalytically chloro-tert-butane.

[0008] To solve one or more or all of the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A smart control system for the non-catalytic production of tert-chlorobutane includes a host computer and a sensing mechanism and an actuator, both electrically connected to the host computer. The sensing mechanism includes a TDLAS analyzer installed at the top outlet of the reactive distillation column, an NIR analyzer installed at the tert-butanol inlet of the mixing preheater and the side outlet of the reactive distillation column, a temperature sensor installed at the top, bottom, and side outlet of the reactive distillation column, a flow meter installed at the inlet, top outlet, and reflux inlet of the reactive distillation column, and multiple pressure sensors installed at equal intervals along the height of the reaction section. The actuator includes electrically operated regulating valves respectively installed at the feed inlet and the reflux liquid inlet at the top of the reactive distillation column; The host computer is used to automatically control the feed flow rate of the reactive distillation column, using the side stream outlet temperature and side stream hydrogen chloride concentration as feedback, the tert-butanol concentration, the mixing preheater temperature, and the reactive distillation column feed flow rate as feedforward, and the top hydrogen chloride concentration as the feedforward weight. Existing technology CN121411368A uses an online gas chromatograph to measure the top product concentration with a sampling cycle of 5 minutes. This long-cycle concentration detection results in significant lag in both feedforward compensation and feedback control, making it unable to quickly respond to disturbances such as fluctuations in feedstock concentration. In the production of catalytically non-catalytically controlled tert-butane, rapid fluctuations in feedstock concentration directly affect the reaction process and product distribution, requiring higher-frequency concentration monitoring to achieve effective advance compensation. Therefore, this application uses TDLAS and NIR, which can perform second-level response, for online monitoring.

[0009] The intelligent control system provided by this invention integrates multiple detection devices and electric regulating valve actuators to construct an intelligent control system covering the entire process from raw material input, reaction process, to product output, achieving multi-dimensional real-time monitoring of the production process. Specifically, the TDLAS analyzer utilizes tunable diode laser absorption spectroscopy technology to achieve second-level online detection of the concentration of gaseous hydrogen chloride at the top of the column, overcoming the delay problem of traditional offline analysis; the NIR analyzer uses near-infrared spectroscopy technology to achieve real-time online analysis of tert-butanol concentration and side-stream product concentration, obtaining accurate quality indicators without manual sampling; temperature sensors, flow meters, and pressure sensors monitor the temperature distribution, material flow rate, and internal pressure gradient of the reaction section at the top, bottom, and side streams, respectively, providing complete data support for the control algorithm.

[0010] In the intelligent control system provided by this invention, the host computer of the control mechanism integrates intelligent control methods, preheating temperature adjustment methods, and filling early warning methods into one unit. By collecting multi-dimensional data, it calculates control parameters in real time and drives the electric regulating valves of the feed inlet and return liquid inlet to automatically adjust the flow rate. This realizes closed-loop automated operation from data acquisition, analysis and calculation to execution control. Compared with the traditional method that relies on human experience and offline analysis, it significantly reduces the intensity of human intervention and operational errors, improves control accuracy and response speed, makes product quality indicators more stable, and makes the production process safer and more reliable.

[0011] For the non-catalytic reactive distillation of tert-butane chloride, a unique process that deeply couples reaction and separation, there are specific requirements such as multi-factor disturbances, deep synergy between process and equipment, and precise control of product purity. Existing technology CN121411368A cannot provide a suitable solution. For example, patent document CN108721934A proposes a fully automated control method and system for a distillation column, achieving automated operation. However, this method is based on fixed process parameters and is unsuitable for the complex production process of tert-butane chloride, nor can it automatically adjust according to different operating conditions. Another example is patent document CN206473848U, which proposes a novel automated intelligent temperature control system for a distillation column. This system uses a PLC controller and temperature sensors to automatically control the circulation of cold and hot water within the distillation column, but it does not respond to disturbances such as upstream feed concentration and feed flow rate, making it unsuitable for non-catalytic reactive distillation of tert-butane chloride. For example, patent document CN101337133A proposes an automatic control device and method for the reflux ratio temperature of a distillation column, which uses a temperature detector and an intelligent regulator to control the reflux ratio of the distillation column. However, this control method is based only on the feed temperature parameter and does not consider other possible influencing factors in the production process, making it unsuitable for the non-catalytic reactive distillation of tert-butane chloride. Another example is patent document CN120742693A, which proposes an intelligent optimization control method and system for a distillation column. This method collects data on steam volume, refrigerant dosage, and tray temperature, uses cluster analysis to determine the thermal balance index, assesses the controllability of temperature change disturbances, dynamically adjusts the iteration number of the model predictive control algorithm, obtains the predicted temperature value using model prediction, and adjusts the steam volume and refrigerant dosage based on the relationship between the predicted temperature and the thermal balance index. However, the process parameters and control methods relied upon by this method differ significantly from those used in tert-butane chloride production, making it unsuitable for the non-catalytic reactive distillation of tert-butane chloride.

[0012] Therefore, based on the intelligent control system, this application proposes an intelligent control method for the production of catalytically non-catalytically controlled tert-butane, comprising: Real-time acquisition and recording of tert-butanol concentration, reactive distillation column feed flow rate, hydrogen chloride concentration collected from the top of the column, hydrogen chloride concentration and discharge temperature of the side stream, and mixing preheater temperature; The feedback terms of the control parameters are determined based on the relative deviations of the side discharge temperature and the side discharge hydrogen chloride concentration. The feedforward terms for the control parameters are obtained based on the relative deviation of tert-butanol concentration, the relative deviation of mixing preheater temperature, and the fluctuation of feed flow rate in the reactive distillation column. The dynamic weights of the feedforward term are determined based on the relative deviation of the hydrogen chloride concentration at the top of the column. The control parameters are obtained based on the feedback term, the feedforward term, and the feedforward weights. Adjust the feed flow rate of the reactive distillation column according to the control parameters.

[0013] Furthermore, the control parameters are expressed as follows: ,in For feedback items, It is a feedforward type. It is a feedforward dynamic weight.

[0014] Furthermore, the feedback term of the control parameter is represented as follows: Where k represents the sampling time index, This represents the relative deviation between the side-flow discharge temperature and the reference temperature. This represents the time interval between two adjacent samples, where N is the length of the integration time window. This represents the relative deviation of the side-line hydrogen chloride concentration. , and These are the proportional coefficient, integral coefficient, and derivative coefficient, respectively, fitted using historical data. This is the feedback coefficient of the lateral hydrogen chloride concentration fitted using historical data.

[0015] Furthermore, the feedforward term of the control parameter is expressed as: ,in The relative deviation between the feed flow rate and the average flow rate at sampling time k. This represents the relative deviation between the concentration of tert-butanol (k) at the sampling time and the reference concentration. The relative deviation between the mixing preheater temperature and the reference temperature. , and These are the sensitivity coefficients obtained by fitting historical data.

[0016] Furthermore, the dynamic weights of the feedforward term are expressed as follows: ,in and The upper and lower limits are preset for dynamic weights. β represents the relative deviation of the hydrogen chloride concentration at the top of the column, and β is the steepness of the transition. This is the preset transition center point.

[0017] Furthermore, it also includes: adjusting the reflux ratio at the top of the column according to control parameters; the reflux ratio is expressed as... ,in The control parameters are... λ is the reference reflux ratio, and λ is the pre-calibrated mapping coefficient.

[0018] The intelligent control method provided in this application breaks through the traditional single feedback mode that relies solely on temperature control by introducing the relative deviation of the side-line hydrogen chloride concentration as a control variable into the feedback term. It achieves dual closed-loop control of temperature and product quality indicators. Since the side-line hydrogen chloride concentration directly reflects the residual hydrogen chloride in the product and is strongly correlated with the purity of the final product, incorporating the side-line hydrogen chloride concentration into the feedback control can significantly improve the control accuracy, effectively reduce the fluctuation of residual hydrogen chloride in the product, and make the product purity more stably maintained at the target level of over 99.9%, thereby reducing rework and scrap losses caused by product quality fluctuations.

[0019] The intelligent control method provided in this application comprehensively considers three key disturbance factors in the feedforward term: feed flow rate fluctuation, tert-butanol concentration deviation, and mixing preheater temperature deviation. This allows for predictive adjustment of control parameters before the disturbance affects product quality. Compared to the lag problem of traditional feedback control, feedforward control can respond to changes in feed conditions in advance. When the raw material concentration or flow rate fluctuates, the control system immediately adjusts the feed flow rate and reflux ratio to compensate, avoiding significant fluctuations in product quality and significantly improving the system's anti-interference capability against feed disturbances. Under the feed condition of a target molar ratio of tert-butanol to concentrated hydrochloric acid of 1:1.05, and given the actual situation where the concentrated hydrochloric acid concentration is stable while the tert-butanol concentration is unstable, this feedforward control method selects the tert-butanol concentration, which has a greater impact on the reaction, as the feedforward indicator. Simultaneously, fluctuations in concentrated hydrochloric acid concentration are reflected in changes in the hydrogen chloride concentration at the top of the column. By setting dynamic weights in the feedforward term reflecting changes in the hydrogen chloride concentration at the top of the column, compensation for fluctuations in concentrated hydrochloric acid concentration is achieved.

[0020] The intelligent control method provided in this application applies control parameters to both feed flow rate adjustment and reflux ratio adjustment, forming a synergistic control strategy of feed-reflux ratio. Feed flow rate adjustment directly affects production capacity and reaction residence time, while reflux ratio adjustment affects gas-liquid balance and separation efficiency within the tower. The synergistic cooperation of the two can simultaneously control product quality from both the source and process dimensions. Compared with single parameter control, it has stronger adjustment capability and a wider control range, and can optimize energy consumption indicators while ensuring product quality, thereby achieving comprehensive optimization of product quality and production efficiency.

[0021] The intelligent control method provided in this application uses a combination of historical data fitting and Bayesian optimization to determine control parameters. First, it collects more than three months of historical data covering multiple operating conditions, uses the least squares method to obtain initial values ​​of linear parameters, and then obtains optimal values ​​of nonlinear parameters through Bayesian optimization or grid search. This data-driven parameter optimization method avoids the problems of traditional PID control parameter tuning relying on experience and having a long trial-and-error cycle, significantly reducing the difficulty and time of system debugging. At the same time, since the parameters are optimized based on a large amount of actual operating data, the control effect is more stable, reliable, and adaptable.

[0022] In traditional production processes, preheating in reactive distillation often uses a fixed temperature. However, in actual production, the feed flow rate frequently fluctuates, and a fixed preheating temperature will affect the stability of subsequent reactions. The existing control method in CN121411368A does not address the dynamic adjustment of the preheating temperature. In actual production, when the feed flow rate fluctuates, if the preheating temperature remains constant, the temperature of the material entering the reactive distillation column will deviate from the optimal reaction temperature, affecting the reaction rate and conversion rate. For the catalytic-free reactive distillation process of tert-butane chloride, the preheating temperature of tert-butanol and concentrated hydrochloric acid before feeding is crucial to the reaction initiation conditions, requiring the establishment of a linkage compensation mechanism between the feed flow rate and the preheating temperature. Therefore, based on an intelligent control system, this application proposes a preheating temperature adjustment method for the production of tert-butane chloride without catalytic control, including: Real-time acquisition and recording of feed flow rate and mixing preheater temperature in the reactive distillation column; The target temperature of the mixing preheater is determined based on the fluctuation of the feed flow rate. The target temperature is expressed as ,in and These are the upper and lower limits of the target temperature, respectively. The temperature of the mixing preheater at sampling time k; γ is the relative deviation between the feed flow rate and the average flow rate at sampling time k, with a value limited to ±0.1; γ is the flow rate influence coefficient.

[0023] The preheating temperature adjustment method provided in this application establishes a linkage compensation mechanism between feed flow rate and preheating temperature by real-time monitoring of feed flow rate fluctuations and dynamic adjustment of the mixing preheater temperature. When the feed flow rate increases, the preheating temperature is automatically increased to compensate for the increased heat demand caused by the increased material flow rate. When the feed flow rate decreases, the preheating temperature is automatically decreased to avoid energy waste and material overheating and decomposition caused by excessive preheating. This solves the problem of traditional fixed preheating temperatures failing to maintain a stable reaction temperature when the feed rate fluctuates, ensuring that the material entering the reactive distillation column always maintains a suitable reaction temperature, providing a fundamental guarantee for the stable progress of subsequent reactions. Through dynamic matching of feed flow rate and preheating temperature, it ensures that the material reaches the standard reaction temperature under different feed rate conditions, avoiding the problem of unstable reaction temperature caused by feed rate fluctuations. This is beneficial for the complete reaction of tert-butanol and hydrochloric acid, improving HCl conversion rate and product purity, and making product quality indicators more stable.

[0024] The preheating temperature adjustment method provided in this application introduces upper and lower limit constraints in the target temperature calculation, while limiting the feed flow fluctuation to within ±0.1, avoiding over-adjustment under extreme operating conditions. This dual limiting protection mechanism effectively prevents preheating temperature runaway caused by drastic flow fluctuations or sensor malfunctions, ensuring that the preheater temperature is always within the equipment and process safety range, reducing the risk of material decomposition and equipment damage caused by excessively high temperatures, as well as the problem of insufficient reaction caused by excessively low temperatures, thereby improving the safety and reliability of the production process.

[0025] Aging and blockage of the packing in the reaction section of a reactive distillation column can lead to changes in pressure drop within the column, affecting gas-liquid mass transfer and reaction efficiency. However, existing technologies lack a quantitative assessment method for packing condition and cannot automatically assess packing condition based on process parameters. The control method in existing technology CN121411368A is primarily used for product quality control and does not address the monitoring and early warning of packing condition within the reactive distillation column. For a non-catalytic tert-butane reactive distillation column, aging or blockage of the packing in the reaction section can cause changes in pressure drop within the column, affecting gas-liquid mass transfer and reaction efficiency. This existing technology lacks a quantitative assessment method for packing condition and cannot promptly detect signs of packing performance degradation. Therefore, based on an intelligent control system, this application proposes a packing early warning method for non-catalytic tert-butane production, including: Construct a correlation model between the pressure drop in the reaction section and the operating time and feed flow rate of the reactive distillation column; Obtain pressure data at each measuring point within the reaction section, and obtain the measured pressure drop within the reaction section; Obtain the feed flow rate of the reactive distillation column and calculate the baseline pressure drop and equivalent operating time; Substitute the measured pressure drop, the baseline pressure drop, the equivalent running time, and the design running time into the correlation model to obtain the time decay exponent; Early warning for packing replacement is provided based on the time decay index.

[0026] Furthermore, the association model is represented as follows: Where k represents the sampling time index, The pressure drop in the reaction section at sampling time k. denoted as b, where b is the design operating time of the packing material and b is the time decay exponent. The reference pressure drop at sampling time k. This is the equivalent operating time of the packing material; , To design pressure drop, Let k be the feed flow rate at sampling time k. To design the feed flow rate, n is the load index; ,in The time interval between two consecutive samples is denoted as m, where m is the load aging index. , This represents the average feed flow rate.

[0027] The packing early warning method provided in this application establishes a correlation model between the pressure drop in the reaction section and the operating time and feed flow rate, thereby realizing a quantitative assessment of the aging state of the packing. By utilizing the pressure drop, a parameter that directly reflects the internal fluid resistance of the packing, and combining the equivalent operating time and the design operating time to calculate the time decay index b, compared with the traditional method of relying solely on experience or periodically replacing the packing, it can accurately and in real time determine the degree of aging and blockage of the packing, promptly detect signs of declining packing performance, provide a scientific basis for packing maintenance and replacement, and avoid waste caused by premature packing replacement or decreased production efficiency caused by late replacement.

[0028] The packing condition early warning method provided in this application introduces the feed flow rate ratio into the correlation model to perform load correction on the baseline pressure drop and equivalent operating time. This eliminates the influence of different production loads on the pressure drop, ensuring the comparability of early warning indicators under different operating conditions. It solves the problems of false alarms caused by underestimating pressure drop at low loads and false alarms caused by overestimating pressure drop at high loads, which are problems of traditional early warning methods. This significantly improves the accuracy and reliability of packing condition early warning. Furthermore, it utilizes readily available feed flow rate data to calculate the equivalent load ratio, reducing the difficulty of implementation.

[0029] The packing early warning method provided in this application adopts a dynamic load aging index, which dynamically adjusts the aging rate according to the average load ratio in recent times. When the average load is higher than the design load, the aging index increases, accelerating the accumulation of equivalent running time and reflecting the actual situation of accelerated packing aging under high load conditions. When the average load is lower than the design load, the aging index decreases, slowing down the accumulation of equivalent running time and reflecting the actual situation of slower packing aging under low load conditions. This dynamic aging index mechanism makes the equivalent running time more consistent with the actual aging state of the packing, and the early warning is more targeted. Attached Figure Description

[0030] The present invention will now be described in further detail with reference to the accompanying drawings.

[0031] Figure 1 Schematic diagram of a catalyst-free tert-chlorobutane production unit; Figure 2 : A schematic diagram of Embodiment 1 of the present invention. Detailed Implementation

[0032] To better understand the present invention, the content of the invention is further clearly illustrated below with reference to embodiments and accompanying drawings. However, the scope of protection of the present invention is not limited to the embodiments described below. Numerous specific details are set forth in the following description to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the present invention can be practiced without one or more of these details.

[0033] The applicant has proposed a catalyst-free method for producing tert-butane chloride (refer to patent document CN202511913327.2), comprising: adding dilute hydrochloric acid as a starting liquid to the bottom of a reactive distillation column, heating the bottom to boiling, and then refluxing at the top of the column; adding tert-butanol and concentrated hydrochloric acid to a mixing preheater and preheating to 50-60°C, and then continuously feeding the mixture from the feed inlet of the reactive distillation column until the temperature at the top of the column is 45-48°C, the temperature at the side outlet is 50-52°C, and the temperature at the bottom of the column is 96-105°C, and then refluxing is performed, controlling the reflux ratio to be 1.5:1-3:1; continuously collecting crude tert-butane chloride from the side outlet, and then cooling, neutralizing, drying, filtering, or obtaining the finished product.

[0034] like Figure 1 As shown, in the production apparatus used in this production method, the reactive distillation column includes a rectification section, a reaction section, and a stripping section. A feed inlet is located at the lower part of the reaction section and is connected to a mixing preheater. A side outlet is located at the junction of the rectification section and the reaction section, and this side outlet is used to collect crude tert-butane chloride. A reflux condenser is also connected to the top of the distillation column for controlling the reflux of the light components distilled from the top (including hydrogen chloride, water, and trace amounts of tert-butanol). A reboiler is also connected to the bottom of the reactive distillation column.

[0035] Example 1: The purpose of this example is to provide an intelligent control system for the production of catalytically uncatalyzed tert-butane, enabling intelligent control of the above-mentioned production method. The control system includes a control mechanism and sensing and execution mechanisms electrically connected to the control mechanism.

[0036] The control mechanism includes a host computer.

[0037] The sensing system includes a TDLAS analyzer, an NIR analyzer, temperature sensors, flow meters, and pressure sensors. The TDLAS analyzer is installed at the top outlet of the reactive distillation column, allowing the host computer to collect real-time hydrogen chloride concentration data from the top. The NIR analyzers are installed at the tert-butanol inlet of the mixing preheater and at the side outlet of the reactive distillation column, allowing the host computer to collect real-time tert-butanol concentration input to the mixing preheater, as well as chlorotert-butane and hydrogen chloride concentrations from the side outlet. Temperature sensors are installed inside the top, bottom, and side outlet of the reactive distillation column, allowing the host computer to collect real-time temperatures at the top, bottom, and side outlet. Flow meters are installed at the feed inlet, top outlet, and top reflux inlet of the reactive distillation column, allowing the host computer to collect flow rates at the corresponding inlet and outlet. Multiple pressure sensors are set at equal intervals along the height of the reaction section. The host computer is used to collect the pressure values ​​at each point in the reaction section in real time using the pressure sensors to obtain pressure drop data. The pressure sensors are implemented using pressure transmitters or distributed fiber optic sensors.

[0038] The actuators include electrically operated regulating valves installed at the feed inlet and the reflux liquid inlet at the top of the reactive distillation column, respectively. The host computer can use the electrically operated regulating valves to adjust the flow rate of the corresponding inlet.

[0039] The host computer is used to execute the control method of Example 2, the preheating temperature adjustment method of Example 3, and the packing warning method of Example 4.

[0040] Example 2: The purpose of this example is to provide an intelligent control method for the production of catalytically uncatalyzed tert-butane, implemented using the control system provided in Example 1. Specifically, the control method includes: S201. Real-time acquisition and recording of tert-butanol concentration, reactive distillation column feed flow rate, hydrogen chloride concentration collected from the top of the column, hydrogen chloride concentration and discharge temperature of the side stream, and mixing preheater temperature.

[0041] S202. Determine the feedback items of the control parameters based on the relative deviation of the side discharge temperature and the relative deviation of the side discharge hydrogen chloride concentration.

[0042] Feedback items are represented as This includes, in sequence, a lateral temperature proportional term, a lateral temperature integral term, a lateral temperature differential term, and a lateral hydrogen chloride concentration feedback term. Where k represents the sampling time index; This represents the relative deviation between the side-flow discharge temperature and the reference temperature. , The reference temperature for the side line (e.g., 51℃). The measured value of the side outlet temperature at sampling time k; This represents the time interval between two adjacent samples. Step S201 uses fixed-frequency sampling, therefore... It is a fixed value; N is the length of the integration time window; This represents the relative deviation of the side-line hydrogen chloride concentration. , This represents the measured value of the hydrogen chloride concentration on the k-side at sampling time. The baseline value for the side-line hydrogen chloride concentration is 0.1% (e.g., 0.1%). , and These are the proportional coefficient, integral coefficient, and differential coefficient in the feedback term, respectively. This is the feedback coefficient for the side-line hydrogen chloride concentration.

[0043] S203. The feedforward term of the control parameters is obtained based on the relative deviation of tert-butanol concentration, the relative deviation of mixing preheater temperature, and the fluctuation of feed flow rate of reactive distillation column.

[0044] The feedforward term is represented as The feedforward terms are, in order, the feedforward terms for feed rate, tert-butanol concentration, and preheating temperature. Here, k represents the sampling time index. The feed flow rate fluctuation at sampling time k (relative deviation from the average flow rate) is given. , Let k be the feed flow rate at sampling time k. The average feed flow rate over a recent period (e.g., within the last hour); This represents the relative deviation between the concentration of tert-butanol (k) at the sampling time and the reference concentration. , Let be the concentration of tert-butanol at sampling time k. The baseline concentration of tert-butanol is (e.g., 85%). The relative deviation between the mixing preheater temperature and the reference temperature. , The temperature of the mixing preheater at sampling time k. This is a reference value for the temperature of the mixing preheater (e.g., 55°C). , and These are the corresponding sensitivity coefficients.

[0045] S204. Determine the dynamic weight of the feedforward term based on the relative deviation of the hydrogen chloride concentration at the top of the column.

[0046] The dynamic weights of the feedforward term are expressed as follows: ,in and The upper and lower limits for the dynamic weights are preset, preferably with values ​​of 0.2 and 0.5, respectively. This represents the relative deviation of the hydrogen chloride concentration at the top of the column. β represents the steepness of the transition; the larger the value, the faster the response to abnormal changes in the concentration of hydrogen chloride at the top of the column, such as a value of 8. The preset transition center point is preferably 0.3.

[0047] Therefore, the optimal formula for dynamic weights is: .

[0048] When the relative deviation of the hydrogen chloride concentration at the top of the column is less than 0.1, the production of tert-butane chloride is under normal operating conditions, and the dynamic weight is fixed at 0.2. When the relative deviation of the hydrogen chloride concentration at the top of the column is between 0.1 and 0.5, the production is under warning conditions. At this time, the dynamic weight is calculated based on the relative deviation, and the higher the deviation, the greater the weight. At the same time, an upper limit of 0.5 is set for the dynamic weight. When the relative deviation of the hydrogen chloride concentration at the top of the column is greater than 0.5, the production is under abnormal operating conditions, and the dynamic weight is fixed at 0.5 to avoid abnormal increases in the dynamic weight.

[0049] S205. Obtain control parameters based on feedback terms, feedforward terms, and feedforward term weights.

[0050] Control parameters .

[0051] In summary, control parameters It can be directly obtained from the following control model: Coefficients in the control model , , , , , , , The data is obtained by fitting historical records. First, at least three months of historical data are collected to ensure coverage of normal, warning, and abnormal operating conditions. The data is then cleaned to remove missing values, outliers, and steady-state transition points. Subsequently, feature engineering is performed on the historical data according to calculation needs, such as calculating the relative deviation variable and integral term as required by the control formula. and differential terms Subsequently, with β fixed at 8, the least squares method was used to fit and obtain the linear parameters. , , , , , , The initial β values ​​are obtained; finally, Bayesian optimization or grid search methods are used to refit linear parameters for each candidate β value during the optimization process, and finally the optimal value of β and the corresponding final value of linear parameters are obtained.

[0052] For example, the method for obtaining linear parameters using least squares fitting is as follows: When β is fixed at 8, the control model can be rewritten as a linear regression model, for example: .

[0053] in ~ Corresponding to the side line temperature ratio terms Sideline temperature integral term Side-line temperature differential term Sideline hydrogen chloride concentration feedback item Feed flow rate feedforward term tert-butanol concentration feedforward term Preheating temperature feedforward term ; For M sets of historical data, construct a matrix equation Where u is the control output vector, θ is the vector of parameters to be estimated. X is ~ The characteristic matrix formed; Using the least squares method The initial values ​​of the correlation coefficient are obtained by solving the problem.

[0054] S206. Adjust the feed flow rate of the reactive distillation column according to the control parameters.

[0055] The obtained control parameters As a control output, it is used to directly adjust the electric regulating valve at the feed inlet of the reactive distillation column to regulate the feed flow rate.

[0056] S207. Adjust the reflux ratio at the top of the tower according to the control parameters.

[0057] reflux ratio ,in The reference reflux ratio is 2.2; λ is the mapping coefficient, which controls the parameters. This is mapped to an adjustment amount for the reflux ratio. λ can be calibrated based on historical data or set based on process experience. For example, if the λ value is set to 0.2–0.3, then when… Within ±0.5, the reflux ratio adjustment range is ±0.1 to ±0.15.

[0058] Based on reflux ratio Given the current flow rate at the top outlet of the column, adjust the electric regulating valve at the top reflux inlet to achieve a reflux ratio at the top of the column. .

[0059] Example 3: The purpose of this example is to provide a method for adjusting the preheating temperature in the production of catalytically uncatalyzed tert-butane, implemented using the control system provided in Example 1. Specifically, the adjustment method includes: S301. Real-time acquisition and recording of feed flow rate and mixing preheater temperature of reactive distillation column.

[0060] S302. Determine the target temperature of the mixing preheater based on the fluctuation of the feed flow rate.

[0061] The feed flow rate of the reactive distillation column is the same as that of the mixing preheater. When the feed flow rate increases, the preheating temperature should be increased appropriately, and vice versa.

[0062] The target temperature obtained in this step ,in The temperature of the mixing preheater at sampling time k. The feed flow rate fluctuation at sampling time k (the relative deviation from the average flow rate, with the value limited to ±0.1). , Let k be the feed flow rate at sampling time k. The average feed flow rate over a recent period (e.g., within the last hour) is γ; γ is the flow rate influence coefficient, with a value of 20-50. For example, if γ is 30 (°C), when the feed flow rate fluctuates by 0.1, the temperature is increased by 3°C based on the reference temperature.

[0063] S303. Adjust the temperature of the mixing preheater to the target temperature.

[0064] Example 4: The purpose of this example is to provide a packing material early warning method for the production of catalytically non-catalytically controlled tert-butane, implemented using the control system provided in Example 1. Specifically, the early warning method includes: S401. Construct a correlation model between the pressure drop in the reaction section and the operating time and feed flow rate of the reactive distillation column.

[0065] The association model is represented as Where k represents the sampling time index, The pressure drop in the reaction section at sampling time k (the pressure difference between the top and bottom of the reaction section); b is the design operating time of the packing material; b is the time decay index. The reference pressure drop at sampling time k. This represents the equivalent operating time of the current packing material.

[0066] In production, the theoretical pressure drop will differ depending on the actual process load. Therefore, the correlation model uses a baseline pressure drop that takes into account the influence of the actual load. The actual process load will also affect the aging or clogging rate of the packing; therefore, the equivalent running time that takes into account the actual load is used in the correlation model. .

[0067] However, in the automated control or early warning of production, process load data is not easy to collect directly, and the load is directly related to the feed flow rate. Therefore, this application cleverly uses the ratio of feed flow rate to design feed flow rate to represent the load situation.

[0068] , For design pressure drop (pressure drop under the design process flow). Let k be the feed flow rate at sampling time k. For the design feed flow rate, n is the load index. The load index n can be calibrated through theoretical calculations, such as using the Ergun equation or the pressure drop correlation diagram (GPDC) method. For the production process and reactive distillation column involved in this application, the calculated value of n is 1.8 to 2, preferably 1.9. The load index n can also be calibrated by measuring pressure drop data under different loads (feed flow rates) with new packing conditions.

[0069] ,in The time interval between two consecutive samples is denoted as m, where m is the load aging index, reflecting the impact of load on the aging rate. , This represents the average feed flow rate over a recent period (e.g., the past month). It can reflect the average load ratio in recent times, and the load ratio has an exponential effect on the aging rate.

[0070] S402. Obtain the pressure data of each measuring point in the reaction section, and obtain the measured pressure drop of the reaction section.

[0071] S404. Obtain the feed flow rate of the reactive distillation column and calculate the baseline pressure drop and equivalent operating time.

[0072] S405. Substitute the measured pressure drop, the reference pressure drop, the equivalent running time, and the design running time into the correlation model to solve for the time decay index.

[0073] For association model , converted Taking the natural logarithm, then Therefore, the value of b can be calculated using the sampled data of the record through linear regression.

[0074] S405. Pre-warning for packing replacement based on time decay index.

[0075] In the correlation model, the time decay index b reflects the aging or blockage rate of the packing. When the b value is close to 0, the packing is in good condition and the pressure drop does not change much over time. When the b value is less than the warning threshold, the packing is in the normal aging stage and the pressure drop also increases slowly. When the b value is greater than the warning threshold, the packing ages rapidly or becomes abnormally blocked, and the pressure drop increases rapidly, requiring the packing to be replaced. When the b value is less than 0, an abnormal situation occurs where the pressure drop decreases over time, which may indicate that the packing is damaged.

[0076] Therefore, when the b value is greater than the warning threshold or less than 0, and the duration exceeds the set time (e.g., 12 hours), an alarm is issued to remind staff to replace the packing material. Preferably, the warning threshold value is 0.25.

[0077] Furthermore, multiple pressure sensors are evenly spaced along the height of the reaction section. The detected pressure values ​​can reflect the uniformity of the pressure drop within the reaction section. If the uniformity is less than 0.8, it indicates that a local blockage has occurred.

[0078] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Any other modifications or equivalent substitutions made by those skilled in the art to the technical solutions of the present invention, as long as they do not depart from the spirit and scope of the technical solutions of the present invention, should be covered within the scope of the claims of the present invention.

Claims

1. An intelligent control system for catalytic-free chlorination of tertiary butane production, characterized by, It includes a host computer and sensing and actuating mechanisms, both electrically connected to the host computer; The sensing mechanism includes a TDLAS analyzer installed at the top outlet of the reactive distillation column, an NIR analyzer installed at the tert-butanol inlet of the mixing preheater and the side outlet of the reactive distillation column, a temperature sensor installed at the top, bottom, and side outlet of the reactive distillation column, a flow meter installed at the inlet, top outlet, and reflux inlet of the reactive distillation column, and multiple pressure sensors installed at equal intervals along the height of the reaction section. The actuator includes electrically operated regulating valves respectively installed at the feed inlet and the reflux liquid inlet at the top of the reactive distillation column; The host computer is used to automatically control the feed flow rate of the reactive distillation column, using the side stream discharge temperature and side stream hydrogen chloride concentration as feedback, the tert-butanol concentration, the mixing preheater temperature and the reactive distillation column feed flow rate as feedforward, and the column top hydrogen chloride concentration as feedforward weight.

2. A method of intelligent control of catalytic-free chlorination of tertiary butane production, characterized by, include: Real-time acquisition and recording of tert-butanol concentration, reactive distillation column feed flow rate, hydrogen chloride concentration collected from the top of the column, hydrogen chloride concentration and discharge temperature of the side stream, and mixing preheater temperature; The feedback terms of the control parameters are determined based on the relative deviations of the side discharge temperature and the side discharge hydrogen chloride concentration. The feedforward terms for the control parameters are obtained based on the relative deviation of tert-butanol concentration, the relative deviation of mixing preheater temperature, and the fluctuation of feed flow rate in the reactive distillation column. The dynamic weights of the feedforward term are determined based on the relative deviation of the hydrogen chloride concentration at the top of the column. The control parameters are obtained based on the feedback term, the feedforward term, and the feedforward weights. Adjust the feed flow rate of the reactive distillation column according to the control parameters.

3. The intelligent control method for the production of catalytically uncatalyzed tert-butane according to claim 2, characterized in that, The control parameter is represented as wherein is a feedback term, is a feedforward term, is a feedforward dynamic weight.

4. The intelligent control method for the production of catalytically uncatalyzed tert-butane according to claim 3, characterized in that, The feedback term of the control parameter is expressed as wherein k represents the sampling time index, is the relative deviation of the side line discharge temperature from the reference temperature, represents the time interval between the adjacent two samplings, and N is the length of the integration time window, is the relative deviation of the side line hydrogen chloride concentration, , and are the proportional coefficient, the integral coefficient and the differential coefficient fitted by the historical data, respectively, is the side line hydrogen chloride concentration feedback coefficient fitted by the historical data.

5. The intelligent control method for the production of catalytically uncatalyzed tert-butane according to claim 3, characterized in that, The feed forward term of the control parameter is expressed as wherein is the relative deviation of the feed flow at sampling time k from the average flow, is the relative deviation of the t-butanol concentration at sampling time k from the reference concentration, is the relative deviation of the mixed preheater temperature from the reference temperature, , and are the sensitivity coefficients fitted through historical data, respectively.

6. The intelligent control method for the production of catalytically uncatalyzed tert-butane according to claim 3, characterized in that, The dynamic weight of the feedforward term is expressed as wherein and is the upper limit and the lower limit of the dynamic weight preset, is the relative deviation of the hydrogen chloride concentration at the top of the column, and β is the transition steepness, is the preset transition center point.

7. The intelligent control method for the production of catalytically uncatalyzed tert-butane according to claim 2, characterized in that, Also includes: The reflux ratio at the top of the column is adjusted according to a control parameter; the reflux ratio is expressed as wherein is the control parameter, is a reference reflux ratio, and λ is a pre-calibrated mapping coefficient.

8. A method of adjusting the preheating temperature for the production of catalytic-free chlorinated tertiary butane, characterized by, include: Real-time acquisition and recording of feed flow rate and mixing preheater temperature in the reactive distillation column; The target temperature of the mixing preheater is determined based on the fluctuation of the feed flow rate. The target temperature is expressed as wherein and are respectively the upper and lower limits of the target temperature, is the mixed preheater temperature at the sampling instant k; is the relative deviation of the feed flow rate from the average flow rate at the sampling instant k, limited to a value within ±0.1; and γ is the flow rate influence coefficient.

9. A method for pre-alarming of a packing of a non-catalytic chlorinated tertiary butane production, characterized by, include: Construct a correlation model between the pressure drop in the reaction section and the operating time and feed flow rate of the reactive distillation column; Obtain pressure data at each measuring point within the reaction section, and obtain the measured pressure drop within the reaction section; Obtain the feed flow rate of the reactive distillation column and calculate the baseline pressure drop and equivalent operating time; Substitute the measured pressure drop, the baseline pressure drop, the equivalent running time, and the design running time into the correlation model to obtain the time decay exponent; Early warning for packing replacement is provided based on the time decay index.

10. The packing-based early warning method for the production of tert-butane without catalysis according to claim 9, characterized in that, The correlation model is expressed as where k represents the sampling time index, is the pressure drop of the reaction section at the sampling time k, is the designed operation time of the packing, and b is the time decay index, is the reference pressure drop at the sampling time k, is the equivalent operation time of the packing; , To design pressure drop, Let k be the feed flow rate at sampling time k. To design the feed flow rate, n is the load index; ,in The time interval between two consecutive samples is denoted as m, where m is the load aging index. , This represents the average feed flow rate.

Images