System and method for real-time optimization of fluidized bed reactor and storage medium

Abstract

The invention relates to a system and method for real-time optimization of a fluidized bed reactor and a storage medium, the system comprises a data reading module, a steady state judgment module and an optimization engine module, the data reading module is used for obtaining operation data of the fluidized bed reactor in real time; the steady state judgment module is used for performing steady state judgment on the operation data to obtain steady state data; the optimization engine module is used for acquiring a mapping relation between the steady-state data and target product variables of the fluidized bed reactor, optimizing a target function based on the mapping relation and a search algorithm, and determining an optimized operation data combination according to the obtained target function value and the mapping relation, the target function is constructed by each target product variable and the corresponding weight. Therefore, the change of the production working condition can be responded in time in the face of nonlinear and dynamic changing working conditions, and the production efficiency and the product quality are improved.

Description

Technical Field

[0001] The present application mainly relates to the technical field of organosilicon industrial process optimization, and in particular to a system, method and storage medium for real-time optimization of a fluidized bed reactor. Background Art

[0002] In the synthesis process of organosilicon monomers, the fluidized bed reactor is the core equipment, involving gas-solid two-phase flow, heat and mass transfer, and complex chemical reactions. Operating conditions such as temperature, pressure, gas flow rate and other parameters are coupled with each other, which has an important impact on the selectivity of dimethyldichlorosilane (M2), the conversion rate of methyl chloride (CH3Cl) and the production cost.

[0003] Currently, traditional control methods rely primarily on the experience of on-site operators or automated control based on distributed control systems. However, these methods struggle to capture real-time dynamic changes within the reactor, resulting in delayed control response, an inability to adjust key parameters in a timely manner, and poor repeatability and reliability.

[0004] As production scale expands and process complexity increases, traditional control methods struggle to effectively adapt to nonlinear and dynamically changing operating conditions, such as fluctuating feedstock composition, varying catalyst activity, and uneven temperature distribution within the reactor. Due to the dynamic and complex nature of organosilicon monomer synthesis, offline optimization often fails to respond promptly to changes in production conditions, resulting in delayed optimization results and impacting production efficiency and product quality. Summary of the Invention

[0005] One objective of this application is to provide a system, method, and storage medium for real-time optimization of a fluidized bed reactor. This approach addresses the problem that conventional control methods, when faced with nonlinear and dynamically changing operating conditions, are unable to effectively adapt to fluctuations in feedstock composition, changes in catalyst activity, and uneven temperature distribution within the reactor, thereby impacting production efficiency and product quality. To this end, this application proposes a solution that can stably operate and optimize the production process under dynamic conditions by combining an optimization model with a real-time control system.

[0006] According to one aspect of the present application, a system for real-time optimization of a fluidized bed reactor is provided, which is applied to the synthesis of organic silicon monomers. The system comprises: a data reading module, a steady-state judgment module, and an optimization engine module, wherein:

[0007] The data reading module is used to obtain the operating data of the fluidized bed reactor in real time;

[0008] The steady-state judgment module is used to perform steady-state judgment on the operation data to obtain steady-state data;

[0009] The optimization engine module is used to obtain the mapping relationship between steady-state data and the target product variables of the fluidized bed reactor, optimize the objective function based on the mapping relationship and the search algorithm, and determine the optimized operation data combination according to the obtained objective function value and the mapping relationship, wherein the objective function is constructed by each target product variable and the corresponding weight.

[0010] Optionally, the steady-state judgment module is used to calculate the variance of the operation data in each sampling period, select a target sampling period that satisfies the variance being less than a threshold, and use the operation data in the target sampling period as the steady-state data.

[0011] Optionally, the system includes a data filtering module for filtering the steady-state data according to a filtering formula, wherein the filtering formula satisfies the following conditions:

[0012] Y(n)=αX(n)+(1-α)Y(n-1);

[0013] Among them, α∈[0, 1] is the adjustable filter coefficient, X(n) represents the current sampling value, and Y(n-1) represents the previous filtering result.

[0014] Optionally, the system includes an outlier correction module for filtering out outliers from the steady-state data based on a limit value of a preset operating variable, and correcting the outliers using a correction method, wherein the correction method includes deleting the overall data corresponding to the outlier or replacing the outlier with a boundary value.

[0015] Optionally, the system optimization engine module has a built-in mathematical model based on a fluidized bed reactor, which is used to describe the relationship between input variables and output variables of the fluidized bed reactor.

[0016] The optimization engine module has a built-in pattern search optimization model, and the execution steps of the optimization model include: setting a direction vector for steady-state data, adjusting the steady-state data according to the step size based on the direction vector, judging the change of the objective function value, determining the recommended target product variable according to the change result, and determining the optimized operation data combination according to the recommended target product variable and the mapping relationship.

[0017] Optionally, the pattern search satisfies the following formula:

[0018] x candidate =x k ±Δ k e i (i=1,2,…,n);

[0019] d=x best -x k ;

[0020] x k+1 =x k +Δ k d k ;

[0021] Δ k+1 =βΔ k (β∈(0,1));

[0022] Among them, x candidate represents the generated candidate points, which are used to detect whether the function value is improved; x k is the current iteration point, indicating the position of the algorithm at step k; Δ k is the step size of the kth iteration, controlling the search range; e i is the standard basis vector; x best represents the optimal point found in the detection phase; d is the pattern movement direction vector, which represents the displacement direction from the current point to the optimal point; d k represents the effective direction selected in the kth iteration; β∈(0,1) is the step size attenuation factor, which controls the shrinkage speed.

[0023] Optionally, the optimization engine module includes a first input interface, a second input interface and a third input interface, wherein the first input interface is used to receive steady-state data sent by the steady-state judgment module, the second input interface is used to receive data feedback from the control system of the fluidized bed reactor, and the third input interface is used to receive the optimized operation data combination.

[0024] Optionally, the optimization engine module also includes a first output interface, a second output interface, a third output interface and an algorithm component, wherein the first output interface is used to send the optimized operation data combination to the downstream components of the system, the second output interface is used to send the optimized operation data combination to the front end of the control system of the fluidized bed reactor, and the third output interface is used to provide steady-state data to the algorithm component, and the algorithm component is used to execute the optimization process according to the steady-state data to obtain the optimized operation data combination.

[0025] Optionally, the system includes a data writing module for writing the optimized operation data combination into the control system of the fluidized bed reactor to perform control.

[0026] According to another aspect of the present application, a method for real-time optimization of a fluidized bed reactor is provided, which is applied to the synthesis of organosilicon monomers. The method comprises:

[0027] Real-time acquisition of fluidized bed reactor operating data;

[0028] Performing steady-state judgment on the operation data to obtain steady-state data;

[0029] Obtain a mapping relationship between steady-state data and target product variables of a fluidized bed reactor, optimize an objective function based on the mapping relationship and a search algorithm, and determine an optimized operating data combination according to the obtained objective function value and the mapping relationship, wherein the objective function is constructed by each target product variable and the corresponding weight.

[0030] According to another aspect of the present application, a device for predicting and optimizing a fluidized bed reactor is provided, the device comprising:

[0031] one or more processors; and

[0032] A memory storing computer-readable instructions, which, when executed, cause the processor to perform the operations of the method described above.

[0033] According to another aspect of the present application, a computer-readable medium is provided, on which computer instructions are stored. The computer-readable instructions can be executed by a processor to implement the method described above.

[0034] Compared to the prior art, the present application provides a system for real-time optimization of a fluidized bed reactor, which is applied to the synthesis of organic silicon monomers. The system comprises: a data reading module, a steady-state judgment module, and an optimization engine module. The data reading module is used to obtain the operating data of the fluidized bed reactor in real time; the steady-state judgment module is used to perform steady-state judgment on the operating data to obtain steady-state data; the optimization engine module is used to obtain a mapping relationship between the steady-state data and the target product variables of the fluidized bed reactor, optimize the objective function based on the mapping relationship and a search algorithm, and determine the optimized operating data combination based on the obtained objective function value and the mapping relationship. The objective function is constructed from each target product variable and its corresponding weight. Thus, when faced with nonlinear and dynamically changing operating conditions, the system can promptly respond to changes in production conditions, thereby improving production efficiency and product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to make the above-mentioned objects, features and advantages of the present application more clearly understood, the specific embodiments of the present application are described in detail below with reference to the accompanying drawings, wherein:

[0036] Figure 1 A schematic diagram of a system structure for real-time optimization of a fluidized bed reactor according to one aspect of the present application is shown;

[0037] Figure 2 A schematic diagram showing temperature data of a fluidized bed reactor in one embodiment of the present application is shown;

[0038] Figure 3 Schematic diagram showing pressure data of a fluidized bed reactor in one embodiment of the present application

[0039] Figure 4 Schematic diagram showing the optimization effect of M2 selectivity in one embodiment of the present application;

[0040] Figure 5 A schematic flow chart of a method for real-time optimization of a fluidized bed reactor according to another aspect of the present application is shown;

[0041] Figure 6 A system block diagram showing an apparatus for real-time optimization of a fluidized bed reactor according to an embodiment of the present application.

[0042] The same or similar reference numerals in the drawings represent the same or similar components. DETAILED DESCRIPTION

[0043] In order to make the above-mentioned objectives, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are described in detail below with reference to the accompanying drawings.

[0044] In the following description, many specific details are set forth to facilitate a full understanding of the present application. However, the present application may also be implemented in other ways different from those described herein. Therefore, the present application is not limited to the specific embodiments disclosed below.

[0045] As used in this application and the claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not intended to refer to the singular but may include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0046] Figure 1A schematic diagram of a system structure for real-time optimization of a fluidized bed reactor provided according to one aspect of the present application is shown, which is applied to the synthesis of organic silicon monomers. The system includes: a data reading module 100, a steady-state judgment module 200, and an optimization engine module 300, wherein the data reading module 100 is used to obtain the operating data of the fluidized bed reactor in real time; the steady-state judgment module 200 is used to perform steady-state judgment on the operating data to obtain steady-state data; the optimization engine module 300 is used to obtain a mapping relationship between the steady-state data and the target product variables of the fluidized bed reactor, optimize the objective function based on the mapping relationship and the search algorithm, and determine the optimized operating data combination based on the obtained objective function value and the mapping relationship, wherein the objective function is constructed by each target product variable and the corresponding weight. Among them, the optimization engine module 300 serves as the execution core of the optimization task, calls the optimization algorithm to optimize the output variable M2 selectivity and methyl chloride conversion rate, and also includes a mathematical model of the fluidized bed reactor, establishes a mathematical relationship between the input variables and the output variables, and establishes a nonlinear mapping relationship between the input variables and the output variables through historical data.

[0047] The data reading module 100 is connected to the on-site PLC point and reads the operating data of the fluidized bed reactor from the distributed control system (DCS) in real time. The operating data is some key data, such as temperature, pressure, catalyst dosage, methyl chloride feed amount and purity, etc. Figure 2 The temperature data shown, such as Figure 3 The DCS is the fluidized bed reactor's control system, used to control the reactor's operating parameters for organosilicon monomer synthesis. The steady-state determination module 200 is used to perform steady-state detection on key data during the production and synthesis of organosilicon monomers in the fluidized bed reactor, determine whether the data read by the data reading module 100 has entered a stable operating state, obtain steady-state data, and then send the steady-state data to downstream modules.

[0048] The optimization engine module 300 serves as the entry point for optimization task management and uses an optimization model to calculate the maximum value of the objective function. The optimization model is constructed based on a mathematical model of the reactor's input and output variables combined with a pattern search algorithm. The input variables include temperature, pressure, catalyst dosage, etc., and the output variables are target product variables, such as M2 selectivity and chloromethane conversion. This mathematical model is established using a data-driven algorithm that establishes a nonlinear mapping relationship between the reactor input variables and the target product variables based on historical operating data. The system can periodically perform optimization calculations based on real-time production data using the optimization model integrated in the optimization engine module. The optimization model predicts the values of the output variables based on the mapping function fitted by the mathematical model and provides recommended values for the operating variables to optimize the target product variables. The objective function is constructed by maximizing the weighted sum of the target product variables, where each target product variable has a corresponding weight, such as M2 selectivity and chloromethane conversion. Users can flexibly adjust the target product weights based on production needs to achieve a balance between different optimization objectives. For example, in some cases, it may be necessary to prioritize improving M2 selectivity, while in other cases, more emphasis may be placed on chloromethane conversion. Through weight adjustment, the optimization engine module can accurately control the operating conditions of the reactor, ensuring that the production process achieves the expected target product selectivity and conversion rate, and thus obtain the optimal operating data combination of the fluidized bed reactor.

[0049] Specifically, the optimization model is invoked and a search algorithm is applied to solve the objective function for the optimization task. The input variable values that maximize the target product are determined within the physical range of the input variables, which is determined based on equipment safety or reaction stability. This objective function is constructed based on the target product variables of the fluidized bed reactor, including M2 selectivity and methyl chloride conversion. Each target product variable and its corresponding weight are used to construct the objective function. The search algorithm uses pattern search to optimize the input variables to maximize the performance of the target product. The pattern search algorithm gradually adjusts the operating variables to explore potential optimal points in the solution space. The algorithm first searches within the neighborhood of the current solution and decides whether to accept the new solution based on the change in the objective function. If the current adjustment improves the objective function value, the solution is accepted and the search continues with the new operating variables. If not, the search direction is changed or the step size is reduced, and the search is repeated until the optimal solution that maximizes the target product performance is found. This search process not only avoids the dilemma of local optimal solutions but also dynamically adjusts the search strategy based on real-time data, thereby maintaining the system's optimization effect under changing production conditions. During execution, the optimization engine module 300 will collaborate with the steady-state judgment module 200 to ensure that the optimization calculation is performed only when the reactor is in a steady state. The use of steady-state data effectively avoids the interference of short-term fluctuations on the optimization calculation results, making the optimization results more stable and reliable. The optimization engine module, in combination with historical data and real-time monitoring data, can continuously adjust the operating conditions to adapt to the changes in the demand of the reactor in different production stages, ensuring the continuous improvement of production efficiency. After optimizing the objective function, the recommended objective function value, i.e. the optimized target product variable, is obtained. Then, according to the mapping relationship, the corresponding operating data combination can be obtained, and then the required optimized operating data combination is obtained, and then the optimized operating data combination is used to be input into the control system of the fluidized bed reactor, and the control of the fluidized bed reactor is executed. In an embodiment of the present application, different optimization schemes or adjustment strategies can be flexibly selected according to production needs, and integrated into the optimization engine module, wherein the user can flexibly select algorithms such as pattern search or composite differential evolution according to the characteristics of specific problems such as field temperature and pressure data and M2 selectivity and methyl chloride conversion, to achieve the optimal solution for controlling the fluidized bed reactor of the organic silicon monomer synthesis process.

[0050] In some embodiments of the present application, the data reading module 100 is used to connect to the on-site PLC point to read the data. Specifically, in order to enable DCS to read the operating data of the fluidized bed reactor in real time, the channel, access and some key data, such as temperature, pressure, catalyst dosage, methyl chloride feed amount and purity, should first be connected and established in accordance with the Open Platform Communications-Data Access Protocol (OPC-DA). Specifically, the channel is configured by aligning the unique channel name and proxy IP address; the access is configured by aligning the unique device name, the IP address and name of the OPC server; and the configuration of each point is achieved by aligning the data item address of each point. After the configuration is completed, the operating data of the fluidized bed reactor read from the DCS can be sent to the downstream module.

[0051] In some embodiments of the present application, the steady-state determination module 200 is configured to perform a steady-state detection on the operating data within each sampling period to determine whether the reactor has entered a stable operating condition. Specifically, the steady-state determination module 200 is configured to calculate the variance of the operating data within each sampling period, select a target sampling period in which the variance is less than a threshold, and use the operating data within the target sampling period as the steady-state data. Here, the calculation of the variance of the operating data is used to determine whether the data meets the steady-state condition. First, the module calculates the variance of the operating data within each sampling period and compares it with a preset threshold. When the variance is less than the threshold, the data within that sampling period is considered to meet the steady-state condition and has entered a stable operating condition. In this embodiment, the steady-state determination module 200 analyzes data from multiple consecutive sampling periods using a sliding window technique. Specifically, the module sets a sliding window of length N (e.g., N=50) and calculates the variance of the data within each window. If the variance within a window is less than the preset threshold, the data in that window is considered to meet the steady-state condition. If the variance within M consecutive sliding windows meets the condition of being less than the threshold, the reactor is considered to have entered a steady-state operating condition. At this time, the steady-state judgment module 200 uses the window data entering the steady-state working condition as the target sampling period, and uses the data within the target sampling period as the steady-state data. The use of steady-state data can effectively reduce the interference of short-term fluctuations in the production process on the optimization calculation results, thereby improving the stability and feasibility of the optimization calculation. For example, when the fluctuation of temperature data is controlled within 0.5°C, or the standard deviation of pressure data is less than 0.01MPa, the data can be judged as steady-state and suitable for subsequent optimization calculations. For sampling periods that do not meet the steady-state conditions, the system will automatically eliminate these data to ensure that the optimization calculation is based on continuous data that meets the steady-state conditions.

[0052] In one embodiment of the present application, the system includes a data filtering module 400 for filtering the steady-state data according to a filtering formula. Here, the data filtering module 400 smoothes the steady-state data filtered by the steady-state judgment module 200 to reduce interference caused by noise or transient fluctuations, thereby improving the accuracy and stability of subsequent optimization calculations. In this embodiment of the present application, a first-order filtering method is used, wherein the filtering formula of the first-order filter satisfies the following conditions:

[0053] Y(n)=αX(n)+(1-α)Y(n-1);

[0054] Where α∈[0,1] is an adjustable filter coefficient, X(n) represents the current sample value, and Y(n-1) represents the previous filter result. The selection of the filter coefficient α has a significant impact on the smoothness of the filter result. In this embodiment, the user can set the value of the filter coefficient α according to the specific production conditions to meet the filtering effect with different accuracy requirements. For data such as temperature and pressure, if the data fluctuates greatly, the user can reduce the α value and increase the reliance on historical filtering results to make the data smoother; if the data fluctuates less, the user can appropriately increase α to speed up the data response while maintaining the filtering effect.

[0055] In one embodiment of the present application, the system includes an outlier correction module 500, which is used to filter out outliers from the steady-state data based on the limit values of the preset operating variables, and correct the outliers using a correction method, wherein the correction method includes out-of-bounds elimination or out-of-bounds replacement, that is, deleting the overall data corresponding to the outlier or replacing the outlier with a boundary value; ensuring that in the production process, abnormal data generated due to equipment failure, sensor error or external interference will not affect the accuracy and stability of the optimization calculation. The limit values of the preset operating variables include upper and lower limits. The module first filters the steady-state data according to the preset limit values, compares the relationship between the steady-state data and the preset limit values, and then determines whether the data is abnormal. If the data exceeds the limit range, the data is considered to be an outlier and needs to be corrected. Specifically, for temperature data and pressure data, the system provides two outlier correction strategies: out-of-bounds elimination and out-of-bounds replacement. The out-of-bounds elimination strategy is that when the temperature data T satisfies T>T max or T <T min When the pressure data P satisfies P <P min Or P>P max When , the system will regard the data point as an outlier and delete it. This strategy is suitable for those serious abnormal data that cannot be corrected. By deleting these data, they will not affect the subsequent optimization calculations. The strategy of out-of-bounds replacement is that when the temperature data T satisfies T>T max When forced to correct to T = T max; When the temperature data T meets T <T min When forced to correct to T = T min ; When the pressure data P meets P <P min When forced to correct to P = P min ; When the pressure data P satisfies P>P max When forced to correct to P = P max This strategy is suitable for data with slight abnormal fluctuations. By correcting these outliers to reasonable boundary values, it ensures that the data can still be used for optimization calculations without losing valid information. By monitoring data input and optimization results in real time, the system can automatically identify and eliminate abnormal data to ensure the accuracy of optimization calculations.

[0056] In one embodiment of the present application, the optimization engine module 300 includes a built-in mathematical model based on a fluidized bed reactor. This model uses historical data to establish a mapping relationship between operating variables and optimization targets, describing the relationship between the input and output variables of the fluidized bed reactor. The system includes a mathematical model based on a fluidized bed reactor. Based on historical operating variables such as temperature, pressure, and catalyst dosage, the system establishes a relationship between operating variables and output variables such as M2 selectivity and methyl chloride conversion. This relationship between variables creates an optimization model. This optimization model can be a pattern search-based optimization model, specifically:

[0057] The optimization engine module 300 is used to set a direction vector for the steady-state data, adjust the steady-state data according to the step size based on the direction vector, judge the change of the objective function value, determine the recommended target product variable based on the change result, and determine the optimized operation data combination based on the recommended target product variable and the mapping relationship. Here, a pattern search algorithm can be used for parameter optimization. By setting a set of direction vectors near the current operating point, tentative adjustments are made to each operating variable with a preset amplitude. The preset amplitude is a small step size, and the preset amplitude does not exceed 5% of the adjustable range. For example, 0.1 or one percent or five percent of the current value is used as the step size for tentative adjustment. If the adjustment in a certain direction can improve the objective function value, such as improving the selectivity or conversion rate, the operating point is accepted and the step size is appropriately adjusted, and the iterative optimization is continued. That is, the objective function value is improved, and the corresponding input operating point, that is, the operating point, can be obtained by using the mapping relationship. If there is no improvement, the step size is reduced and re-explored until the termination condition is met, and a recommended value is given. The termination condition is that the search step size is less than the set threshold. The step size is gradually reduced as the optimization process progresses. When the reduced step size is less than the set threshold, the search is stopped, and the recommended value searched before stopping the search is given. The recommended value is the recommended operating data combination. Figure 4The optimization effect diagram of M2 selectivity is shown; the parameters of the pattern search algorithm are optimized for each operation data to obtain a set of recommended values, that is, the optimized operation data combination.

[0058] The optimization model of pattern search is to make a tentative adjustment to the operating variable by a small adjustment value of ±1% of the current value according to a selected set of direction vectors, and transmit the adjusted operating variable value to the mathematical model to obtain the adjusted output value. The core of the optimization model of pattern search includes four main steps: coordinate exploration, pattern movement, iterative update and step size adjustment. First, coordinate exploration generates candidate points to find the direction of local improvement, thereby optimizing the objective function value. Secondly, pattern movement uses successful directions to accelerate the search and increase the convergence speed of the search. The iterative update stage further optimizes the search strategy through flexible adjustment of the direction set to find the optimal solution more efficiently. Finally, step size adjustment ensures the convergence and stability of the optimization process. The above steps are in accordance with the following formula:

[0059] x candidate =x k ±Δ k e i (i=1,2,…,n);

[0060] d=x best -x k ;

[0061] x k+1 =x k +Δ k d k ;

[0062] Δ k+1 =βΔ k (β∈(0,1));

[0063] where x k is the current iteration point, and the step size of the kth iteration is Δ k (step length>0), used to control the search range, d is the pattern movement direction vector, indicating the displacement direction from the current point to the optimal point, and the effective direction selected in the kth iteration is d k , the optimal point is represented by x best Specifically, the candidate point x generated by coordinate exploration candidate Used to detect changes in the objective function value; if the current iteration point x k Indicates the position of the search algorithm at step k. If the adjustment can bring about an improvement in the objective function value, the search will continue; if there is no improvement, the step size will be gradually reduced and a new direction will be explored again; e i is the standard basis vector; x bestrepresents the optimal point found in the exploration phase. The step size attenuation factor β controls the speed at which the step size shrinks. Its value is in the range of (0,1), ensuring that the global optimal solution is gradually approached during the optimization process.

[0064] In one embodiment of the present application, the optimization engine module 300 includes a first input interface, a second input interface, and a third input interface, wherein the first input interface is used to receive the steady-state data sent by the steady-state judgment module, the second input interface is used to receive data fed back by the control system of the fluidized bed reactor, and the third input interface is used to receive the optimized operation data combination. Based on the data received by the first input interface, the optimization engine module 300 calls the optimization model and uses a pattern search algorithm to optimize the parameters of the operating variables. Here, the optimization engine module serves as the optimization task management entrance and includes three input interfaces, namely in1, in2, and in3, wherein in1 receives point data, that is, receives the steady-state data sent; in2 receives the reverse data of the front-end configuration system, which is the control system of the fluidized bed, that is, the distributed control system DCS. in3 receives the final optimization result of optimizing the objective function using steady-state data.

[0065] Continuing with the above embodiment, the optimization engine module 300 further includes a first output interface, a second output interface, a third output interface, and an algorithm component, wherein the first output interface is used to send the optimized operating data combination to the downstream components of the system, the second output interface is used to send the optimized operating data combination to the front end of the fluidized bed reactor control system, and the third output interface is used to provide steady-state data to the algorithm component, and the algorithm component is used to perform the optimization process based on the steady-state data to obtain the optimized operating data combination. Here, the optimization engine module also includes three output interfaces and an algorithm component. The first output interface out1 is used to output the optimization results to the downstream components after the optimization is completed, wherein the downstream components are components that require the optimization results. The second output interface out2 is used to output the optimization results to the front end of the DCS system for display. The third output interface out3 is connected to the algorithm component and sends data and parameters to the algorithm component. The algorithm component then uses the received data to optimize the objective function using a pattern search algorithm, executes the optimization process, constructs multi-directional tentative adjustments based on key operating variables under the current operating conditions, iteratively searches for variable combinations that can improve the objective function, and generates recommended operating parameter combinations based on the found variable combinations.

[0066] In one embodiment of the present application, the system includes a data writing module 600 for writing the optimized operational data combination to the fluidized bed reactor control system for control. Here, the optimized recommended values are output to the distributed control system (DCS) for display to the operator, along with specific adjustment suggestions. The operator then adjusts the corresponding operating parameters of the fluidized bed reactor based on actual conditions, thereby controlling the reactor to synthesize the organosilicon monomer and improve the selectivity and conversion rate of the target product.

[0067] Figure 5 A schematic flow chart of a method for real-time optimization of a fluidized bed reactor provided according to another aspect of the present application is shown, which is applied to the synthesis of organosilicon monomers. The method comprises: steps S11 to S13, wherein, in step S11, operating data of the fluidized bed reactor is acquired in real time; step S12, steady-state determination is performed on the operating data to obtain steady-state data; step S13, a mapping relationship between the steady-state data and the target product variables of the fluidized bed reactor is acquired, an objective function is optimized based on the mapping relationship and a search algorithm, and an optimized operating data combination is determined based on the obtained objective function value and the mapping relationship, wherein the objective function is constructed from each target product variable and its corresponding weight. Here, the operating data of the fluidized bed reactor is read in real time. The operating data is some key data, such as temperature, pressure, catalyst dosage, methyl chloride feed amount and purity. Next, steady-state detection is performed on the key data in the process of producing and synthesizing organosilicon monomers in the fluidized bed reactor. For example, steady-state detection is performed on a data sequence of N consecutive sampling periods to determine whether the read data has entered a stable operating condition, that is, when the variance is less than a threshold, it is determined to be steady-state, and steady-state data is obtained. A first-order filtering process is performed on the steady-state data, and outliers are screened out from the steady-state data based on the limit values of the preset operating variables. The outliers are corrected using a correction method, wherein the correction method includes out-of-bounds elimination or out-of-bounds replacement, that is, deleting the entire data corresponding to the outlier or replacing the outlier with a boundary value; ensuring that in the production process, outliers generated due to equipment failure, sensor error, or external interference will not affect the accuracy and stability of the optimization calculation. The steady-state data is used to solve the objective function using a search algorithm. The search algorithm can use a pattern search algorithm. The objective function is constructed based on the target product variables of the fluidized bed reactor. The target product variables include M2 selectivity, methyl chloride conversion rate, etc. The optimized operating data combination is obtained based on the solved values and the mapping relationship. The optimized operating data combination is then used to input into the control system of the fluidized bed reactor to execute the control of the fluidized bed reactor. Thus, the operating conditions are adjusted in time according to the operating state of the reactor to improve the target product, such as improving the selectivity of dimethyldichlorosilane and the conversion rate of methyl chloride, and ensuring the stability and reliability of the optimization process.

[0068] Figure 6 A system block diagram of an apparatus for real-time optimization of a fluidized bed reactor according to an embodiment of the present application is shown. Figure 6 As shown, the apparatus 600 for real-time optimization of a fluidized bed reactor may include an internal communication bus 601, a processor 602, a read-only memory (ROM) 603, a random access memory (RAM) 604, a communication port 605, and a hard disk 607. The internal communication bus 601 enables data communication between the components of the apparatus for real-time optimization of a fluidized bed reactor. The processor 602 may make judgments and issue prompts. In some embodiments, the processor 602 may be composed of one or more processors.

[0069] Communication port 605 enables data transmission and communication between the apparatus for real-time optimization of a fluidized bed reactor and external input / output devices. In some embodiments, the apparatus for real-time optimization of a fluidized bed reactor can send and receive information and data from a network via communication port 605. In some embodiments, the apparatus for real-time optimization of a fluidized bed reactor can transmit and communicate data with external input / output devices via a wired connection via input / output port 606.

[0070] The apparatus for real-time optimization of a fluidized bed reactor may also include various forms of program storage units and data storage units, such as a hard disk 607, a read-only memory (ROM) 603, and a random access memory (RAM) 604, capable of storing various data files used for computer processing and / or communication, as well as possible program instructions executed by the processor 602. The processor 602 executes these instructions to implement the main part of the method. The results of the processing by the processor 602 are transmitted to an external output device via a communication port 605 and displayed on a user interface of the output device.

[0071] For example, the implementation process file of the above-mentioned apparatus for real-time optimization of a fluidized bed reactor may be a computer program, which is stored in the hard disk 607 and can be recorded in the processor 602 for execution to implement the method of the present application.

[0072] The present application also provides a computer-readable medium having computer instructions stored thereon, wherein the computer-readable instructions can be executed by a processor to implement the method for real-time optimization of a fluidized bed reactor as described above.

[0073] When the method for real-time optimization of a fluidized bed reactor is implemented as a computer program, it can also be stored in a computer-readable storage medium as an article of manufacture. For example, a computer-readable storage medium can include, but is not limited to, magnetic storage devices (e.g., hard disks, floppy disks, magnetic strips), optical disks (e.g., compact disks (CDs), digital versatile disks (DVDs)), smart cards, and flash memory devices (e.g., electrically erasable programmable read-only memories (EPROMs), cards, sticks, key drives). In addition, the various storage media described herein can represent one or more devices and / or other machine-readable media for storing information. The term "machine-readable medium" can include, but is not limited to, wireless channels and various other media (and / or storage media) that can store, contain, and / or carry code and / or instructions and / or data.

[0074] It should be understood that the embodiments described above are merely illustrative. The embodiments described herein may be implemented in hardware, software, firmware, middleware, microcode, or any combination thereof. For hardware implementation, the processor may be implemented within one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, and / or other electronic units designed to perform the functions described herein, or a combination thereof.

[0075] Some aspects of the present application can be performed entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The above hardware or software can be referred to as "data blocks", "modules", "engines", "units", "components" or "systems". The processor can be one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DAPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors or combinations thereof. In addition, various aspects of the present application may be expressed as computer products located in one or more computer-readable media, which include computer-readable program code. For example, computer-readable media may include, but are not limited to, magnetic storage devices (e.g., hard disks, floppy disks, tapes...), optical disks (e.g., compact disks CDs, digital versatile disks DVDs...), smart cards, and flash memory devices (e.g., cards, sticks, key drives...).

[0076] A computer-readable medium may include a propagated data signal embodying computer program code, for example, in baseband or as part of a carrier wave. The propagated signal may be in a variety of forms, including electromagnetic, optical, etc., or a suitable combination thereof. A computer-readable medium may be any computer-readable medium other than a computer-readable storage medium that can be connected to an instruction execution system, apparatus, or device to communicate, propagate, or transmit the program for use. The program code on the computer-readable medium may be transmitted via any suitable medium, including radio, cable, fiber optic cable, radio frequency signal, or similar medium, or any combination of the above.

[0077] The basic concepts have been described above. It will be apparent to those skilled in the art that the above disclosures are merely illustrative and do not constitute limitations on this application. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and amendments to this application. Such modifications, improvements, and amendments are suggested in this application and remain within the spirit and scope of the exemplary embodiments of this application.

[0078] At the same time, this application uses specific terms to describe the embodiments of this application. For example, "one embodiment," "an embodiment," and / or "some embodiments" refer to a certain feature, structure, or characteristic related to at least one embodiment of this application. Therefore, it should be emphasized and noted that "one embodiment," "an embodiment," or "an alternative embodiment" mentioned twice or multiple times in different locations in this specification does not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of this application may be appropriately combined.

[0079] In some embodiments, numbers are used to describe the quantity of components and attributes. It should be understood that such numbers used in the description of the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples. Unless otherwise stated, "about", "approximately" or "substantially" indicate that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the description and claims are approximate values, which may change according to the required features of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining digits. Although the numerical domains and parameters used to confirm the breadth of their range in some embodiments of the present application are approximate values, in specific embodiments, the settings of such numerical values are as accurate as possible within the feasible range.

Claims

1. A system for real-time optimization of fluidized bed reactors, applied to organosilicon monomer synthesis, characterized in that: The system includes: a data reading module, a steady-state judgment module and an optimization engine module, wherein: The data reading module is used to obtain the operating data of the fluidized bed reactor in real time; The steady-state judgment module is used to perform steady-state judgment on the operation data to obtain steady-state data; The optimization engine module is used to obtain the mapping relationship between steady-state data and the target product variables of the fluidized bed reactor, optimize the objective function based on the mapping relationship and the search algorithm, and determine the optimized operation data combination according to the obtained objective function value and the mapping relationship, wherein the objective function is constructed by each target product variable and the corresponding weight.

2. The system according to claim 1, wherein: The steady-state judgment module is used to calculate the variance of the operation data in each sampling period, select a target sampling period that satisfies the variance less than a threshold, and use the operation data in the target sampling period as steady-state data.

3. The system according to claim 1, wherein: The system includes a data filtering module for filtering the steady-state data according to a filtering formula, wherein the filtering formula satisfies the following conditions: Y(n)=αX(n)+(1-α)Y(n-1); Among them, α∈[0, 1] is the adjustable filter coefficient, X(n) represents the current sampling value, and Y(n-1) represents the previous filtering result.

4. The system according to claim 1, wherein: The system includes an outlier correction module for filtering out outliers from the steady-state data based on a limit value of a preset operating variable, and correcting the outliers using a correction method, wherein the correction method includes deleting the overall data corresponding to the outlier or replacing the outlier with a boundary value.

5. The system according to claim 1, wherein: The optimization engine module has a built-in mathematical model based on a fluidized bed reactor, which establishes a mapping relationship between operating variables and optimization objects through historical data and describes the relationship between input variables and output variables of the fluidized bed reactor.

6. The system according to claim 1, wherein: The optimization engine module has a built-in pattern search optimization model, and the execution steps of the optimization model include: setting a direction vector for steady-state data, adjusting the steady-state data according to the step size based on the direction vector, judging the change of the objective function value, determining the recommended target product variable according to the change result, and determining the optimized operation data combination according to the recommended target product variable and the mapping relationship.

7. The system according to claim 6, characterized in that The pattern search satisfies the following formula: x candidate =x k ±Δ k e i (i=1,2,…,n); d=x best -x k ; x k+1 =x k +Δ k d k ; D k+1 =bD k (β∈(0,1)); Among them, x candidate represents the generated candidate points, which are used to detect whether the function value is improved; x k is the current iteration point, indicating the position of the algorithm at step k; Δ k is the step size of the kth iteration, controlling the search range; e i is the standard basis vector; x best represents the optimal point found in the detection phase; d is the pattern movement direction vector, which represents the displacement direction from the current point to the optimal point; d k represents the effective direction selected in the kth iteration; β∈(0,1) is the step size attenuation factor, which controls the shrinkage speed.

8. The system according to claim 1, wherein: The optimization engine module includes a first input interface, a second input interface and a third input interface, wherein the first input interface is used to receive steady-state data issued by the steady-state judgment module, the second input interface is used to receive data feedback from the control system of the fluidized bed reactor, and the third input interface is used to receive the optimized operation data combination.

9. The system according to claim 1, wherein: The optimization engine module also includes a first output interface, a second output interface, a third output interface and an algorithm component, wherein the first output interface is used to send the optimized operation data combination to the downstream components of the system, the second output interface is used to send the optimized operation data combination to the front end of the control system of the fluidized bed reactor, and the third output interface is used to provide steady-state data to the algorithm component, and the algorithm component is used to execute the optimization process according to the steady-state data to obtain the optimized operation data combination.

10. The system according to claim 1, wherein: The system includes a data writing module for writing the optimized operation data combination into the control system of the fluidized bed reactor to perform control.

11. A method for real-time optimization of a fluidized bed reactor, applied to organosilicon monomer synthesis, characterized in that: The method comprises: Real-time acquisition of fluidized bed reactor operating data; Performing steady-state judgment on the operation data to obtain steady-state data; Obtain a mapping relationship between steady-state data and target product variables of a fluidized bed reactor, optimize an objective function based on the mapping relationship and a search algorithm, and determine an optimized operating data combination according to the obtained objective function value and the mapping relationship, wherein the objective function is constructed by each target product variable and the corresponding weight.

12. A device for prediction and optimization of a fluidized bed reactor, characterized in that: The device comprises: one or more processors; and A memory storing computer readable instructions that, when executed, cause the processor to perform the operations of the method of claim 11.

13. A computer-readable medium having computer instructions stored thereon, wherein the computer-readable instructions can be executed by a processor to implement the method according to claim 11.

Images