A manufacturing parameter control decision system and method for energetic materials

Abstract

The present invention discloses a manufacturing parameter control decision system and method for energetic materials, belonging to the field of parameter control technology. The present invention comprises the steps of: collecting reaction records of nitration reactions and storing them in a database to form a training set; establishing a deep learning model and completing the training of the deep learning model; obtaining the reaction start time and temperature change diagram of the current nitration reaction, analyzing the cooling start time, first temperature characteristics and predicted characteristics of the current nitration reaction; judging whether the current nitration reaction needs to be regulated once according to the predicted characteristics of the current nitration reaction, and adjusting the preset flow rate; analyzing the actual second temperature characteristics of the current nitration reaction, judging whether the current nitration reaction needs to be regulated twice, and performing the secondary regulation; the present invention realizes intelligent regulation of manufacturing parameters, overcomes the problems of response lag and insufficient regulation accuracy in the prior art, and improves the reliability and adaptability of the energetic material manufacturing process.

Description

Technical Field

[0001] The present invention belongs to the technical field of parameter control, and specifically relates to a manufacturing parameter control decision system and method for energetic materials. Background Art

[0002] The manufacturing process of energetic materials involves multiple steps such as nitration, mixing, pressing, coating, and drying. The manufacturing parameters include temperature, pressure, acid concentration, stirring rate, etc. The precise control of manufacturing parameters in the manufacturing process of energetic materials is crucial to product quality and safety.

[0003] However, due to the complexity of chemical reactions and the uncertainty of environmental factors, traditional manufacturing models have the following problems: The lack of intelligent adjustment mechanisms results in delayed parameter control responses. In particular, in the nitration process, the temperature of the nitration reaction cannot be accurately controlled based on real-time data, resulting in poor stability of the manufactured energetic materials and high safety risks.

[0004] Therefore, people urgently need a manufacturing parameter control decision system for energetic materials to solve the above problems. Summary of the Invention

[0005] The purpose of the present invention is to provide a manufacturing parameter control decision system for energetic materials to solve the problems raised in the above background technology.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] A manufacturing parameter control decision-making method for energetic materials, the method comprising the following steps:

[0008] S1. Collecting reaction records of nitration reactions and storing them in a database; screening the reaction records in the database, analyzing the first temperature characteristics and the second temperature characteristics of the reaction records to form a training set; establishing a deep learning model, and completing the training of the deep learning model using the training set;

[0009] S2. Obtaining the reaction start time and temperature change diagram of the current nitration reaction, analyzing the cooling start time and first temperature characteristics of the current nitration reaction; extracting the raw material usage in the current nitration reaction, as well as the initial temperature and preset flow rate of the coolant from the database, and analyzing the prediction characteristics of the current nitration reaction in combination with the deep learning model;

[0010] S3. Determine whether the current nitrification reaction needs to be regulated based on the predicted characteristics of the current nitrification reaction; if the current nitrification reaction needs to be regulated, determine a flow rate adjustment value for the regulation and adjust the preset flow rate;

[0011] S4. Analyze the actual second temperature characteristic of the current nitrification reaction according to the temperature change diagram of the current nitrification reaction, determine whether secondary regulation of the current nitrification reaction is required, and perform secondary regulation.

[0012] According to the above technical solution, step S1 includes:

[0013] S1-1. Each time a nitration reaction is performed, a reaction record is generated and stored in a database; the reaction record includes the amount of raw materials used, the reaction start time, the cooling start time, the initial temperature of the coolant, the average flow rate of the coolant, the reaction end time, the cooling end time, and a temperature change graph;

[0014] S1-2. Extracting a standard temperature range for nitration reaction from a database, extracting maximum and minimum temperature values ​​from the temperature change graph in each reaction record, and screening out reaction records whose maximum and minimum temperature values ​​are within the standard temperature range for nitration reaction to form a first record set;

[0015] S1-3. Based on the first record set, obtain, from a temperature change graph of a reaction record in the first record set, a temperature difference between the cooling start time and the reaction start time in the reaction record, and obtain a time interval between the reaction start time and the cooling start time in the reaction record; and use a ratio of the temperature difference between the cooling start time and the reaction start time in the reaction record to the time interval between the reaction start time and the cooling start time as a first temperature feature of the reaction record;

[0016] S1-4. Extracting the temperature of each reaction record in the first record set at the cooling start time, and taking the average of the temperatures of all reaction records in the first record set at the cooling start time as the first temperature threshold;

[0017] Obtaining the time interval between the cooling start time and the cooling end time of each reaction record in the first record set, and taking the average of the time intervals between the cooling start time and the cooling end time of all reaction records in the first record set as the first time parameter;

[0018] Calculating the ratio of the temperature difference between the cooling end time and the cooling start time in a certain reaction record to the time interval from the cooling start time to the cooling end time in the reaction record as the second temperature feature of the reaction record;

[0019] The raw material amount, coolant initial temperature, coolant average flow rate, and first temperature characteristic of each reaction record in the first record set are used as inputs of the training set; the second temperature characteristic of each reaction record in the first record set is used as outputs of the training set to form a training set;

[0020] S1-5. Establish a deep learning model and complete the training of the deep learning model through the training set;

[0021] By screening the calculated data to train the model, the accuracy of predicting temperature changes in the nitration reaction can be improved, forming an intelligent model that can be used for decision-making, improving the automation level of the manufacturing process, and making parameter control no longer rely on fixed experience, but on intelligent judgment based on historical data.

[0022] According to the above technical solution, step S2 includes:

[0023] S2-1. Obtaining the reaction start time and temperature change diagram of the current nitration reaction through a sensing device. If the temperature at a certain time after the reaction start time of the current nitration reaction is greater than or equal to a first temperature threshold, use this time as the cooling start time of the current nitration reaction;

[0024] S2-2. Calculating a first temperature characteristic of the current nitration reaction based on the reaction start time, cooling start time, and temperature change graph of the current nitration reaction;

[0025] Extracting the current raw material usage, as well as the initial temperature and preset flow rate of the coolant, from the database, inputting the current raw material usage, the first temperature characteristic, the cooling start time, and the initial temperature and preset flow rate of the coolant into the trained deep learning model, and outputting the second temperature characteristic of the current nitration reaction as a prediction feature of the current nitration reaction;

[0026] By monitoring temperature changes in real time, we can predict the occurrence of abnormal situations in advance, improve the foresight of temperature control, avoid delayed adjustments, and conduct intelligent analysis based on historical data to improve prediction accuracy and make control more precise.

[0027] According to the above technical solution, step S3 includes:

[0028] S3-1. Calculate the value of the multiplication of the predicted characteristic of the current nitrification reaction and the first time parameter, and record it as the predicted change amount; calculate the value of the sum of the first temperature threshold and the predicted change amount, and record it as the first control judgment amount;

[0029] S3-2. If the primary control judgment amount is not within the standard temperature range of the nitrification reaction, the current nitrification reaction needs to be regulated once;

[0030] S3-3. If the current nitrification reaction needs to be regulated, extract the minimum value within the standard temperature range of the nitrification reaction, calculate the value of the difference between the minimum value within the standard temperature range of the nitrification reaction and the first temperature threshold, and record it as a regulation change; calculate the ratio of the regulation change to the first time parameter, and record it as a regulation feature;

[0031] S3-4. Using the current nitration reaction raw material usage, the first temperature characteristic, the cooling start time, and the initial temperature of the coolant as input, and the primary control characteristic as output, combined with the deep learning model, obtain the average flow rate of the coolant as the flow rate adjustment value for the primary control; adjust the preset flow rate to a value corresponding to the flow rate adjustment value for the primary control; and adjust the predicted characteristic to a value corresponding to the primary control characteristic;

[0032] Timely adjust the preset flow rate of the coolant to prevent the temperature from being too high or too low. Through intelligent calculation of the adjustment range, avoid the errors that may be caused by human adjustment, make the reaction temperature closer to the ideal state, and improve product quality and stability.

[0033] According to the above technical solution, step S4 includes:

[0034] S4-1. Regulating the temperature of the current nitration reaction according to the cooling start time and a preset flow rate of the coolant; setting a unit time interval, obtaining the temperature at a corresponding time after the unit time interval from the cooling start time, calculating the difference between the temperature at the cooling start time and the corresponding time after the unit time interval, and calculating the ratio of the difference to the unit time interval, recording the difference as the actual second temperature characteristic of the current nitration reaction;

[0035] S4-2, if the absolute value of the difference between the actual second temperature characteristic of the current nitrification reaction and the predicted characteristic is recorded as the secondary control judgment amount;

[0036] If the secondary control judgment amount is greater than or equal to the preset error value, the current nitrification reaction needs to be secondary controlled;

[0037] S4-3. When secondary control of the current nitrification reaction is required, if the difference between the actual second temperature characteristic of the current nitrification reaction and the predicted characteristic is negative, the preset flow rate is reduced, and the ratio of the value of the preset flow rate before the reduction minus the preset flow rate after the reduction to the preset flow rate before the reduction is the value obtained by multiplying the preset weight by the secondary control judgment amount;

[0038] If the difference between the actual second temperature characteristic of the current nitrification reaction and the predicted characteristic is positive, the preset flow rate is increased, and the ratio of the value of the preset flow rate after the increase minus the preset flow rate before the increase to the preset flow rate before the increase is the value obtained by multiplying the preset weight by the secondary control judgment amount;

[0039] Through secondary adjustment, the accuracy of temperature control is improved, the impact of errors is reduced, the problems of delayed flow rate adjustment and slow response in the traditional mode are solved, safety is improved, a closed-loop control system is formed, temperature stability is ensured, and quality fluctuations and safety hazards are reduced.

[0040] A manufacturing parameter control decision system for energetic materials, comprising an acquisition and training module, a reaction prediction module, a primary control module, and a secondary control module;

[0041] The acquisition and training module is responsible for collecting, screening and analyzing historical reaction records of the nitrification reaction, forming a training set, establishing a deep learning model and training it, and providing data support for subsequent regulation; the reaction prediction module is used to monitor the current nitrification reaction in real time, analyze the cooling start time and the first temperature characteristics of the current nitrification reaction, and use the deep learning model to predict the reaction characteristics to provide a basis for regulation and control decisions; the primary regulation module is used to determine whether the current nitrification reaction needs to be regulated based on the predicted characteristics, and adjust the preset flow rate of the coolant; the secondary regulation module is used to analyze the actual second temperature characteristics of the current nitrification reaction based on the actual temperature change status of the current nitrification reaction, and determine whether the current nitrification reaction needs to be regulated secondary, so as to accurately control the current nitrification reaction and ensure that the temperature of the current nitrification reaction remains stable.

[0042] According to the above technical solution, the acquisition and training module includes a data acquisition unit and a model training unit;

[0043] The data acquisition unit is used to collect reaction records of the nitration reaction, including the amount of raw materials used, the reaction start time, the cooling start time, the initial temperature of the coolant, the average flow rate of the coolant, the reaction end time, the cooling end time and the temperature change diagram, and store them in a database; the model training unit is used to filter the reaction records from the database, and calculate the first temperature feature and the second temperature feature corresponding to each reaction record after filtering, form a training set, establish a deep learning model, and complete the training of the deep learning model.

[0044] According to the above technical solution, the reaction prediction module includes a feature extraction unit and a feature prediction unit;

[0045] The feature extraction unit is used to obtain temperature change data of the current nitration reaction through a sensing device and analyze the cooling start time and the first temperature feature of the current nitration reaction; the feature prediction unit is used to obtain the predicted features of the current nitration reaction based on the raw material dosage, the first temperature feature, the cooling start time, the initial temperature and the preset flow rate of the coolant of the current nitration reaction in combination with a deep learning model.

[0046] According to the above technical solution, the primary control module includes a control determination unit and a flow rate adjustment unit;

[0047] The control judgment unit is used to predict the characteristics of the current nitrification reaction and determine whether the current nitrification reaction needs to be regulated once; the flow rate adjustment unit is used to analyze the flow rate adjustment value for the current regulation if the current nitrification reaction needs to be regulated once, and adjust the preset flow rate.

[0048] According to the above technical solution, the secondary control module includes an error analysis unit and a dynamic adjustment unit;

[0049] The error analysis unit is used to calculate the error between the actual temperature characteristics and the predicted characteristics, and determine whether the current nitrification reaction needs to be secondary regulated; the dynamic adjustment unit is used to dynamically adjust the preset flow rate according to the error, optimize the flow rate parameters, and ensure that the temperature change meets expectations.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] The present invention uses a deep learning model to accurately predict the temperature characteristics of the nitration reaction, and dynamically adjusts it in combination with real-time sensor data, thereby realizing intelligent regulation of manufacturing parameters and improving the stability and safety of the reaction process. At the same time, the present invention establishes a standardized temperature characteristic analysis method based on historical reaction data, optimizes the cooling strategy of the nitration reaction, effectively reduces temperature fluctuations, and improves product consistency. Secondly, the present invention adopts a phased regulation strategy, which accurately controls the coolant flow rate through a combination of primary regulation and secondary regulation, shortens the regulation response time, reduces the need for human intervention, and improves production efficiency. In addition, the present invention combines an error analysis method to perform real-time correction of deviations in the regulation process, ensuring that the manufacturing parameters are always within the optimal range, overcoming the problems of response lag and insufficient regulation accuracy in the prior art, and improving the reliability and adaptability of the energetic material manufacturing process. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0053] Figure 1 It is a flow chart of a manufacturing parameter control decision-making method for energetic materials according to the present invention;

[0054] Figure 2 It is a structural schematic diagram of a manufacturing parameter control decision system for energetic materials according to the present invention. DETAILED DESCRIPTION

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by a person of ordinary skill in the art without making any creative effort shall fall within the scope of protection of the present invention.

[0056] See also Figure 1 , the present invention provides a technical solution:

[0057] A manufacturing parameter control decision-making method for energetic materials, the method comprising the following steps:

[0058] S1. Collecting reaction records of nitration reactions and storing them in a database; screening the reaction records in the database, analyzing the first temperature characteristics and the second temperature characteristics of the reaction records to form a training set; establishing a deep learning model, and completing the training of the deep learning model using the training set;

[0059] According to the above technical solution, step S1 includes:

[0060] S1-1. Each time a nitration reaction is performed, a reaction record is generated and stored in a database; the reaction record includes the amount of raw materials used, the reaction start time, the cooling start time, the initial temperature of the coolant, the average flow rate of the coolant, the reaction end time, the cooling end time, and a temperature change graph;

[0061] S1-2. Extracting a standard temperature range for nitration reaction from a database, extracting maximum and minimum temperature values ​​from the temperature change graph in each reaction record, and screening out reaction records whose maximum and minimum temperature values ​​are within the standard temperature range for nitration reaction to form a first record set;

[0062] S1-3. Based on the first record set, obtain, from a temperature change graph of a reaction record in the first record set, a temperature difference between the cooling start time and the reaction start time in the reaction record, and obtain a time interval between the reaction start time and the cooling start time in the reaction record; and use a ratio of the temperature difference between the cooling start time and the reaction start time in the reaction record to the time interval between the reaction start time and the cooling start time as a first temperature feature of the reaction record;

[0063] S1-4. Extracting the temperature of each reaction record in the first record set at the cooling start time, and taking the average of the temperatures of all reaction records in the first record set at the cooling start time as the first temperature threshold;

[0064] Obtaining the time interval between the cooling start time and the cooling end time of each reaction record in the first record set, and taking the average of the time intervals between the cooling start time and the cooling end time of all reaction records in the first record set as the first time parameter;

[0065] Calculating the ratio of the temperature difference between the cooling end time and the cooling start time in a certain reaction record to the time interval from the cooling start time to the cooling end time in the reaction record as the second temperature feature of the reaction record;

[0066] The raw material amount, coolant initial temperature, coolant average flow rate, and first temperature characteristic of each reaction record in the first record set are used as inputs of the training set; the second temperature characteristic of each reaction record in the first record set is used as outputs of the training set to form a training set;

[0067] S1-5. Establish a deep learning model and complete the training of the deep learning model through the training set;

[0068] By screening the calculated data to train the model, the accuracy of predicting temperature changes in the nitration reaction can be improved, forming an intelligent model that can be used for decision-making, improving the automation level of the manufacturing process, and making parameter control no longer rely on fixed experience, but on intelligent judgment based on historical data.

[0069] S2. Obtaining the reaction start time and temperature change diagram of the current nitration reaction, analyzing the cooling start time and first temperature characteristics of the current nitration reaction; extracting the raw material usage in the current nitration reaction, as well as the initial temperature and preset flow rate of the coolant from the database, and analyzing the prediction characteristics of the current nitration reaction in combination with the deep learning model;

[0070] According to the above technical solution, step S2 includes:

[0071] S2-1. Obtaining the reaction start time and temperature change diagram of the current nitration reaction through a sensing device. If the temperature at a certain time after the reaction start time of the current nitration reaction is greater than or equal to a first temperature threshold, use this time as the cooling start time of the current nitration reaction;

[0072] S2-2. Calculating a first temperature characteristic of the current nitration reaction based on the reaction start time, cooling start time, and temperature change graph of the current nitration reaction;

[0073] Extracting the current raw material usage, as well as the initial temperature and preset flow rate of the coolant, from the database, inputting the current raw material usage, the first temperature characteristic, the cooling start time, and the initial temperature and preset flow rate of the coolant into the trained deep learning model, and outputting the second temperature characteristic of the current nitration reaction as a prediction feature of the current nitration reaction;

[0074] By monitoring temperature changes in real time, we can predict the occurrence of abnormal situations in advance, improve the foresight of temperature control, avoid delayed adjustments, and conduct intelligent analysis based on historical data to improve prediction accuracy and make control more precise.

[0075] S3. Determine whether the current nitrification reaction needs to be regulated based on the predicted characteristics of the current nitrification reaction; if the current nitrification reaction needs to be regulated, determine a flow rate adjustment value for the regulation and adjust the preset flow rate;

[0076] According to the above technical solution, step S3 includes:

[0077] S3-1. Calculate the value of the multiplication of the predicted characteristic of the current nitrification reaction and the first time parameter, and record it as the predicted change amount; calculate the value of the sum of the first temperature threshold and the predicted change amount, and record it as the first control judgment amount;

[0078] S3-2. If the primary control judgment amount is not within the standard temperature range of the nitrification reaction, the current nitrification reaction needs to be regulated once;

[0079] S3-3. If the current nitrification reaction needs to be regulated, extract the minimum value within the standard temperature range of the nitrification reaction, calculate the value of the difference between the minimum value within the standard temperature range of the nitrification reaction and the first temperature threshold, and record it as a regulation change; calculate the ratio of the regulation change to the first time parameter, and record it as a regulation feature;

[0080] S3-4. Using the current nitration reaction raw material usage, the first temperature characteristic, the cooling start time, and the initial temperature of the coolant as input, and the primary control characteristic as output, combined with the deep learning model, obtain the average flow rate of the coolant as the flow rate adjustment value for the primary control; adjust the preset flow rate to a value corresponding to the flow rate adjustment value for the primary control; and adjust the predicted characteristic to a value corresponding to the primary control characteristic;

[0081] Timely adjust the preset flow rate of the coolant to prevent the temperature from being too high or too low. Through intelligent calculation of the adjustment range, avoid the errors that may be caused by human adjustment, make the reaction temperature closer to the ideal state, and improve product quality and stability.

[0082] S4. Analyze the actual second temperature characteristic of the current nitrification reaction according to the temperature change graph of the current nitrification reaction, determine whether secondary regulation of the current nitrification reaction is required, and perform secondary regulation;

[0083] According to the above technical solution, step S4 includes:

[0084] S4-1. Regulating the temperature of the current nitration reaction according to the cooling start time and a preset flow rate of the coolant; setting a unit time interval, obtaining the temperature at a corresponding time after the unit time interval from the cooling start time, calculating the difference between the temperature at the cooling start time and the corresponding time after the unit time interval, and calculating the ratio of the difference to the unit time interval, recording the difference as the actual second temperature characteristic of the current nitration reaction;

[0085] S4-2, if the absolute value of the difference between the actual second temperature characteristic of the current nitrification reaction and the predicted characteristic is recorded as the secondary control judgment amount;

[0086] If the secondary control judgment amount is greater than or equal to the preset error value, the current nitrification reaction needs to be secondary controlled;

[0087] S4-3. When secondary control of the current nitrification reaction is required, if the difference between the actual second temperature characteristic of the current nitrification reaction and the predicted characteristic is negative, the preset flow rate is reduced, and the ratio of the value of the preset flow rate before the reduction minus the preset flow rate after the reduction to the preset flow rate before the reduction is the value obtained by multiplying the preset weight by the secondary control judgment amount;

[0088] If the difference between the actual second temperature characteristic of the current nitrification reaction and the predicted characteristic is positive, the preset flow rate is increased, and the ratio of the value of the preset flow rate after the increase minus the preset flow rate before the increase to the preset flow rate before the increase is the value obtained by multiplying the preset weight by the secondary control judgment amount;

[0089] For example:

[0090] The cooling start time of the current nitration reaction is t0 = 300, the temperature at the cooling start time is 85, and the preset flow rate of the coolant is 10. The unit time interval is set to 60, and the temperature at t1 = t0 + 60 = 360 is 80. The actual second temperature characteristic of the current nitration reaction is calculated as (80-85) / 60 = -0.833.

[0091] The predicted characteristic is -0.1; the calculated secondary control judgment value is |(−0.0833)−(−0.1)|=0.0167;

[0092] The default error value is 0.015. Since 0.0167 ≥ 0.015, secondary adjustment is required;

[0093] The difference between the actual second temperature characteristic and the predicted characteristic of the current nitrification reaction is (−0.0833)−(−0.1)=0.0167, which is positive. Increase the preset flow rate.

[0094] The default weight is 10;

[0095] The value of the preset weight multiplied by the secondary control judgment amount is 10×0.0167=0.167;

[0096] The ratio of the value of the preset flow rate after the increase minus the preset flow rate before the increase to the preset flow rate before the increase is 0.167, so the preset flow rate after the increase is 10.167;

[0097] Through secondary adjustment, the accuracy of temperature control is improved, the impact of errors is reduced, the problems of delayed flow rate adjustment and slow response in the traditional mode are solved, safety is improved, a closed-loop control system is formed, temperature stability is ensured, and quality fluctuations and safety hazards are reduced.

[0098] Please refer to Figure 2 , a manufacturing parameter control decision system for energetic materials, the system includes an acquisition training module, a reaction prediction module, a primary control module and a secondary control module;

[0099] The acquisition and training module is responsible for collecting, screening and analyzing historical reaction records of the nitrification reaction, forming a training set, establishing a deep learning model and training it, and providing data support for subsequent regulation; the reaction prediction module is used to monitor the current nitrification reaction in real time, analyze the cooling start time and the first temperature characteristics of the current nitrification reaction, and use the deep learning model to predict the reaction characteristics to provide a basis for regulation and control decisions; the primary regulation module is used to determine whether the current nitrification reaction needs to be regulated based on the predicted characteristics, and adjust the preset flow rate of the coolant; the secondary regulation module is used to analyze the actual second temperature characteristics of the current nitrification reaction based on the actual temperature change status of the current nitrification reaction, and determine whether the current nitrification reaction needs to be regulated secondary, so as to accurately control the current nitrification reaction and ensure that the temperature of the current nitrification reaction remains stable.

[0100] According to the above technical solution, the acquisition and training module includes a data acquisition unit and a model training unit;

[0101] The data acquisition unit is used to collect reaction records of the nitration reaction, including the amount of raw materials used, the reaction start time, the cooling start time, the initial temperature of the coolant, the average flow rate of the coolant, the reaction end time, the cooling end time and the temperature change diagram, and store them in a database; the model training unit is used to filter the reaction records from the database, and calculate the first temperature feature and the second temperature feature corresponding to each reaction record after filtering, form a training set, establish a deep learning model, and complete the training of the deep learning model.

[0102] According to the above technical solution, the reaction prediction module includes a feature extraction unit and a feature prediction unit;

[0103] The feature extraction unit is used to obtain temperature change data of the current nitration reaction through a sensing device and analyze the cooling start time and the first temperature feature of the current nitration reaction; the feature prediction unit is used to obtain the predicted features of the current nitration reaction based on the raw material dosage, the first temperature feature, the cooling start time, the initial temperature and the preset flow rate of the coolant of the current nitration reaction in combination with a deep learning model.

[0104] According to the above technical solution, the primary control module includes a control determination unit and a flow rate adjustment unit;

[0105] The control judgment unit is used to predict the characteristics of the current nitrification reaction and determine whether the current nitrification reaction needs to be regulated once; the flow rate adjustment unit is used to analyze the flow rate adjustment value for the current regulation if the current nitrification reaction needs to be regulated once, and adjust the preset flow rate.

[0106] According to the above technical solution, the secondary control module includes an error analysis unit and a dynamic adjustment unit;

[0107] The error analysis unit is used to calculate the error between the actual temperature characteristics and the predicted characteristics, and determine whether the current nitrification reaction needs to be secondary regulated; the dynamic adjustment unit is used to dynamically adjust the preset flow rate according to the error, optimize the flow rate parameters, and ensure that the temperature change meets expectations.

[0108] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0109] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A manufacturing parameter control decision-making method for energetic materials, characterized by: The method comprises the following steps: S1. Collecting reaction records of nitration reactions and storing them in a database; screening the reaction records in the database, analyzing the first temperature characteristics and the second temperature characteristics of the reaction records to form a training set; establishing a deep learning model, and completing the training of the deep learning model using the training set; S2. Obtaining the reaction start time and temperature change diagram of the current nitration reaction, analyzing the cooling start time and first temperature characteristics of the current nitration reaction; extracting the raw material usage in the current nitration reaction, as well as the initial temperature and preset flow rate of the coolant from the database, and analyzing the prediction characteristics of the current nitration reaction in combination with the deep learning model; S3. Determine whether the current nitrification reaction needs to be regulated based on the predicted characteristics of the current nitrification reaction; if the current nitrification reaction needs to be regulated, determine a flow rate adjustment value for the regulation and adjust the preset flow rate; S4. Analyze the actual second temperature characteristic of the current nitrification reaction according to the temperature change diagram of the current nitrification reaction, determine whether secondary regulation of the current nitrification reaction is required, and perform secondary regulation.

2. The manufacturing parameter control decision-making method for energetic materials according to claim 1, characterized in that: The step S1 comprises: S1-1. Each time a nitration reaction is performed, a reaction record is generated and stored in a database; the reaction record includes the amount of raw materials used, the reaction start time, the cooling start time, the initial temperature of the coolant, the average flow rate of the coolant, the reaction end time, the cooling end time, and a temperature change graph; S1-2. Extracting a standard temperature range for nitration reaction from a database, extracting maximum and minimum temperature values ​​from the temperature change graph in each reaction record, and screening out reaction records whose maximum and minimum temperature values ​​are within the standard temperature range for nitration reaction to form a first record set; S1-3. Based on the first record set, obtain, from a temperature change graph of a reaction record in the first record set, a temperature difference between the cooling start time and the reaction start time in the reaction record, and obtain a time interval between the reaction start time and the cooling start time in the reaction record; and use a ratio of the temperature difference between the cooling start time and the reaction start time in the reaction record to the time interval between the reaction start time and the cooling start time as a first temperature feature of the reaction record; S1-4. Extracting the temperature of each reaction record in the first record set at the cooling start time, and taking the average of the temperatures of all reaction records in the first record set at the cooling start time as the first temperature threshold; Obtaining the time interval between the cooling start time and the cooling end time of each reaction record in the first record set, and taking the average of the time intervals between the cooling start time and the cooling end time of all reaction records in the first record set as the first time parameter; Calculating the ratio of the temperature difference between the cooling end time and the cooling start time in a certain reaction record to the time interval from the cooling start time to the cooling end time in the reaction record as the second temperature feature of the reaction record; The raw material amount, coolant initial temperature, coolant average flow rate, and first temperature characteristic of each reaction record in the first record set are used as inputs of the training set; the second temperature characteristic of each reaction record in the first record set is used as outputs of the training set to form a training set; S1-5. Establish a deep learning model and complete the training of the deep learning model through the training set.

3. The manufacturing parameter control decision-making method for energetic materials according to claim 2, characterized in that: The step S2 comprises: S2-1. Obtaining the reaction start time and temperature change diagram of the current nitration reaction through a sensing device. If the temperature at a certain time after the reaction start time of the current nitration reaction is greater than or equal to a first temperature threshold, use this time as the cooling start time of the current nitration reaction; S2-2. Calculating a first temperature characteristic of the current nitration reaction based on the reaction start time, cooling start time, and temperature change graph of the current nitration reaction; The raw material usage, the initial temperature, and the preset flow rate of the coolant for the current nitration reaction are extracted from the database. The raw material usage, the first temperature characteristic, the cooling start time, and the initial temperature and the preset flow rate of the coolant for the current nitration reaction are input into the trained deep learning model, and the second temperature characteristic of the current nitration reaction is output as the prediction feature of the current nitration reaction.

4. The manufacturing parameter control decision-making method for energetic materials according to claim 3, characterized in that: The step S3 comprises: S3-1. Calculate the value of the multiplication of the predicted characteristic of the current nitrification reaction and the first time parameter, and record it as the predicted change amount; calculate the value of the sum of the first temperature threshold and the predicted change amount, and record it as the first control judgment amount; S3-2. If the primary control judgment amount is not within the standard temperature range of the nitrification reaction, the current nitrification reaction needs to be regulated once; S3-3. If the current nitrification reaction needs to be regulated, extract the minimum value within the standard temperature range of the nitrification reaction, calculate the value of the difference between the minimum value within the standard temperature range of the nitrification reaction and the first temperature threshold, and record it as a regulation change; calculate the ratio of the regulation change to the first time parameter, and record it as a regulation feature; S3-4. The raw material dosage, the first temperature characteristic, the cooling start time, and the initial temperature of the coolant of the current nitration reaction are used as inputs, and the one-time control characteristic is used as output. Combined with the deep learning model, the average flow rate of the coolant is obtained as the flow rate adjustment value for the one-time control; the preset flow rate is adjusted to the value corresponding to the flow rate adjustment value for the one-time control; and the predicted characteristic is adjusted to the value corresponding to the one-time control characteristic.

5. The manufacturing parameter control decision-making method for energetic materials according to claim 4, characterized in that: The step S4 comprises: S4-1. Regulating the temperature of the current nitration reaction according to the cooling start time and a preset flow rate of the coolant; setting a unit time interval, obtaining the temperature at a corresponding time after the unit time interval from the cooling start time, calculating the difference between the temperature at the cooling start time and the corresponding time after the unit time interval, and calculating the ratio of the difference to the unit time interval, recording the difference as the actual second temperature characteristic of the current nitration reaction; S4-2, if the absolute value of the difference between the actual second temperature characteristic of the current nitrification reaction and the predicted characteristic is recorded as the secondary control judgment amount; If the secondary control judgment amount is greater than or equal to the preset error value, the current nitrification reaction needs to be secondary controlled; S4-3. When secondary control of the current nitrification reaction is required, if the difference between the actual second temperature characteristic of the current nitrification reaction and the predicted characteristic is negative, the preset flow rate is reduced, and the ratio of the value of the preset flow rate before the reduction minus the preset flow rate after the reduction to the preset flow rate before the reduction is the value obtained by multiplying the preset weight by the secondary control judgment amount; If the difference between the actual second temperature characteristic of the current nitrification reaction and the predicted characteristic is positive, the preset flow rate is increased, and the ratio of the value of the preset flow rate after the increase minus the preset flow rate before the increase to the preset flow rate before the increase is the value multiplied by the preset weight and the secondary control judgment amount.

6. A manufacturing parameter control decision system for energetic materials, used to implement the manufacturing parameter control decision method for energetic materials according to any one of claims 1 to 5, characterized in that: The system includes an acquisition and training module, a reaction prediction module, a primary control module, and a secondary control module; The acquisition and training module is responsible for collecting, screening and analyzing historical reaction records of the nitrification reaction, forming a training set, establishing a deep learning model and training it, and providing data support for subsequent regulation; the reaction prediction module is used to monitor the current nitrification reaction in real time, analyze the cooling start time and the first temperature characteristics of the current nitrification reaction, and use the deep learning model to predict the reaction characteristics to provide a basis for regulation and control decisions; the primary regulation module is used to determine whether the current nitrification reaction needs to be regulated based on the predicted characteristics, and adjust the preset flow rate of the coolant; the secondary regulation module is used to analyze the actual second temperature characteristics of the current nitrification reaction based on the actual temperature change status of the current nitrification reaction, and determine whether the current nitrification reaction needs to be regulated secondary, so as to accurately control the current nitrification reaction and ensure that the temperature of the current nitrification reaction remains stable.

7. The manufacturing parameter control decision system for energetic materials according to claim 6, characterized in that: The acquisition and training module includes a data acquisition unit and a model training unit; The data acquisition unit is used to collect reaction records of the nitration reaction, including the amount of raw materials used, the reaction start time, the cooling start time, the initial temperature of the coolant, the average flow rate of the coolant, the reaction end time, the cooling end time and the temperature change diagram, and store them in a database; the model training unit is used to filter the reaction records from the database, and calculate the first temperature feature and the second temperature feature corresponding to each reaction record after filtering, form a training set, establish a deep learning model, and complete the training of the deep learning model.

8. The manufacturing parameter control decision system for energetic materials according to claim 6, characterized in that: The reaction prediction module includes a feature extraction unit and a feature prediction unit; The feature extraction unit is used to obtain temperature change data of the current nitration reaction through a sensing device and analyze the cooling start time and the first temperature feature of the current nitration reaction; the feature prediction unit is used to obtain the predicted features of the current nitration reaction based on the raw material dosage, the first temperature feature, the cooling start time, the initial temperature and the preset flow rate of the coolant of the current nitration reaction in combination with a deep learning model.

9. The manufacturing parameter control decision system for energetic materials according to claim 6, characterized in that: The primary control module includes a control determination unit and a flow rate adjustment unit; The control judgment unit is used to predict the characteristics of the current nitrification reaction and determine whether the current nitrification reaction needs to be regulated once; the flow rate adjustment unit is used to analyze the flow rate adjustment value for the current regulation if the current nitrification reaction needs to be regulated once, and adjust the preset flow rate.

10. The manufacturing parameter control decision system for energetic materials according to claim 6, characterized in that: The secondary control module includes an error analysis unit and a dynamic adjustment unit; The error analysis unit is used to calculate the error between the actual temperature characteristics and the predicted characteristics, and determine whether the current nitrification reaction needs to be secondary regulated; the dynamic adjustment unit is used to dynamically adjust the preset flow rate according to the error, optimize the flow rate parameters, and ensure that the temperature change meets expectations.

Images