Phosphine reaction kettle control parameter optimization method, device, equipment and storage medium

Abstract

The application provides a phosphine reaction kettle control parameter optimization method and device, equipment and storage medium, and belongs to the automatic control technical field. The method comprises the following steps: acquiring a temperature limit interval, a pressure limit interval and a stirring speed limit interval of a reaction kettle; randomly initializing a plurality of population individuals based on the temperature limit interval, the pressure limit interval and the stirring speed limit interval to obtain an initialized population; wherein each population individual is used for representing the value of a group of control parameters, and the group of control parameters comprises a temperature control parameter, a pressure control parameter and a stirring speed control parameter of the reaction kettle; performing a plurality of population iteration operations on the initialized population by using a genetic algorithm until a stop condition is met to obtain an optimal population individual; and optimizing the control parameters of the reaction kettle based on the optimal population individual. The phosphine reaction kettle control parameter optimization method and device, equipment and storage medium provided by the application can improve the production efficiency and operation safety of the reaction kettle.

Description

Technical Field

[0001] This application belongs to the field of automatic control technology, and more specifically, relates to a method, apparatus, equipment, and storage medium for optimizing control parameters of a phosphine reactor. Background Technology

[0002] The phosphine reactor is the core equipment used for the preparation and treatment of phosphine (PH3). Its working process includes: first, adding raw materials such as yellow phosphorus, sodium hydroxide, and lime milk into the reactor, and then introducing nitrogen gas to replace the air and isolate oxygen; then, initiating the reaction by heating (controlling the temperature) and stirring (controlling the speed) (e.g., P4 + 3NaOH + 3H2O → 3NaH2PO2 + PH3↑); during the reaction, the pressure needs to be controlled, and the phosphine generation rate is balanced by the exhaust valve to avoid overpressure; the generated phosphine is collected after purification, and the residue and waste liquid are discharged after treatment.

[0003] The coordinated optimization of control parameters such as temperature, pressure, and stirring speed is a key aspect of phosphine reactor control. Traditional trial-and-error methods rely on repeated manual adjustments based on human experience, which is inefficient and easily subject to subjective limitations. Local optimization methods (such as gradient descent) search for local optima from the initial point, which can easily lead to local extrema and are difficult to handle multi-parameter coupled scenarios, resulting in limited optimization accuracy.

[0004] Therefore, there is an urgent need to propose a more efficient method for the coordinated optimization of control parameters such as reactor temperature, pressure, and stirring speed. Summary of the Invention

[0005] The purpose of this application is to provide a method, apparatus, equipment, and storage medium for optimizing the control parameters of a phosphine reactor, so as to improve the production efficiency and operational safety of the reactor.

[0006] A first aspect of this application provides a method for optimizing control parameters of a phosphine reactor, comprising:

[0007] Obtain the temperature limit range, pressure limit range, and stirring speed limit range for the reactor;

[0008] Multiple population individuals are randomly initialized based on the temperature limit range, pressure limit range, and stirring speed limit range to obtain an initial population; wherein, each population individual is used to represent the value of a set of control parameters, which includes the temperature control parameters, pressure control parameters, and stirring speed control parameters of the reactor;

[0009] The initial population is iterated multiple times using a genetic algorithm until the stopping condition is met, and the optimal population individual is obtained.

[0010] The control parameters of the reactor are optimized based on the optimal population individuals.

[0011] Each population iteration operation includes:

[0012] The peak phosphine yield and safety risk assessment value for each individual in the initial population are determined based on the values ​​of the control parameters corresponding to that individual.

[0013] The fitness value of each individual in the population is determined based on the peak phosphine yield and safety risk assessment value.

[0014] Multiple parent individuals are selected from multiple populations based on the fitness values ​​of individuals in each population; the parent individuals are populations whose corresponding fitness values ​​are greater than the fitness threshold.

[0015] The initial population is updated based on multiple parent individuals;

[0016] The stopping condition is that the number of iterations reaches a set number, or the change in the maximum fitness value obtained in N consecutive iterations is less than a first threshold.

[0017] The optimal population individual is the one with the highest fitness value among the multiple population individuals obtained in the last iteration.

[0018] A second aspect of this application provides a device for optimizing control parameters of a phosphine reactor, comprising:

[0019] The data acquisition module is used to acquire the temperature limit range, pressure limit range, and stirring speed limit range of the reactor.

[0020] The population initialization module is used to randomly initialize multiple population individuals based on the temperature limit range, pressure limit range, and stirring speed limit range to obtain an initialized population; wherein, each population individual is used to represent the value of a set of control parameters, the set of control parameters including the temperature control parameters, pressure control parameters, and stirring speed control parameters of the reactor;

[0021] The population optimization module is used to perform multiple population iteration operations on the initial population using a genetic algorithm until the stopping condition is met, thereby obtaining the optimal population individuals.

[0022] The control parameter optimization module is used to optimize the control parameters of the reactor based on the optimal population individuals;

[0023] In any population iteration operation, the population optimization module is used for:

[0024] The peak phosphine yield and safety risk assessment value for each individual in the initial population are determined based on the values ​​of the control parameters corresponding to that individual.

[0025] The fitness value of each individual in the population is determined based on the peak phosphine yield and safety risk assessment value.

[0026] Multiple parent individuals are selected from multiple populations based on the fitness values ​​of individuals in each population; the parent individuals are populations whose corresponding fitness values ​​are greater than the fitness threshold.

[0027] The initial population is updated based on multiple parent individuals;

[0028] The stopping condition is that the number of iterations reaches a set number, or the change in the maximum fitness value obtained in N consecutive iterations is less than a first threshold.

[0029] The optimal population individual is the one with the highest fitness value among the multiple population individuals obtained in the last iteration.

[0030] A third aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the above-described method for optimizing control parameters of a phosphine reactor.

[0031] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method for optimizing control parameters of a phosphine reactor.

[0032] The beneficial effects of the phosphine reactor control parameter optimization method, apparatus, equipment, and storage medium provided in this application embodiment are as follows:

[0033] This application uses a genetic algorithm as an optimization tool. Through population iteration, it can search for the global optimal solution within the safe operating range of each control parameter. Compared with traditional trial and error methods or local optimization methods, it can improve the optimization accuracy.

[0034] Based on this, during each population iteration, the fitness value of each individual in the population is determined based on its peak phosphine yield and safety risk assessment value. This fitness value serves as a comprehensive indicator for judging the quality of individuals in the population. The control parameters of the reactor are then optimized based on the values ​​of a set of control parameters corresponding to the optimal individual in the final population. This approach ensures both a high phosphine yield and keeps safety risks at a low level. Compared to existing single-objective control methods (such as focusing solely on high yield or high safety), this approach avoids safety accidents (such as overpressure explosion risks) caused by solely pursuing high yields, or low yields due to excessive conservatism, thus achieving a balance between production efficiency and operational safety. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 A schematic flowchart illustrating a method for optimizing control parameters of a phosphine reactor according to an embodiment of this application;

[0037] Figure 2 This is a structural block diagram of a phosphine reactor control parameter optimization device provided in an embodiment of this application;

[0038] Figure 3 This is a schematic block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0039] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0040] To make the objectives, technical solutions, and advantages of this application clearer, the following description will be provided in conjunction with the accompanying drawings and specific embodiments.

[0041] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a method for optimizing control parameters of a phosphine reactor according to an embodiment of this application. The method can be executed by an electronic device and may include:

[0042] S101: Obtain the temperature limit range, pressure limit range, and stirring speed limit range of the reactor.

[0043] In this embodiment, the raw materials for phosphine production include yellow phosphorus, lime slurry, and sodium hydroxide. The process flow for phosphine production using a reactor is as follows: First, a measured amount of lime slurry is added to the reactor using a metering pump, and the flow meter is used to measure the amount of lime slurry. After addition, the reactor is stirred. A measured amount of purified water is used to rinse the lime slurry pipeline before adding it to the reactor; the amount of purified water added is measured using a flow meter. Steam is introduced into the reactor to raise the temperature to 50°C, then the steam is turned off, and a measured amount of liquid yellow phosphorus is fed into the reactor using a flow meter. Steam is introduced into the reactor to raise the temperature to 65°C, then the steam is turned off, and a measured amount of sodium hydroxide solution is slowly added using a flow meter. The sodium hydroxide solution is added dropwise after 1-2 hours. As the reaction proceeds, the exothermic reaction causes the temperature inside the reactor to gradually rise. Circulating water is introduced into the reactor to control the temperature to 80-110°C (the optimal temperature range). After 4 hours, the reaction stage ends. The crude product PH3 produced by the reaction unit enters the dehydration unit, while the condensate from the dehydration unit is returned to the reactor. The salt solution produced by the reaction unit is sent to the tailings unit. When the tailings are collected, the reactor is pressurized to 0.2 MPa using nitrogen gas, and the pressure difference is used to force the tailings into the purification unit.

[0044] Those skilled in the art can set the safe operating range of the core control parameters of the reactor based on the equipment performance, chemical characteristics of the phosphine reaction, and safety regulations. This includes the temperature limit range, pressure limit range, and stirring speed limit range. For example, the temperature limit range could be 50–200°C, the pressure limit range could be 0.1 MPa (atmospheric pressure) to 0.5 MPa, and the stirring speed limit range could be 50 r / min to 500 r / min.

[0045] Based on this, a genetic algorithm can be used to search for the global optimal solution within the safe operating range of the core control parameters, thereby optimizing the temperature control parameters, pressure control parameters, and stirring speed control parameters of the reactor.

[0046] S102: Randomly initialize multiple population individuals based on temperature limit range, pressure limit range, and stirring speed limit range to obtain an initial population; wherein, each population individual is used to represent the value of a set of control parameters, the set of control parameters including the temperature control parameters, pressure control parameters, and stirring speed control parameters of the reactor.

[0047] In this embodiment, multiple populations can be randomly initialized based on temperature limit ranges, pressure limit ranges, and stirring speed limit ranges. Each population individual can be represented by (T, P, S), where T is the value of the temperature control parameter, P is the value of the pressure control parameter, and S is the value of the stirring speed control parameter. A value is randomly selected from each of the temperature limit range, pressure limit range, and stirring speed limit range to form a population individual. Using the same method, multiple populations can be obtained, and these multiple populations are used as the initial population for genetic algorithm optimization.

[0048] The initial population size, i.e., the number of individuals in the initial population, can be set according to actual needs, such as any number between 50 and 200.

[0049] S103: Use a genetic algorithm to perform multiple population iterations on the initial population until the stopping condition is met, and obtain the optimal population individuals.

[0050] Each population iteration operation includes:

[0051] The peak phosphine yield and safety risk assessment value for each individual in the initial population are determined based on the values ​​of the control parameters corresponding to that individual.

[0052] The fitness value of each individual in the population is determined based on the peak phosphine yield and safety risk assessment value.

[0053] Multiple parent individuals are selected from multiple populations based on the fitness values ​​of individuals in each population; the parent individuals are population individuals whose corresponding fitness values ​​are greater than the fitness threshold.

[0054] The population is initialized based on multiple parent individuals.

[0055] For the embodiments of this application, the stopping condition is that the number of iterations reaches a set number, or the change in the maximum fitness value obtained in N consecutive iterations is less than a first threshold.

[0056] Among them, the optimal population individual is the one with the largest fitness value among the multiple population individuals obtained in the last iteration.

[0057] In this embodiment, starting with an initial population, individuals within the population are screened through multiple population iterations to ultimately obtain the optimal individuals. During each population iteration, the production efficiency of phosphine under a set of control parameters corresponding to each individual is first determined, i.e., the peak phosphine yield. Then, the leakage risk, explosion risk, and corrosion risk of the reactor are assessed based on the values ​​of the control parameters corresponding to each individual, yielding a corresponding safety risk assessment value. Finally, the fitness value of each individual is determined based on its peak phosphine yield and safety risk assessment value (e.g., a weighted sum of the peak phosphine yield and the safety risk assessment value), serving as a comprehensive indicator for judging the quality of individuals within the population.

[0058] Based on this, individuals with better performance (i.e., fitness values ​​higher than the fitness threshold) are selected from the initial population according to their fitness values ​​to serve as parent individuals. New individuals are generated based on the parent individuals through the core operations of the genetic algorithm (selection, crossover, mutation), replacing individuals with lower fitness in the initial population to form the next generation population, which can ensure that high-quality genes are passed on.

[0059] Perform multiple population iterations as described above until the set number of iterations is reached, or the optimization effect tends to stabilize (that is, the change in the maximum fitness value of the population is less than the first threshold in N consecutive iterations).

[0060] After multiple population iterations, the individual with the highest fitness value in the last population iteration is selected as the optimal population individual, and its corresponding control parameters are the optimization result.

[0061] S104: Optimize the control parameters of the reactor based on the optimal population individuals.

[0062] In this embodiment, the values ​​of a set of control parameters corresponding to the optimal population individuals are applied to the actual control of the phosphine reactor, thereby achieving the parameter optimization goal.

[0063] Specifically, the temperature control parameter corresponding to the optimal population individual can be used as the target temperature, and the heating power of the heater can be corrected in real time through PID control (proportional, integral, derivative control) to ensure that the actual temperature tracks the target temperature; the pressure control parameter corresponding to the optimal population individual can be used as the target pressure, and the opening of the exhaust valve can be corrected in real time through PID control to ensure that the actual pressure tracks the target pressure; the stirring speed control parameter corresponding to the optimal population individual can be used as the target stirring speed, and the speed of the stirring motor can be corrected in real time through PID control to ensure that the actual stirring speed tracks the target stirring speed.

[0064] Taking PID control of temperature control parameters as an example, in the initial stage of the reaction, the temperature inside the reactor needs to rise rapidly to the target temperature in a short time. At this time, the control cycle of PID control can be set to a smaller first cycle, such as 1 second, that is, PID control is performed once per second. In the middle and end stages of the reaction, the temperature has stabilized in the optimal range set by the genetic algorithm, the raw material consumption and product generation rate are relatively stable, and the parameter change rate is low. At this time, the control cycle of PID control can be set to a larger second cycle, such as 10 seconds, that is, PID control is performed once every 10 seconds.

[0065] Furthermore, considering that it takes time for the heat generated by the heater to transfer to the interior of the reactor, traditional PID control suffers from control lag in adjusting the temperature control parameters. To address this issue, this embodiment first constructs the temperature control transfer function by conducting a step disturbance experiment on the heating power of the reactor:

[0066] ;

[0067] in, A represents the amplification factor, used to characterize the effect of changes in heating power on the steady-state temperature value. For example, if the heating power increases by 10% and the temperature eventually rises by 5℃, then A = 0.5. This represents the inertial time constant, used to characterize the time it takes for the reactor temperature to rise from its initial value to 63.2% of its steady-state value. The pure time delay can be obtained through a step experiment, such as by suddenly increasing the heating power and recording the delay time from the change in heating power to the start of temperature change.

[0068] A specific example of a step disturbance experiment is as follows:

[0069] (1) Set the initial temperature to T1=50℃ and record the initial heating power;

[0070] (2) Step up the heating power ;

[0071] (3) Record the temperature curve T(t) in real time until it stabilizes at T2;

[0072] Based on this, the specific values ​​of each coefficient in the above transfer function calculation formula can be determined:

[0073] (1) Magnification factor A:

[0074] ;

[0075] For example, if the heating power increases by 20%, causing the temperature to rise by 8°C, then A = 0.4.

[0076] (2) Pure time delay :

[0077] The difference between the moment when the temperature begins to change and the moment of the step change, for example, 20 seconds;

[0078] (3) Inertial time constant :

[0079] The time it takes for the temperature to rise from T1 to 0.632 (T2-T1), for example, 30 seconds.

[0080] Therefore, the transfer function for temperature control is obtained as follows:

[0081] .

[0082] Based on this, the temperature of the reactor can be predicted using the temperature control transfer function, as shown below:

[0083] ;

[0084] in, This indicates the estimated temperature of the reactor. This indicates the heating power.

[0085] Without considering hysteresis, the actual temperature of the reactor is estimated, and the hysteresis-free estimated temperature is obtained as follows:

[0086] ;

[0087] in, This indicates a no-hysteresis temperature forecast. This represents a hysteresis-free transfer function.

[0088] Therefore, the ideal deviation without lag is obtained as follows:

[0089] ;

[0090] in, This indicates an ideal deviation without lag. Indicates the target temperature. This represents the measured value of temperature. The error between the actual lag and the estimated lag is used to offset the error. This error is added to the lag-free estimated temperature to obtain the lag-compensated actual temperature. Through lag compensation, the lag-free ideal deviation can be obtained. Based on the lag-free ideal deviation, PID control can be performed to achieve advance adjustment of temperature control parameters.

[0091] Specifically, the calculation formula for PID control is as follows:

[0092] ;

[0093] in, This represents the output of the PID controller. This represents the proportional gain of the PID controller. This represents the integral coefficient of the PID controller. This represents the derivative coefficient of the PID controller.

[0094] As can be seen from the above, this embodiment uses a genetic algorithm as an optimization tool. Through population iteration, it can search for the global optimal solution within the safe operating range of each control parameter. Compared with traditional trial and error methods or local optimization methods, it can improve the optimization accuracy.

[0095] Based on this, during each population iteration, the fitness value of each individual in the population is determined based on its peak phosphine yield and safety risk assessment value. This fitness value serves as a comprehensive indicator for judging the quality of individuals in the population. The control parameters of the reactor are then optimized based on the values ​​of a set of control parameters corresponding to the optimal individual in the final population. This approach ensures both a high phosphine yield and keeps safety risks at a low level. Compared to existing single-objective control methods (such as focusing solely on high yield or high safety), this approach avoids safety accidents (such as overpressure explosion risks) caused by solely pursuing high yields, or low yields due to excessive conservatism, thus achieving a balance between production efficiency and operational safety.

[0096] In one embodiment of this application, updating the initial population based on multiple parent individuals includes:

[0097] By performing crossover operations on multiple parent individuals, multiple first-generation individuals are obtained;

[0098] Mutation operations are performed on multiple first-generation individuals to obtain multiple second-generation individuals;

[0099] Multiple parent individuals, multiple first-generation individuals, and multiple second-generation individuals are identified as the updated initial population.

[0100] In this embodiment, multiple parent individuals can be randomly paired, and the control parameters (temperature, pressure, stirring speed) corresponding to the two parent individuals in the pair can be cross-combined to obtain the first offspring individual. For example, the two parent individuals are... and ,Will and After crossover, the first generation individuals obtained are as follows:

[0101] ;

[0102] in, and Represents two first-generation individuals. The crossover factor is a constant or random number between 0 and 1. The first offspring generated can incorporate the superior traits of the parent individuals, thus expanding population diversity.

[0103] Furthermore, the control parameters of the first generation individuals are randomly mutated (e.g., Gaussian mutation, uniform mutation), and a certain parameter value (e.g., temperature ±2℃) is slightly adjusted within the parameter limit range to generate the second generation individuals. Mutation operations can prevent the population from getting trapped in local optima and maintain evolutionary vitality.

[0104] The parent individuals, first-generation individuals, and second-generation individuals are merged into a new population to replace the initial population and enter the next iteration. The above process not only preserves the superior genes of the parent generation, but also introduces new individuals through crossover and mutation, thereby achieving dynamic optimization of the population.

[0105] In one embodiment of this application, a crossover operation is performed on multiple parent individuals to obtain multiple first child individuals, including:

[0106] Obtain the real-time yield of phosphine;

[0107] If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the crossover probability is set to the first value.

[0108] If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is greater than the second percentage, the crossover probability is set to the second value; the second percentage is greater than the first percentage, and the second value is greater than the first value.

[0109] If the change in phosphine yield is less than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the crossover probability is set to the third value; the third value is greater than the first value and less than the second value.

[0110] Based on the set crossover probability, multiple parent individuals are crossovered to obtain multiple first-generation individuals.

[0111] In this embodiment, the frequency of gene recombination in parent individuals can be controlled by setting the crossover probability. A lower crossover probability can reduce the destruction of high-quality parameter combinations in parent individuals and ensure the stability of the population's evolutionary direction. A higher crossover probability can generate more new parameter combinations through gene recombination in parent individuals, introduce diversity, and avoid the algorithm from getting stuck in local optima (such as repeatedly searching only in a certain temperature range).

[0112] Considering the varying requirements for parameter stability and exploratory (diversity) at different stages of the phosphine reaction (initial, intermediate, and final stages), this embodiment first determines the current reaction stage of the reactor based on the real-time yield (production rate) of phosphine. Specifically, if the real-time yield continuously increases within a set first time period (product change > 0), and the real-time yield / target yield peak value < first percentage (e.g., 30%), it indicates that the reaction has just started and is in the initial stage. If the real-time yield continuously increases within the first time period (yield change > 0), and the real-time yield / target yield peak value > second percentage (e.g., 50%), it indicates that the reaction has entered the intermediate stage and is approaching the target yield peak value. If the real-time yield continuously decreases within the first time period (change < 0), and the real-time yield / target yield peak value < first percentage (e.g., 30%), it indicates that the reaction is gradually declining and is in the final stage. The target yield peak value can be the maximum yield peak value obtained from multiple phosphine production processes in the same reactor, and the first time period is a preset constant, such as 20 minutes.

[0113] Based on this, in the initial stage of the reaction, setting the crossover probability to a relatively small first value (e.g., 0.4) can reduce excessive perturbation of the initial optimal parameter combination and ensure a stable start-up of the reaction; in the middle stage of the reaction, setting the crossover probability to a relatively large second value (e.g., 0.8) can explore better parameter combinations through frequent gene recombination and push for peak yield; in the final stage of the reaction, reducing the crossover probability from the second value to the third value (e.g., 0.6) can adapt to the dynamic changes in raw material consumption and avoid local optima.

[0114] As can be seen from the above, this embodiment determines the reaction stage of the reactor based on the real-time yield of phosphine, and dynamically adjusts the crossover probability based on the reaction stage, so that the population can stably inherit high-quality genes at different stages and introduce new individuals in a timely manner, thereby improving the global optimization ability.

[0115] In one embodiment of this application, multiple first-generation individuals are mutated to obtain multiple second-generation individuals, including:

[0116] Obtain the real-time yield of phosphine;

[0117] If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the mutation probability is set to the fourth value.

[0118] If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is greater than the second percentage, the mutation probability is set to the fifth value; the second percentage is greater than the first percentage, and the fifth value is less than the fourth value.

[0119] If the change in phosphine yield is less than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the mutation probability is set to the sixth value; the sixth value is greater than the fifth value and less than the fourth value.

[0120] Based on the set mutation probability, multiple first-generation individuals are mutated to obtain multiple second-generation individuals.

[0121] In this embodiment, the mutation probability is used to control the frequency of random gene mutations in individual populations. When crossover operations fail to produce better solutions, mutation can introduce entirely new genes by randomly adjusting parameters (such as small temperature fluctuations), helping the population escape local optima. A lower mutation probability can reduce random interference with the already formed high-quality gene combinations in the population and reduce the risk of destroying high-quality parameters. However, an excessively low mutation probability may lead to insufficient population diversity and make it prone to getting trapped in local optima after long-term iterations. A higher mutation probability can introduce new parameter combinations through more random gene changes, enhancing population diversity and helping to escape local optima. However, an excessively high mutation probability may destroy the high-quality genes accumulated by the parent generation, leading to a decrease in population stability.

[0122] Considering the different requirements for parameter stability and exploratory (diversity) at different stages of the phosphine reaction (initial, intermediate and final stages), this embodiment adopts the same processing method as the crossover probability. First, the reaction stage of the reactor is determined based on the real-time yield (production rate) of phosphine, and the mutation probability is dynamically adjusted according to the reaction stage of the reactor.

[0123] Specifically, in the early stage of the reaction, setting the mutation probability to a larger fourth value (e.g., 0.08) can increase the parameter exploration and quickly find the stable start-up range; in the middle stage of the reaction, in conjunction with a larger crossover probability, setting the mutation probability to a smaller fifth value (e.g., 0.03) can finely optimize within the safe range; in the late stage of the reaction, increasing the mutation probability from the fifth value to the sixth value (e.g., 0.06) can avoid getting trapped in local optima (e.g., a decrease in yield due to raw material consumption).

[0124] As can be seen from the above, this embodiment determines the reaction stage of the reactor based on the real-time yield of phosphine, and dynamically adjusts the mutation probability based on the reaction stage, which can enable the population to achieve the optimal exploration intensity in each reaction stage and accelerate the convergence of the algorithm to the global optimum.

[0125] In one embodiment of this application, the raw materials in the phosphine reactor include yellow phosphorus and sodium hydroxide, and the target yield peak is determined by:

[0126] The target yield peak was obtained by finding a mapping relationship between the purity of yellow phosphorus and the concentration of sodium hydroxide. The mapping relationship was obtained by fitting the historical data corresponding to the purity of yellow phosphorus, the concentration of sodium hydroxide, and the target yield peak.

[0127] In this embodiment, considering that even in the same reactor, the purity of yellow phosphorus and the concentration of sodium hydroxide in different batches of raw materials may vary, resulting in different peak yields of phosphine, when determining the reaction stage of the reactor based on the target yield peak, if the raw material purity decreases but the target yield peak at high purity is still used, it will lead to a distortion in the determination of the reaction stage, resulting in the setting of crossover probability or mutation probability not meeting the actual needs of the reaction stage.

[0128] To avoid the above problems, this embodiment can collect phosphine yield data corresponding to different yellow phosphorus purities (e.g., 98.5%, 99.0%, 99.5%) and different sodium hydroxide concentrations (e.g., 8%, 10%, 12%) during multiple production processes, record the actual highest yield under each set of raw material conditions (i.e., the historical yield peak under that condition), use yellow phosphorus purity (c1) and sodium hydroxide concentration (c2) as input variables, and the historical highest yield (y) under the corresponding conditions as output variables, and use data fitting methods (e.g., multiple linear regression or polynomial fitting) to construct a mapping relationship model: y=f(c1,c2).

[0129] During the phosphine reaction, by inputting the purity of the yellow phosphorus and the concentration of sodium hydroxide used in this process into the above mapping model, the target yield peak under the current raw material conditions can be obtained.

[0130] As can be seen from the above, this embodiment dynamically calculates the target yield peak based on the characteristics of the raw materials, ensuring that the "ratio of real-time yield to target peak yield" can truly reflect the reaction stage, and providing a reliable basis for the dynamic adjustment of crossover probability and mutation probability.

[0131] In one embodiment of this application, the raw materials in the phosphine reactor include yellow phosphorus and sodium hydroxide. The peak phosphine yield for each individual in the initial population is determined based on the values ​​of control parameters corresponding to that individual, including:

[0132] The values ​​of the control parameters corresponding to each individual in the initial population, as well as the purity of yellow phosphorus and the concentration of sodium hydroxide, are input into a pre-built regression model to obtain the peak phosphine yield corresponding to that individual in the population.

[0133] In this embodiment, multiple sets of yellow phosphorus purity (e.g., 98.5%, 99.0%, 99.5%) and sodium hydroxide concentration (e.g., 8%, 10%, 12%) corresponding to multiple production processes, along with the corresponding control parameter combinations (temperature, pressure, stirring speed) and the actual measured peak phosphine yield, can be collected. Using yellow phosphorus purity, sodium hydroxide concentration, temperature, pressure, and stirring speed as input features, and the corresponding peak phosphine yield as the output label, a regression model can be trained.

[0134] During the phosphine reaction, the purity of the yellow phosphorus, the concentration of sodium hydroxide, and the values ​​of the control parameters corresponding to each individual in the population are input into the regression model that has been trained above, so as to obtain the peak phosphine yield corresponding to that individual in the population.

[0135] As can be seen from the above, this embodiment simultaneously considers the characteristics of raw materials (purity of yellow phosphorus, concentration of sodium hydroxide) and control parameters (temperature, pressure, stirring speed) to determine the peak phosphine yield corresponding to each individual in the population, making the calculation results more consistent with the actual reaction law.

[0136] In one embodiment of this application, for each individual in the initial population, a safety risk assessment value is determined based on the value of the control parameter corresponding to that individual, including:

[0137] The first risk assessment value for each individual in the population is determined based on the temperature control parameters corresponding to that individual.

[0138] The second risk assessment value for each individual in the population is determined based on the stress control parameters corresponding to that individual.

[0139] The third risk assessment value for each individual in the population is determined based on the stirring speed control parameters corresponding to that individual.

[0140] The first risk assessment value, the second risk assessment value, and the third risk assessment value are weighted and summed to obtain the safety risk assessment value corresponding to the individual in the population.

[0141] In this embodiment, considering that high temperatures accelerate chemical corrosion within the reactor (sodium hydroxide exhibits enhanced corrosivity at high temperatures), leading to oxidation and embrittlement of the reactor's metal materials (such as stainless steel), reducing equipment strength, and potentially causing leaks in the long term; excessive pressure can exceed the equipment's design pressure limit, causing reactor deformation, seal failure, or even physical explosion, resulting in material leakage and personnel casualties. Simultaneously, excessive pressure may cause leaks at valves, flanges, and other sealing points, allowing phosphine gas to diffuse into the working environment; and excessively fast stirring speeds can intensify friction between the materials (yellow phosphorus, sodium hydroxide solution) and the reactor wall and stirring paddle, generating localized high temperatures that may ignite yellow phosphorus or accelerate phosphine volatilization. Therefore, this embodiment determines the safety risk assessment value for each individual in the population based on the values ​​of the control parameters corresponding to that individual.

[0142] Specifically, the first risk assessment value for each individual in the population can be calculated using the following first formula:

[0143] ;

[0144] in, This represents the first risk assessment value, and exp() represents the exponential operation with e as the minimum value. This indicates the value of the temperature control parameter corresponding to an individual in the population. This represents the upper limit of the optimal temperature range. This indicates the upper limit of the temperature range. This represents the preset first proportional coefficient.

[0145] Furthermore, the second risk assessment value corresponding to each individual in the population can be calculated using the following second formula:

[0146] ;

[0147] in, This indicates the second risk assessment value. This indicates the value of the stress control parameter for each individual in the population. This represents the upper limit of the optimal pressure range. This indicates the upper limit of the pressure limit range. This indicates the change in pressure. This represents the rate of change of pressure for an individual in the population. This represents a reference value indicating the amount of pressure change. The greater the positive rate of pressure change, the larger the corresponding second evaluation value. 'b' and 'b' represent the preset second and third proportional coefficients, respectively. In calculating the second risk assessment value, in addition to considering the degree to which the pressure deviates from the optimal pressure range, the pressure change rate is also introduced. This can capture rapid changes in pressure within a short period, improving the comprehensiveness of the pressure risk assessment.

[0148] Furthermore, the third risk assessment value corresponding to each individual in the population can be calculated using the following third formula:

[0149] ;

[0150] in, This represents the third risk assessment value. This indicates the value of the stirring rate control parameter corresponding to each individual in the population. This indicates the upper limit of the optimal stirring speed range. This indicates the upper limit of the range of stirring speed limits.

[0151] It should be noted that the optimal temperature range refers to the range of temperature control parameters that achieve the best overall reaction efficiency and safety; the optimal pressure range refers to the range of pressure control parameters that achieve the best overall reaction efficiency and safety; and the optimal stirring speed range refers to the range of stirring speed control parameters that achieve the best overall reaction efficiency and safety. Therefore, the optimal temperature range is a subset of the temperature limit range, the optimal pressure range is a subset of the pressure limit range, and the optimal stirring speed range is a subset of the stirring speed limit range. If the temperature control parameter value corresponding to an individual in a population exceeds the upper limit of the optimal temperature range, or the pressure control parameter value corresponding to an individual in a population exceeds the upper limit of the optimal pressure range, or the stirring speed control parameter value corresponding to an individual in a population exceeds the upper limit of the optimal stirring speed range, the safety risk assessment value of the corresponding individual in the population will increase significantly.

[0152] In this embodiment, the optimal temperature range, optimal pressure range, and optimal stirring speed range can be obtained by statistically analyzing historical reaction data. For example, the optimal temperature range can be 80–110°C, the optimal pressure range can be 0.1–0.2 MPa, and the optimal stirring speed range can be 100–300 rpm.

[0153] As can be seen from the above, this embodiment determines the corresponding safety risk assessment value based on the degree to which the control parameters corresponding to each individual in the population deviate from the optimal parameter range (optimal temperature range, optimal pressure range, or optimal stirring speed range), which can make the calculation result of the safety risk assessment value accurately match the actual safety status.

[0154] Corresponding to the phosphine reactor control parameter optimization method in the above embodiment, Figure 2 This is a structural block diagram of a phosphine reactor control parameter optimization device provided in one embodiment of this application. For ease of explanation, only the parts relevant to the embodiment of this application are shown. References Figure 2The phosphine reactor control parameter optimization device 20 includes: a data acquisition module 21, a population initialization module 22, a population optimization module 23, and a control parameter optimization module 24.

[0155] Among them, the data acquisition module 21 is used to acquire the temperature limit range, pressure limit range and stirring speed limit range of the reactor;

[0156] The population initialization module 22 is used to randomly initialize multiple population individuals based on temperature limit range, pressure limit range and stirring speed limit range to obtain an initial population; wherein, each population individual is used to represent the value of a set of control parameters, the set of control parameters including the temperature control parameters, pressure control parameters and stirring speed control parameters of the reactor;

[0157] Population optimization module 23 is used to perform multiple population iteration operations on the initial population using a genetic algorithm until the stopping condition is met, and to obtain the optimal population individuals.

[0158] The control parameter optimization module 24 is used to optimize the control parameters of the reactor based on the optimal population individuals;

[0159] In any population iteration operation, the population optimization module is used for:

[0160] The peak phosphine yield and safety risk assessment value for each individual in the initial population are determined based on the values ​​of the control parameters corresponding to that individual.

[0161] The fitness value of each individual in the population is determined based on the peak phosphine yield and safety risk assessment value.

[0162] Multiple parent individuals are selected from multiple populations based on the fitness values ​​of individuals in each population; the parent individuals are population individuals whose corresponding fitness values ​​are greater than the fitness threshold.

[0163] The population is initialized based on multiple parent individuals;

[0164] The stopping condition is that the number of iterations reaches the set number, or the change in the maximum fitness value obtained in N consecutive iterations is less than the first threshold.

[0165] Among them, the optimal population individual is the one with the largest fitness value among the multiple population individuals obtained in the last iteration.

[0166] In one embodiment of this application, the population optimization module 23 is specifically used for:

[0167] By performing crossover operations on multiple parent individuals, multiple first-generation individuals are obtained;

[0168] Mutation operations are performed on multiple first-generation individuals to obtain multiple second-generation individuals;

[0169] Multiple parent individuals, multiple first-generation individuals, and multiple second-generation individuals are identified as the updated initial population.

[0170] In one embodiment of this application, the population optimization module 23 is further configured to:

[0171] Obtain the real-time yield of phosphine;

[0172] If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the crossover probability is set to the first value.

[0173] If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is greater than the second percentage, the crossover probability is set to the second value; the second percentage is greater than the first percentage, and the second value is greater than the first value.

[0174] If the change in phosphine yield is less than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the crossover probability is set to the third value; the third value is greater than the first value and less than the second value.

[0175] Based on the set crossover probability, multiple parent individuals are crossovered to obtain multiple first-generation individuals.

[0176] In one embodiment of this application, the population optimization module 23 is further configured to:

[0177] Obtain the real-time yield of phosphine;

[0178] If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the mutation probability is set to the fourth value.

[0179] If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is greater than the second percentage, the mutation probability is set to the fifth value; the second percentage is greater than the first percentage, and the fifth value is less than the fourth value.

[0180] If the change in phosphine yield is less than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the mutation probability is set to the sixth value; the sixth value is greater than the fifth value and less than the fourth value.

[0181] Based on the set mutation probability, multiple first-generation individuals are mutated to obtain multiple second-generation individuals.

[0182] In one embodiment of this application, the raw materials in the phosphine reactor include yellow phosphorus and sodium hydroxide, and the population optimization module 23 is further used for:

[0183] The target yield peak was obtained by finding a mapping relationship between the purity of yellow phosphorus and the concentration of sodium hydroxide. The mapping relationship was obtained by fitting the historical data corresponding to the purity of yellow phosphorus, the concentration of sodium hydroxide, and the target yield peak.

[0184] In one embodiment of this application, the raw materials in the phosphine reactor include yellow phosphorus and sodium hydroxide, and the population optimization module 23 is specifically used for:

[0185] The values ​​of the control parameters corresponding to each individual in the initial population, as well as the purity of yellow phosphorus and the concentration of sodium hydroxide, are input into a pre-built regression model to obtain the peak phosphine yield corresponding to that individual in the population.

[0186] In one embodiment of this application, the raw materials in the phosphine reactor include yellow phosphorus and sodium hydroxide, and the population optimization module 23 is specifically used for:

[0187] The first risk assessment value for each individual in the population is determined based on the temperature control parameters corresponding to that individual.

[0188] The second risk assessment value for each individual in the population is determined based on the stress control parameters corresponding to that individual.

[0189] The third risk assessment value for each individual in the population is determined based on the stirring speed control parameters corresponding to that individual.

[0190] The first risk assessment value, the second risk assessment value, and the third risk assessment value are weighted and summed to obtain the safety risk assessment value corresponding to the individual in the population.

[0191] See Figure 3 , Figure 3 This is a schematic block diagram of an electronic device provided according to an embodiment of this application. Figure 3The electronic device 300 in this embodiment may include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memories 304 store computer programs, including program instructions. The processors 301 execute the program instructions stored in the memories 304. Specifically, the processors 301 are configured to invoke the program instructions to perform the functions of each module / unit in the above-described device embodiments, for example... Figure 2 The functions of the data acquisition module 21, population initialization module 22, population optimization module 23, and control parameter optimization module 24 are shown.

[0192] It should be understood that, in the embodiments of this application, the processor 301 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0193] Input device 302 may include a touchpad, a fingerprint sensor (for collecting the user's fingerprint information and fingerprint orientation information), a microphone, etc., and output device 303 may include a display (LCD, etc.), a speaker, etc.

[0194] The memory 304 may include read-only memory and random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include non-volatile random access memory. For example, the memory 304 may also store preset constants such as fitness threshold and first threshold.

[0195] In specific implementations, the processor 301, input device 302, and output device 303 described in the embodiments of this application can execute the implementation method described in the phosphine reactor control parameter optimization method provided in the embodiments of this application, or they can execute the implementation method of the electronic device described in the embodiments of this application, which will not be repeated here.

[0196] In another embodiment of this application, a computer-readable storage medium is provided. This computer-readable storage medium stores a computer program, which includes program instructions. When executed by a processor, the program instructions implement all or part of the processes in the methods described above. Alternatively, the computer program can instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0197] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the foregoing embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device. Furthermore, the computer-readable storage medium can include both internal and external storage units of the electronic device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0198] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0199] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the electronic devices and units described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0200] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, modules may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connections shown or discussed may be indirect coupling or communication connections through some interfaces or units, or they may be electrical, mechanical, or other forms of connection.

[0201] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of this application, depending on actual needs.

[0202] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0203] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for optimizing control parameters of a phosphine reactor, characterized in that, include: Obtain the temperature limit range, pressure limit range, and stirring speed limit range for the reactor; Multiple population individuals are randomly initialized based on the temperature limit range, pressure limit range, and stirring speed limit range to obtain an initial population; wherein, each population individual is used to represent the value of a set of control parameters, which includes the temperature control parameters, pressure control parameters, and stirring speed control parameters of the reactor; The initial population is iterated multiple times using a genetic algorithm until the stopping condition is met, and the optimal population individual is obtained. The control parameters of the reactor are optimized based on the optimal population individuals. Each population iteration operation includes: The peak phosphine yield and safety risk assessment value for each individual in the initial population are determined based on the values ​​of the control parameters corresponding to that individual. The fitness value of each individual in the population is determined based on the peak phosphine yield and safety risk assessment value. Multiple parent individuals are selected from multiple populations based on the fitness values ​​of individuals in each population; the parent individuals are populations whose corresponding fitness values ​​are greater than the fitness threshold. The initial population is updated based on multiple parent individuals; The stopping condition is that the number of iterations reaches a set number, or the change in the maximum fitness value obtained in N consecutive iterations is less than a first threshold. The optimal population individual is the population individual with the largest fitness value among the multiple population individuals obtained in the last iteration process; The process of updating the initial population based on multiple parent individuals includes: Perform a crossover operation on the multiple parent individuals to obtain multiple first child individuals; Mutation operations are performed on the multiple first-generation individuals to obtain multiple second-generation individuals; The plurality of parent individuals, the plurality of first-generation individuals, and the plurality of second-generation individuals are determined as the updated initial population; Performing a crossover operation on the multiple parent individuals yields multiple first-generation individuals, including: Obtain the real-time yield of phosphine; If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the crossover probability is set to the first value. If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is greater than the second percentage, the crossover probability is set to the second value; the second percentage is greater than the first percentage, and the second value is greater than the first value. If the change in phosphine yield is less than zero within the set first time period, and the ratio of real-time phosphine yield to the target yield peak is less than the first proportion, the crossover probability is set to a third value; the third value is greater than the first value and less than the second value. Based on the set crossover probability, crossover operations are performed on the multiple parent individuals to obtain multiple first child individuals.

2. The method for optimizing control parameters of a phosphine reactor as described in claim 1, characterized in that, Mutation operations are performed on the plurality of first-generation individuals to obtain a plurality of second-generation individuals, including: Obtain the real-time yield of phosphine; If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the mutation probability is set to the fourth value. If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is greater than the second percentage, the mutation probability is set to the fifth value; the second percentage is greater than the first percentage, and the fifth value is less than the fourth value. If the change in phosphine yield is less than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the mutation probability is set to the sixth value; the sixth value is greater than the fifth value, and the sixth value is less than the fourth value. Based on the set mutation probability, mutation operations are performed on the plurality of first-generation individuals to obtain a plurality of second-generation individuals.

3. The method for optimizing control parameters of a phosphine reactor as described in claim 1 or 2, characterized in that, The raw materials in the phosphine reactor include yellow phosphorus and sodium hydroxide, and the method for determining the target peak yield includes: The target yield peak is obtained by finding a mapping relationship between the purity of the yellow phosphorus and the concentration of the sodium hydroxide; wherein the mapping relationship is obtained by data fitting based on the historical data corresponding to the purity of the yellow phosphorus, the concentration of the sodium hydroxide, and the target yield peak.

4. The method for optimizing control parameters of a phosphine reactor as described in claim 1, characterized in that, The raw materials in the phosphine reactor include yellow phosphorus and sodium hydroxide. The peak phosphine yield for each individual in the initial population is determined based on the values ​​of the control parameters corresponding to that individual, including: The values ​​of the control parameters corresponding to each individual in the initial population, as well as the purity of the yellow phosphorus and the concentration of the sodium hydroxide, are input into a pre-constructed regression model to obtain the peak phosphine yield corresponding to that individual.

5. The method for optimizing control parameters of a phosphine reactor as described in claim 1, characterized in that, For each individual in the initial population, a safety risk assessment value is determined based on the value of the control parameters corresponding to that individual, including: The first risk assessment value for each individual in the population is determined based on the temperature control parameters corresponding to that individual. The second risk assessment value for each individual in the population is determined based on the stress control parameters corresponding to that individual. The third risk assessment value for each individual in the population is determined based on the stirring speed control parameters corresponding to that individual. The first risk assessment value, the second risk assessment value, and the third risk assessment value are weighted and summed to obtain the safety risk assessment value corresponding to the individual in the population.

6. A device for optimizing control parameters of a phosphine reactor, characterized in that, include: The data acquisition module is used to acquire the temperature limit range, pressure limit range, and stirring speed limit range of the reactor. The population initialization module is used to randomly initialize multiple population individuals based on the temperature limit range, pressure limit range, and stirring speed limit range to obtain an initialized population; wherein, each population individual is used to represent the value of a set of control parameters, the set of control parameters including the temperature control parameters, pressure control parameters, and stirring speed control parameters of the reactor; The population optimization module is used to perform multiple population iteration operations on the initial population using a genetic algorithm until the stopping condition is met, thereby obtaining the optimal population individuals. The control parameter optimization module is used to optimize the control parameters of the reactor based on the optimal population individuals; In any population iteration operation, the population optimization module is used for: The peak phosphine yield and safety risk assessment value for each individual in the initial population are determined based on the values ​​of the control parameters corresponding to that individual. The fitness value of each individual in the population is determined based on the peak phosphine yield and safety risk assessment value. Multiple parent individuals are selected from multiple populations based on the fitness values ​​of individuals in each population; the parent individuals are populations whose corresponding fitness values ​​are greater than the fitness threshold. The initial population is updated based on multiple parent individuals; The stopping condition is that the number of iterations reaches a set number, or the change in the maximum fitness value obtained in N consecutive iterations is less than a first threshold. The optimal population individual is the population individual with the largest fitness value among the multiple population individuals obtained in the last iteration process; The population optimization module is specifically used for: By performing crossover operations on multiple parent individuals, multiple first-generation individuals are obtained; Mutation operations are performed on multiple first-generation individuals to obtain multiple second-generation individuals; Multiple parent individuals, multiple first-generation individuals, and multiple second-generation individuals are identified as the updated initial population; The population optimization module is also specifically used for: Obtain the real-time yield of phosphine; If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the crossover probability is set to the first value. If the change in phosphine yield is greater than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is greater than the second percentage, the crossover probability is set to the second value; the second percentage is greater than the first percentage, and the second value is greater than the first value. If the change in phosphine yield is less than zero within the set first time period, and the ratio of the real-time yield of phosphine to the target yield peak is less than the first proportion, the crossover probability is set to the third value; the third value is greater than the first value and less than the second value. Based on the set crossover probability, multiple parent individuals are crossovered to obtain multiple first-generation individuals.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 5.

Images