Internal mixer temperature control method based on machine learning algorithm

Through the machine learning algorithm-based temperature control method of the machine mixer, the database is used to analyze and predict the temperature changes of the machine mixer and automatically adjust the temperature of the machine mixer, which solves the problem of insufficient personalized requirements and control accuracy in the traditional method, and achieves rapid and Accurate temperature control.

CN118952497APending Publication Date: 2024-11-15QINGDAO UNIV OF SCI & TECH
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410336667.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-03-22
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The traditional temperature control method of mixing machine cannot meet the personalized needs of different users, and there are problems such as large over-pressure and long adjustment time, especially when the temperature change rate of mixing machine is large, the control effect is poor.

Method used

Using the machine learning algorithm-based temperature control method of the mixer, the data set is constructed by monitoring and analyzing the parameters of the mixer such as the set temperature, medium temperature, flow rate, speed, pressure and material temperature, and constructing a database, calculating the temperature change rate, and dividing the data set , predict and automatically adjust the set temperature to verify its accuracy and reduce user operations.

Benefits of technology

It realizes automatic and accurate calculation of the set temperature of the mixer, shortens the temperature adjustment time, improves control accuracy and robustness, adapts to different user needs, and reduces the frequency of users changing the set temperature.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118952497A_ABST
    Figure CN118952497A_ABST
Patent Text Reader

Abstract

The invention discloses an internal mixer temperature control method based on a machine learning algorithm, which belongs to the technical field of temperature control of molding material preprocessing, and comprises the following steps: monitoring data of an internal mixer, drawing a medium temperature change curve by using data stored in a database, calculating the duration of a detection moment, and calculating a temperature change rate T rate, the method comprises the following steps: dividing a data set, associating a temperature change rate T rate with the data set, obtaining independent variable data, calculating a temperature change rate T rate'at a to-be-measured moment, searching for the same temperature change rate T rate, selecting the data set, calculating the distance between the independent variable data and each data row of the selected data set, and determining a prediction data row. The set temperature T'corresponding to the prediction data line is obtained, whether the set temperature T 'deviates from historical similar records or not is verified through the medium temperature change curve, and the effects that the set temperature of the internal mixer can be automatically and accurately calculated, and frequent operation of a user is not needed are achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of temperature control of molding material pretreatment, and in particular to a temperature control method of an internal mixer based on a machine learning algorithm. Background Art

[0002] Rubber has the characteristics of high elasticity, high electrical insulation, low thermal conductivity, and excellent air tightness. The plastication of raw rubber and the mixing of rubber materials are carried out in internal mixers. In recent years, the steady development of internal mixers has been to improve the productivity of various mixed rubbers. Improvements in rotor design and a better understanding of internal mixer technology have led to higher requirements for internal mixers. As production increases, the temperature control capability of internal mixers should keep pace with other developments.

[0003] Reference Figure 1 , the heated equipment passes through the medium, so that the medium and the heated equipment exchange energy, thereby controlling the temperature of a specific area of ​​the heated equipment. The working principle of the internal mixer temperature control is to use oil or water as the medium to control the temperature when it flows through the internal mixer to achieve the best mixing effect. The internal mixer temperature control adopts the method of indirect cooling of the return oil through the cooler, and then pressurized by the pump and heated by the heater to send it to the internal mixer to achieve the requirements of heating and constant temperature.

[0004] Different users have different temperature requirements, so the set temperatures are also different. The same user may have different preferences for setting temperatures under different external environments. Traditional internal mixer temperature control methods cannot meet users' personalized needs for internal mixer temperature control. Moreover, the material temperature jump rate varies greatly within different ranges of internal mixer temperature change rate. At present, most internal mixer temperature control products have the same configuration across the country, that is, a closed constant temperature control system that sets the temperature of the internal mixer to a fixed value. This control system is a large time-delay control system with poor control quality problems such as large overshoot and long adjustment time. In particular, the control effect is poor in the case of model mismatch and interference in the internal mixer temperature control system.

[0005] The above-mentioned prior art solutions have the following defects: the use of the system cannot well meet the use requirements of different users and cannot meet the needs of various industries and mines, which requires the user to change the set temperature every time. Summary of the invention

[0006] In order to reduce the number of times that users frequently change the set temperature of an internal mixer, the present application provides a temperature control method for an internal mixer based on a machine learning algorithm.

[0007] On the one hand, the present application provides a method for controlling the temperature of an internal mixer based on a machine learning algorithm, which adopts the following technical solution: A method for controlling temperature of an internal mixer based on a machine learning algorithm comprises the following steps: S1, monitor the set temperature T of the internal mixer, the temperature of the temperature control medium T1 of the internal mixer, the medium flow Q, the rotor speed V, the upper push bolt pressure P, the material temperature T2 of the internal mixer and the Mooney viscosity N and save them into the database; S2. Use the historical mixer temperature control medium temperature T1 stored in the database to form a medium temperature change curve, calculate the slope of each point of the medium temperature change curve, compare the absolute value of the slope with a preset value, mark the slope whose absolute value exceeds the preset value, select the two nearest marked slopes, use the time between the two marked slopes as the detection time duration, and calculate the temperature change rate T rate according to the detection time duration and the mixer temperature control medium temperature T1; S3, dividing the database into multiple data sets according to each historical temperature change rate T rate, and associating the temperature change rate T rate with the data set; S4, obtaining the independent variables at the time to be measured, including the temperature of the temperature-controlled medium of the internal mixer T1', the medium flow rate Q', the rotor speed V', the upper ejector pressure P', the material temperature of the internal mixer T2' and the Mooney viscosity N'; S5, calculating the temperature change rate T rate ' at the time to be measured according to the current temperature T1 of the temperature control medium of the internal mixer and the temperature T1 of the temperature control medium of the internal mixer which has a time equal to the detection time since the current time; S6, searching for the same temperature change rate T rate according to the temperature change rate T rate 'at the time to be measured, searching for the data set according to the found temperature change rate T rate, and selecting the data set according to the found temperature change rate T rate; S7, calculating the distance between the independent variable data and each data row of the selected data set, and determining the predicted data row; S8, obtaining the set temperature T' corresponding to the predicted data row; S9. Mark the mixer temperature control medium temperature T1 equal to the set temperature T' on the medium temperature change curve, call the slope of the marked point, select the marked point with the smallest absolute value of the slope, call the medium flow Q, rotor speed V, upper push bolt pressure P, mixer material temperature T2 and Mooney viscosity N associated with the marked point, compare the called data with the independent variable data, and if the differences in each item are lower than the preset value, output the set temperature T'.

[0008] By adopting the above scheme, the set temperature can be automatically calculated according to the current data through historical data, and the database can be constructed by using big data pattern recognition technology. Because the change trend of the set temperature is different in different heating rate intervals, the database range is narrowed by dividing the data set, which is convenient for more rapid determination of the data row corresponding to the independent variable data. Then, the appropriate detection time duration is automatically determined by the medium temperature change curve, and after the set temperature is calculated, the set temperature is verified, and the historical data is used to determine whether the calculation is wrong. The set temperature will be output only when it is determined that there is no problem with the set temperature. The set temperature of the internal mixer can be calculated automatically and accurately without frequent user operation.

[0009] Preferably, the step S3 further comprises: Calculate the temperature change rate T corresponding to each data row 速率 , temperature change rate T 速率 The calculation formula is as follows: n Where n is the number of data rows, is the set temperature T when the number of data rows is n, and 0≤T 速率 The data rows with <5℃ / min are put into data set A, and 5≤T 速率 The data rows with a value less than 10℃ / min are placed in the B data set, and the T 速率 The data rows with a value of ≥10℃ / min are put into the C data set.

[0010] By adopting the above scheme, the temperature change rate T 速率 The calculation is done through data rows, and the calculation process is simple, fast and accurate.

[0011] Preferably, the step S7 further includes: The formula used to calculate the distance between the independent variable data and each data row of the selected data set is as follows: Where dist is the data row.

[0012] Preferably, the step S7 further includes: Sort the data rows in ascending order of distance.

[0013] Preferably, the step S7 further includes: Select the first K data rows, determine the corresponding T values ​​in the K data rows, and calculate the average value of T.

[0014] By adopting the above scheme, the K nearest neighbor algorithm can obtain more accurate and reasonable prediction values.

[0015] Preferably, the step S1 further comprises: The data are normalized and denormalized, and the normalization formula is: Where X is T1, Q, V, P, N or T2, and Y is the normalized value.

[0016] By adopting the above scheme, all feature values ​​are mapped to the [0,1] interval, achieving proportional scaling of the original values ​​and achieving more accurate prediction and calculation.

[0017] Preferably, the method further comprises the steps of: S10. If any difference is higher than a preset value, a calculation failure is displayed.

[0018] By adopting the above scheme, when it is determined that there is a problem with the calculated set temperature T', fault information is automatically displayed to remind the user to operate in time.

[0019] Preferably, the step S10 further includes: S10, obtaining the set temperature T of the internal mixer corresponding to the called data, calculating the difference between the called set temperature T of the internal mixer and the set temperature T', and displaying the set temperature T, the set temperature T' and the difference; S11, receiving the input corrected set temperature T', using the corrected set temperature T' to overwrite the set temperature T', and importing the overwritten set temperature T' into the database.

[0020] By adopting the above scheme, after determining that the set temperature T' is faulty, the system automatically calculates and corrects the set temperature T', and imports the modified set temperature T' into the database for recalculation.

[0021] On the other hand, the present application provides a smart terminal based on a machine learning algorithm that adopts the following technical solution: A smart terminal based on a machine learning algorithm includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute any of the above methods.

[0022] On the other hand, the present application provides a computer-readable storage medium based on a machine learning algorithm using the following technical solution: A computer-readable storage medium based on a machine learning algorithm includes a computer program that can be loaded by a processor and execute any of the above methods.

[0023] In summary, the present invention has the following beneficial effects: 1. The set temperature can be automatically calculated based on the current data through historical data. The database is constructed by using big data pattern recognition technology. Because the temperature change trend is different in different heating rates, the database range is narrowed by dividing the data set, which is convenient for more rapid determination of the data row corresponding to the independent variable data. The set temperature of the internal mixer can be automatically and accurately calculated without frequent user operations.

[0024] 2. Automatically determine the appropriate detection time and duration through the medium temperature change curve, and verify the set temperature after calculating it. Determine whether the calculation is wrong through historical data. The set temperature will be output only when it is determined that there is no problem with the set temperature.

[0025] 3. This control method is more robust and can achieve good control effect even when the control object model is mismatched or interference occurs. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a schematic diagram of the temperature control composition of an internal mixer in the background technology of this application.

[0027] Figure 2 It is an overall flow chart of the first embodiment of the present application. DETAILED DESCRIPTION Example

[0028] The embodiment of the present application discloses a temperature control method for an internal mixer based on a machine learning algorithm, and the specific steps are as follows: S100, monitoring the set temperature T of the internal mixer, the temperature of the temperature control medium T1 of the internal mixer, the medium flow Q, the rotor speed V, the upper push bolt pressure P, the material temperature T2 of the internal mixer and the Mooney viscosity N and saving them into the database.

[0029] In this embodiment, part of the data in the database is listed in the following table. Table 1 Part of the data in the database S101, normalizing and denormalizing the data, the normalization formula is: Where X is T1, Q, V, P, N or T2, Y is the normalized value, Xmax and Xmin are the maximum and minimum values ​​of the original data range.

[0030] S200, using the historical internal mixer temperature control medium temperature T1 stored in the database to form a medium temperature change curve.

[0031] S201, calculating the slope of each point of the medium temperature change curve, comparing the absolute value of the slope with a preset value, marking the slope whose absolute value exceeds the preset value, selecting the two marked slopes closest to each other, and using the duration between the two marked slopes as the detection time duration.

[0032] S202, calculate the temperature change rate T according to the detection time and the temperature T1 of the temperature control medium of the internal mixer 速率 .

[0033] S300, respectively calculate the temperature change rate T corresponding to each data row 速率 , temperature change rate T 速率 The calculation formula is as follows: n Where n is the number of data rows, is the set temperature T when the number of data rows is n, and 0≤T 速率 The data rows with <5℃ / min are put into data set A, and 5≤T 速率 The data rows with a value less than 10℃ / min are placed in the B data set, and the T 速率 The data rows with a temperature change rate T 速率 Associated with a dataset.

[0034] In this embodiment, the serial number items in Table 1 are used as data labels. It can be seen that the data rows with serial numbers 2, 3, 8, 12, 13, and 18 in Table 1 are used as data sets A, the data rows with serial numbers 1, 4, 5, 6, 7, 11, 14, 16, 17, 19, 20, 21, 22, 23, 24, 25, 26, and 27 are used as data sets B, and the data rows with serial numbers 9, 10, 15, 28, 29, and 30 are used as data sets C.

[0035] If the range of the set temperature of the mixing chamber in the internal mixer is 0≤T 速率 <5℃ / min, the material temperature in the mixing chamber does not change much, but if the temperature setting range of the mixing chamber is 5≤T 速率 If the temperature is less than 10℃ / min, the material temperature in the mixer will rise rapidly. If the set temperature is modified according to the traditional temperature control PID method, the material temperature will overheat due to the long response time. The above algorithm can respond quickly. Compared with the traditional PID control method, this algorithm has higher accuracy and faster control speed.

[0036] S400, obtaining independent variables at the time to be measured, including the temperature T1' of the temperature-controlled medium of the internal mixer, the medium flow Q', the rotor speed V', the upper ejector pressure P', the material temperature T2' of the internal mixer, and the Mooney viscosity N'.

[0037] S500, calculating the temperature change rate T at the time to be measured based on the current temperature T1 of the temperature control medium of the internal mixer and the temperature T1 of the temperature control medium of the internal mixer equal to the time from the current time to the detection time 速率 '.

[0038] S600, according to the temperature change rate T at the time to be measured 速率 'Find the same temperature change rate T 速率 , find the temperature change rate T according to the data set found 速率 Select a dataset.

[0039] S700, calculate the distance between the independent variable data and each data row of the selected data set using the following formula: Where dist is the data row, determine the predicted data row, and sort the data rows in ascending order of distance.

[0040] S800, obtaining a set temperature T' corresponding to the predicted data row.

[0041] S900, mark the mixer temperature control medium temperature T1 equal to the set temperature T' on the medium temperature change curve, call the slope of the marked point, select the marked point with the smallest absolute value of the slope, call the medium flow Q, rotor speed V, upper push bolt pressure P, mixer material temperature T2 and Mooney viscosity N associated with the marked point, and compare the called data with the independent variable data.

[0042] S901. If all the differences are lower than the preset values, the set temperature T' is output.

[0043] S1000: If any difference is higher than the preset value, a calculation failure is displayed.

[0044] S1001. Obtain the internal mixer set temperature T corresponding to the called data, calculate the difference between the called internal mixer set temperature T and the set temperature T', and display the set temperature T, the set temperature T' and the difference.

[0045] S1100, receiving the input corrected set temperature T', using the corrected set temperature T' to overwrite the set temperature T', and importing the overwritten set temperature T' into the database.

[0046] The implementation principle of a temperature control method for an internal mixer based on a machine learning algorithm in an embodiment of the present application is as follows: the set temperature can be automatically calculated according to the current data through historical data, and a database is constructed by using big data pattern recognition technology. Because the change trend of the set temperature is different in different intervals of the heating rate, the database range is narrowed by dividing the data set, so as to more quickly determine the data row corresponding to the independent variable data. Then, the appropriate detection time duration is automatically determined by the medium temperature change curve, and after the set temperature is calculated, the set temperature is verified, and whether the calculation is wrong is determined by historical data. The set temperature will be output only when it is determined that there is no problem with the set temperature. The set temperature of the internal mixer can be calculated automatically and accurately without frequent user operation. Example

[0047] An embodiment of the present application discloses an intelligent terminal based on a machine learning algorithm, including a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute any of the above methods. Example

[0048] An embodiment of the present application discloses a computer-readable storage medium based on a machine learning algorithm, including a computer program that can be loaded by a processor and execute any of the methods described above.

[0049] The embodiments of this specific implementation method are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A temperature control method for an internal mixer based on a machine learning algorithm, characterized in that: The following steps are involved: S1, monitor the set temperature T of the internal mixer, the temperature of the temperature control medium T1 of the internal mixer, the medium flow Q, the rotor speed V, the upper push bolt pressure P, the material temperature T2 of the internal mixer and the Mooney viscosity N and save them into the database; S2, using the historical mixer temperature control medium temperature T1 stored in the database to form a medium temperature change curve, calculating the slope of each point of the medium temperature change curve, comparing the absolute value of the slope with the preset value, marking the slope whose absolute value exceeds the preset value, selecting the two nearest marked slopes, and taking the time between the two marked slopes as the detection time duration, and calculating the temperature change rate T according to the detection time duration and the mixer temperature control medium temperature T1 速率 ; S3, according to the historical temperature change rate T 速率 Divide the database into multiple data sets and calculate the temperature change rate T 速率 Associated with a dataset; S4, obtaining the independent variables at the time to be measured, including the temperature of the temperature-controlled medium of the internal mixer T1', the medium flow rate Q', the rotor speed V', the upper ejector pressure P', the material temperature of the internal mixer T2' and the Mooney viscosity N'; S5, calculate the temperature change rate T at the time to be measured based on the current temperature T1 of the internal mixer temperature control medium and the temperature T1 of the internal mixer temperature control medium that is equal to the time from the current time to the detection time 速率 '; S6, according to the temperature change rate T at the time to be measured 速率 'Find the same temperature change rate T 速率 , find the temperature change rate T according to the data set found 速率 Select the dataset; S7, calculating the distance between the independent variable data and each data row of the selected data set, and determining the predicted data row; S8, obtaining the set temperature T' corresponding to the predicted data row; S9. Mark the mixer temperature control medium temperature T1 equal to the set temperature T' on the medium temperature change curve, call the slope of the marked point, select the marked point with the smallest absolute value of the slope, call the medium flow Q, rotor speed V, upper push bolt pressure P, mixer material temperature T2 and Mooney viscosity N associated with the marked point, compare the called data with the independent variable data, and if the differences in each item are lower than the preset value, output the set temperature T'.

2. The internal mixer temperature control method based on machine learning algorithm according to claim 1, characterized in that: The step S3 further comprises: Calculate the temperature change rate T corresponding to each data row 速率 , temperature change rate T 速率 The calculation formula is as follows: n Where n is the number of data rows, is the set temperature T when the number of data rows is n, and 0≤T 速率 The data rows with <5℃ / min are put into data set A, and 5≤T 速率 The data rows with a value less than 10℃ / min are placed in the B data set, and the T 速率 The data rows with a value of ≥10℃ / min are put into the C data set.

3. The internal mixer temperature control method based on machine learning algorithm according to claim 1, characterized in that: The step S7 further comprises: The formula used to calculate the distance between the independent variable data and each data row of the selected data set is as follows: Where dist is the data row.

4. The internal mixer temperature control method based on machine learning algorithm according to claim 1, characterized in that: The step S7 further comprises: Sort the data rows in ascending order of distance.

5. The internal mixer temperature control method based on machine learning algorithm according to claim 4, characterized in that: The step S7 further comprises: Select the first K data rows, determine the corresponding T values ​​in the K data rows, and calculate the average value of T.

6. The internal mixer temperature control method based on machine learning algorithm according to claim 1, characterized in that: The step S1 further comprises: The data are normalized and denormalized, and the normalization formula is: Where X is T1, Q, V, P, N or T2, and Y is the normalized value.

7. The method for controlling the temperature of an internal mixer based on a machine learning algorithm according to claim 1, characterized in that: Also includes the steps: S10. If any difference is higher than a preset value, a calculation failure is displayed.

8. The internal mixer temperature control method based on machine learning algorithm according to claim 7, characterized in that: The step S10 further includes: S10, obtaining the internal mixer set temperature T corresponding to the called data, calculating the difference between the called internal mixer set temperature T and the set temperature T', and displaying the set temperature T, the set temperature T' and the difference; S11, receiving the input corrected set temperature T', using the corrected set temperature T' to overwrite the set temperature T', and importing the overwritten set temperature T' into the database.

9. An intelligent terminal based on a machine learning algorithm, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executes any one of the methods in claims 1-8.

10. A computer-readable storage medium based on a machine learning algorithm, characterized in that: The device comprises a computer program which can be loaded by a processor and executes the method according to any one of claims 1 to 8.

Citation Information

Cited By

  • Dynamic calibration method and system for actual volume of internal mixer and storage medium

    CN120663440A