Temperature control system for flexible film production
By adaptively adjusting smoothing parameters through real-time data acquisition and random tree cutting analysis, the problem of unstable melt temperature in the multi-layer co-extrusion process was solved, and efficient temperature control and stable product quality were achieved.
Patent Information
- Application Number
- CN202511104933.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-07
- Publication Date
- 2025-09-16
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
During the multi-layer co-extrusion process, unstable fluctuations in the melt temperature of each layer lead to unstable product quality. The traditional exponential smoothing algorithm cannot adapt to local temperature changes, resulting in prediction lag and delayed temperature control response.
The data acquisition module is used to obtain the melt temperature in real time. The change degree acquisition module and the final smoothing parameter acquisition module are used to construct a random cutting tree to analyze the reliability of the melt temperature change. The smoothing parameters are adaptively adjusted for prediction to ensure that each layer of melt reaches the target temperature before merging.
The temperature control accuracy is improved, ensuring that each layer of melt reaches the target temperature before converging, improving the interlayer bonding strength and overall film performance, and significantly improving product quality and consistency.
Smart Images
Figure CN120653042A_ABST
Abstract
Description
Technical Field
[0001] The present invention generally relates to the field of data processing technology. More specifically, the present invention relates to a temperature control system for producing flexible films. Background Art
[0002] In the manufacture of high-performance flexible films, multi-layer co-extrusion is a key process. This process combines molten materials with different characteristics through a co-extrusion die to achieve comprehensive properties such as barrier and puncture resistance. During the multi-layer co-extrusion process, the melts of different materials need to converge at the die to form a stable interface. The quality of this interface plays a decisive role in the composite performance of the film. During the multi-layer co-extrusion process, the temperature of each layer of melt needs to be precisely controlled by independent heating and cooling systems to ensure temperature consistency at the die outlet for efficient bonding. However, due to the influence of the purity differences in material batches, there will be a significant deviation between the actual melt temperature and the target temperature at the same heating power. This deviation will cause unstable fluctuations in the melt temperature of each layer, directly affecting the stability of product quality. To ensure the efficient operation of the temperature control system, the melt temperature of each layer needs to be accurately predicted in real time to enable rapid power adjustment to ensure that the temperature can quickly respond to changes and avoid temperature lags and fluctuations.
[0003] When traditional exponential smoothing algorithms are used to predict the melt temperature of each layer, they will make smooth predictions about the future melt temperature based on the past melt temperatures of each layer, usually using a fixed smoothing coefficient. However, due to the fact that the melt temperature may change significantly when switching materials during the actual production of flexible films, the fixed smoothing parameters set at this time cannot adapt to the characteristics of local temperature changes, which may lead to prediction lags and delayed temperature control responses. As a result, it is impossible to ensure that each layer of melt reaches the target temperature before converging, and it is impossible to improve the interlayer bonding strength and overall film performance. Summary of the Invention
[0004] In order to improve the accuracy of predicting the melt temperature of each layer, the present invention proposes a temperature control system for flexible film production. The system includes the following modules: Data acquisition module, used to collect the melt temperature of each layer at each moment; A change degree acquisition module is used to obtain the local moment of each moment, and obtain the change degree of each layer melt temperature at each moment based on the difference between the melt temperature of each layer at each moment and the melt temperature of each layer at its local moment; A final smoothing parameter acquisition module is used to construct a random cut tree for each layer of melt temperature at each moment and its local moments, obtain the change reliability of each layer of melt temperature at each moment based on the tree depth of the random cut tree and the standard deviation of the timestamps corresponding to the data in all single subsets of the random cut tree; adjust the degree of change of each layer of melt temperature at each moment based on the change reliability to obtain the corrected degree of change of each layer of melt temperature at each moment; obtain the final smoothing parameter of each layer of melt temperature at each moment based on the corrected degree of change; and obtain the predicted melt temperature of each layer at the next moment based on the final smoothing parameter; The temperature control module is used to adjust the operating power of the temperature control equipment on each layer based on the predicted melt temperature of each layer at the next moment at each moment, so that the melt temperature of each layer approaches the target temperature.
[0005] The innovation of the present invention lies in first obtaining the degree of change of the melt temperature of each layer at each moment based on the previous difference of the melt temperature of each layer, which can reflect the change of the melt temperature of each layer; then obtaining the reliability of the change of the melt temperature of each layer at each moment based on the tree depth of the random cut tree and the standard deviation of the timestamps corresponding to the data in all single subsets of the random cut tree; correcting the degree of change according to the said change reliability, effectively filtering out occasional abnormal interference; finally, adaptively obtaining the final smoothing parameter according to the corrected degree of change, and predicting the melt temperature of each layer at the next moment, which can adapt to the characteristics of local temperature changes, solve the problems of prediction lag and temperature control response delay, improve temperature control accuracy, ensure that each layer of melt reaches the target temperature before merging, enhance the interlayer bonding strength and the overall performance of the film, thereby significantly improving product quality and consistency.
[0006] Preferably, obtaining the degree of change in the melt temperature of each layer at each moment includes: ; Where, Represents the degree of change in the melt temperature of layer a at the i-th moment; Represents the number of local moments at the i-th moment; Represents the timestamp corresponding to the i-th moment; Represents the timestamp corresponding to the jth local moment at the i-th moment; Represents the melt temperature value of layer a at the i-th moment; Represents the melt temperature value of layer a at the jth local moment at the i-th moment; exp() represents the exponential function with natural constants as the base; norm() represents the normalization function.
[0007] It can reflect the change of melt temperature in each layer, making it easier to adaptively obtain smoothing parameters in the subsequent process.
[0008] Preferably, constructing a random cut tree for the melt temperature of each layer at each moment and its local moments includes: The melt temperature of the a-th layer at the i-th moment and all its local moments is taken as the reference data set, the mean of the maximum and minimum values in the reference data set is taken as the segmentation threshold of the reference data set, and the reference data set is segmented to obtain the first layer of the random cut tree; the mean of the maximum and minimum values of each subset of the first layer is taken as the segmentation threshold of each subset of the first layer, and each subset of the first layer is segmented to obtain the second layer of the random cut tree; and so on, until each data of the reference data set is segmented out and the random cut tree of the melt temperature of the a-th layer at the i-th moment is obtained.
[0009] Preferably, obtaining the reliability of the change in the melt temperature of each layer at each moment includes: Get the depth of the random cut tree of the melt temperature of the a-th layer at the i-th moment and the single subset of the random cut tree of the melt temperature of the a-th layer at the i-th moment; ; Where, Represents the reliability of the change of the melt temperature of layer a at the i-th moment; The depth of the random cut tree representing the melt temperature of layer a at the i-th moment; The maximum value among the depths of the random cut tree representing the melt temperature of layer a at all local moments at moment i; The standard deviation of the timestamps corresponding to the data in all single subsets of the random cut tree representing the melt temperature of the a-th layer at the i-th moment; norm() represents the normalization function.
[0010] It can reflect the reliability of the temperature change of each layer, and then correct the degree of change to effectively filter out occasional abnormal interference.
[0011] Preferably, the step of obtaining the depth of the random cut tree of the melt temperature of the a-th layer at the i-th moment and the single subset of the random cut tree of the melt temperature of the a-th layer at the i-th moment includes: The number of layers of the random cut tree of the melt temperature of the a-th layer at the i-th moment is used as the depth of the random cut tree of the melt temperature of the a-th layer at the i-th moment; The subset containing only one data in the random cut tree of the melt temperature of the a-th layer at the i-th moment is recorded as a single subset of the random cut tree of the melt temperature of the a-th layer at the i-th moment.
[0012] Preferably, obtaining the corrected change degree of the melt temperature of each layer at each moment includes: ; Where, Represents the degree of correction change of the melt temperature of layer a at the i-th moment; Represents the reliability of the change of the melt temperature of layer a at the i-th moment; Represents the degree of change in the melt temperature of layer a at the i-th moment; Represents the stability index value of the melt temperature change in layer a at the i-th moment.
[0013] It can more accurately reflect the temperature change of each layer of melt.
[0014] Preferably, the step of obtaining the final smoothing parameter of the melt temperature of each layer at each moment includes: ; Where, represents the final smoothing parameter of the melt temperature of layer a at the i-th moment; Represents the degree of correction change of the melt temperature of layer a at the i-th moment; Represents the degree of change in the melt temperature of layer a at the i-th moment; Represents the preset smoothing parameter value.
[0015] Preferably, obtaining the predicted melt temperature of each layer at the next moment at each moment includes: According to the final smoothing parameter of the melt temperature of the a-th layer at the i-th moment, the exponential smoothing method is used to smooth the melt temperatures of all a-th layers at the i-th moment and its local moments to obtain the predicted melt temperature of the a-th layer at the i+1-th moment.
[0016] Adaptive smoothing parameters improve the accuracy of predictions.
[0017] Preferably, the operation power of the temperature control equipment of each layer is adjusted according to the predicted melt temperature of each layer at the next moment at each moment, including: The melt temperature of the a-th layer at all historical moments at the current moment and the power of the temperature control equipment of the a-th layer are input into the PID control algorithm to construct a control model. The predicted value of the melt temperature of the a-th layer at the next moment after the current moment is input into the control model to obtain the predicted power of the temperature control equipment of the a-th layer at the next moment after the current moment, and the power is adjusted so that the melt temperature of the a-th layer at the next moment after the current moment approaches the predetermined target value.
[0018] Preferably, collecting the melt temperature of each layer at each moment includes: The temperature is collected every 1 second by default, and high-precision melting temperature sensors are built into the flow channel areas of each layer. During the film production process, the melt temperature of each layer at each moment is collected in real time.
[0019] The present invention has the following technical effects: according to the previous difference of each layer melt temperature, the degree of change of each layer melt temperature at each moment is obtained, which can reflect the change of each layer melt temperature; then, according to the tree depth of the random cut tree and the standard deviation of the timestamps corresponding to the data in all single subsets of the random cut tree, the change reliability of each layer melt temperature at each moment is obtained; according to the said change reliability, the degree of change is corrected to effectively filter out occasional abnormal interference; finally, according to the corrected change degree, the final smoothing parameter is adaptively obtained, and the melt temperature of each layer at the next moment is predicted, which can adapt to the characteristics of local temperature changes, solve the problems of prediction lag and temperature control response delay, improve temperature control accuracy, ensure that each layer of melt reaches the target temperature before merging, enhance the interlayer bonding strength and the overall performance of the film, thereby significantly improving product quality and consistency. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] Figure 1 This is a system block diagram of a temperature control system for flexible film production according to an embodiment of the present invention. DETAILED DESCRIPTION
[0021] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.
[0022] The present invention provides a temperature control system for flexible film production. Figure 1 As shown, a temperature control system for flexible film production includes modules 101 to 104, which are described in detail below.
[0023] The data acquisition module 101 is used to collect the melt temperature of each layer at each moment.
[0024] It should be noted that in the actual production process of flexible films, multiple different material layers are formed into a multi-layer structure through co-extrusion technology. Among them, the flow channels corresponding to each layer have independent heating and air cooling systems to ensure that the melt of each layer reaches the target temperature at the die outlet. Therefore, it is necessary to obtain the melt temperature in the flow channel area of each layer. If the temperature is found to be substandard, the temperature control system parameters should be adjusted immediately to ensure that the target temperature is reached at the die outlet.
[0025] In the embodiment of the present invention, the temperature is preset to be collected once every 1 second, and a high-precision melting temperature sensor is built in the flow channel area of each layer. During the production process of the film, the melt temperature of each layer at each moment is collected in real time.
[0026] The variation degree acquisition module 102 is used to acquire the local moment of each moment, and acquire the variation degree of each layer melt temperature at each moment according to the difference between each layer melt temperature at each moment and each layer melt temperature at its local moment.
[0027] It should be noted that in multi-layer co-extrusion production, the temperature of each layer of melt needs to be precisely controlled by an independent heating and air cooling system to ensure temperature consistency at the die outlet, thereby achieving efficient bonding. However, due to the difference in purity of material batches, there is a significant deviation between the actual temperature of the melt and the target temperature at the same heating power. This deviation will cause unstable fluctuations in the temperature of each layer of melt, directly affecting the stability of product quality. In order to ensure the efficient operation of the temperature control system, the temperature of each layer of melt needs to be accurately predicted in real time so that the power can be adjusted quickly to ensure that the temperature can respond quickly to changes and avoid temperature lags and fluctuations.
[0028] When traditional exponential smoothing algorithms are used to predict the melt temperature of each layer, they will smooth the prediction of the future melt temperature based on the past melt temperatures of each layer, usually using a fixed smoothing coefficient. However, due to the fact that the melt temperature may change significantly when switching materials during the actual production of flexible films, the fixed smoothing parameters set at this time cannot adapt to the characteristics of local temperature changes, which may lead to prediction lag and delayed temperature control response.
[0029] It should be further explained that in the production process of flexible films, when the purity of the materials used is relatively stable, the melt temperature of each layer after processing under the fixed power of the heating or cooling system will be maintained near the target temperature. However, the purity of different batches of materials may be different, and the original stable state of the melt temperature of each layer will be broken, resulting in the melt temperature of the layer corresponding to the material with changed purity deviating significantly from the target temperature under the fixed power of the heating or cooling system, thereby destroying the stability of the melt temperature of this layer in time. Therefore, in order to promptly deal with abnormal temperature fluctuations caused by changes in material purity, the degree of change in the melt temperature of any layer at the current moment is quantified by comparing the difference between the melt temperature of any layer at the current moment and the melt temperature of the layer in a previous period of time. If the temperature deviates significantly, it means that the purity of the material of this layer has changed.
[0030] In an embodiment of the present invention, the preset number of local moments M=50. In other embodiments, implementers may preset the number of local moments according to the specific implementation method, and use the M moments between the i-th moment as the local moments of the i-th moment.
[0031] Get the degree of change in the melt temperature of layer a at the i-th moment: ; Where, Represents the degree of change in the melt temperature of layer a at the i-th moment; Represents the number of local moments at the i-th moment; Represents the timestamp corresponding to the i-th moment; Represents the timestamp corresponding to the jth local moment at the i-th moment; Represents the melt temperature value of layer a at the i-th moment; represents the melt temperature value of layer a at the jth local moment at the i-th moment; exp() represents the exponential function with the natural constant as the base; norm() represents the normalization function; The value of represents the difference between the melt temperature of layer a at the i-th moment and its j-th local moment. The larger the value, the greater the fluctuation amplitude of the melt temperature of layer a. Represents the weight of the difference between the melt temperature of the a-th layer at the i-th moment and its j-th local moment. If the interval between the i-th moment and its j-th local moment is shorter, it means that the j-th local moment of the i-th moment is closer to the i-th moment. If the melt temperature of the a-th layer at the i-th moment suddenly changes, the weight of the difference between the melt temperature of the a-th layer at the local moment close to the i-th moment and the melt temperature of the a-th layer at the i-th moment is greater. The larger the value of , the greater the difference between the melt temperature of the a-th layer at the i-th moment and each of its local moments. At this time, the greater the degree of change in the melt temperature of the a-th layer at the i-th moment.
[0032] The final smoothing parameter acquisition module 103 constructs a random cut tree for each layer of melt temperature at each moment and its local moment, obtains the change reliability of each layer of melt temperature at each moment based on the tree depth of the random cut tree and the temporal continuity of the abnormal data, adjusts the degree of change of each layer of melt temperature at each moment based on the change reliability, and obtains the corrected degree of change of each layer of melt temperature at each moment; obtains the final smoothing parameter of each layer of melt temperature at each moment based on the corrected degree of change, and then obtains the predicted melt temperature of each layer at the next moment after each moment.
[0033] It should be noted that environmental factors on actual film production lines (such as vibration from other equipment and electromagnetic interference from electrical components) can cause occasional sudden changes in the readings of high-precision melting temperature sensors, resulting in significant changes in the melt temperature of the corresponding layer. However, these sudden changes are random and instantaneous, and are not caused by changes in material purity, so they are pseudo-anomalies. In contrast, changes in material purity can cause sudden changes in melt temperature, which can intensify with continued processing of the batch. In other words, changes in material purity can cause persistent deviations in melt temperature. These changes are true anomalies, manifesting as significant and persistent changes in melt temperature over a period of time. Therefore, it is necessary to distinguish between pseudo-anomalies caused by environmental factors (such as equipment vibration and electromagnetic interference) and true anomalies caused by changes in material purity. The reliability of melt temperature changes can be determined by analyzing the distribution characteristics of melt temperature over time. If the melt temperature changes are affected by production line environmental factors, the reliability of the melt temperature changes is low. If the melt temperature changes are true anomalies caused by changes in material purity, the reliability of the melt temperature changes is high.
[0034] It should be further explained that the random cutting tree divides the data set into different subsets by segmenting the data set. This cutting process can effectively reflect the distribution characteristics of the data set and identify abnormal data in the data set. Therefore, it is necessary to construct a random cutting tree for analysis of the melt temperature over a period of time. It is known that the mutation of the melt temperature data caused by sporadic mutations often manifests as isolated characteristics. Therefore, if there are sporadic mutations in the melt temperature over a period of time, the fewer melt data with sporadic mutations will result in a faster segmentation process when constructing the random cutting tree. The depth of the obtained random cutting tree is smaller and the segmented melt data with sporadic mutations are more frequent in time series. The more discontinuous it is, the lower the reliability of the melt temperature change is; it is known that changes in material purity will cause continuous deviations in the melt temperature. Therefore, if there is a continuous deviation in the melt temperature over a period of time, then the more continuously deviated melt temperature data there are, the slower the segmentation process will be when constructing the random cut tree. The greater the depth of the obtained random cut tree and the more continuous the segmented continuously deviated melt temperature data are in time series, the higher the reliability of the melt temperature change is. Therefore, the present invention constructs a random cut tree for the melt temperature over a period of time, and quantifies the reliability of the melt temperature change according to the depth of the random cut tree and the time series continuity of the abnormal data in the random cut tree.
[0035] In an embodiment of the present invention, the melt temperature of the a-th layer at the i-th moment and all its local moments is used as a reference data set, the average of the maximum and minimum values in the reference data set is used as the segmentation threshold of the reference data set, and the reference data set is segmented to obtain the first layer of a random cut tree; the average of the maximum and minimum values of each subset of the first layer is used as the segmentation threshold of each subset of the first layer, and each subset of the first layer is segmented to obtain the second layer of the random cut tree; and so on, until each data of the reference data set is segmented and stopped, obtaining a random cut tree of the melt temperature of the a-th layer at the i-th moment; The number of layers of the random cut tree of the melt temperature of the a-th layer at the i-th moment is used as the depth of the random cut tree of the melt temperature of the a-th layer at the i-th moment; Denote the subset containing only one data in the random cut tree of the melt temperature of the a-th layer at the i-th moment as a single subset of the random cut tree of the melt temperature of the a-th layer at the i-th moment, and obtain the timestamp corresponding to the data in each single subset of the random cut tree of the melt temperature of the a-th layer at the i-th moment; It should be noted that in constructing a random cut tree, abnormal data usually appears as isolated subsets, and in multi-layer co-extrusion production, the melt temperature of each layer must be consistent with the target temperature. Therefore, a stable melt temperature is more likely to be divided into a subset in a random cut tree. Therefore, a subset containing only one data in a random cut tree is more likely to be abnormal data, that is, abnormal data caused by a mutation in the melt temperature data due to an occasional mutation, or abnormal data caused by a continuous deviation in the melt temperature due to a change in material purity. Therefore, the temporal continuity of the abnormal data is quantified according to the standard deviation of the timestamps corresponding to the data in all single subsets of the random cut tree.
[0036] Get the reliability of the change of the melt temperature of layer a at the i-th moment: ; Where, Represents the reliability of the change of the melt temperature of layer a at the i-th moment; The depth of the random cut tree representing the melt temperature of layer a at the i-th moment; The maximum value among the depths of the random cut tree representing the melt temperature of layer a at all local moments at moment i; Represents the standard deviation of the timestamps corresponding to the data in all single subsets of the random cut tree of the melt temperature of the a-th layer at the i-th moment; norm() represents the normalization function; The smaller the value of , the faster the segmentation process is when constructing the random cut tree of the melt temperature of the a-th layer at the i-th moment, indicating that there are fewer abnormal data in the melt temperature of the a-th layer at the i-th moment and all its local moments, that is, there are abnormal data caused by mutations in the melt temperature data caused by occasional mutations; The larger the value of , the slower the segmentation process is when constructing the random cut tree of the melt temperature of the a-th layer at the i-th moment, indicating that there are more abnormal data in the melt temperature of the a-th layer at the i-th moment and all its local moments, that is, there are abnormal data caused by the continuous deviation of the melt temperature due to the change of material purity; like The larger the value of , the greater the discreteness of the timestamps corresponding to the data in all single subsets of the random cut tree of the melt temperature of the a-th layer at the i-th moment, indicating that the data in the single subset are discontinuous in time series, that is, the mutation of the melt temperature data caused by occasional mutations is manifested as discontinuity in time series; The smaller the value of , the greater the clustering of the timestamps corresponding to the data in all single subsets of the random cut tree of the melt temperature of the a-th layer at the i-th moment, indicating that the data in the single subset are continuous in time series, that is, the deviation of the melt temperature caused by the change in material purity is continuous in time series; therefore The smaller the value of The larger the value of , the more it indicates that the change in melt temperature is affected by the environmental factors of the production line. At this time, the reliability of the change in the melt temperature of the a-th layer at the i-th moment is lower.
[0037] It should be noted that if the degree of change of the melt temperature of layer a at any time is small, the melt temperature is stable and there is no need to correct the degree of change; If the reliability of the change of the melt temperature of layer a at any moment is lower, the change of the melt temperature at this moment is due to the influence of the environmental factors of the production line. Since the degree of change of the melt temperature of layer a at this moment is large, it is necessary to reduce the degree of change of the melt temperature of layer a at this moment according to the reliability of the change of the melt temperature of layer a at this moment, and obtain the corrected degree of change of the melt temperature of layer a at this moment; if the reliability of the change of the melt temperature of layer a at any moment is higher, the change of the melt temperature at this moment is due to the change of material purity causing the melt temperature to deviate continuously. Therefore, the degree of change of the melt temperature of layer a at this moment is large. It is necessary to increase the degree of change of the melt temperature of layer a at this moment according to the reliability of the change of the melt temperature of layer a at this moment, to ensure that the abnormal section performs outstandingly.
[0038] In the embodiment of the present invention, the N moments before the i-th moment are taken as the historical moments of the i-th moment, and the minimum value of the change degree of the melt temperature of the a-th layer at all historical moments of the i-th moment is taken as the stability index value of the melt temperature change of the a-th layer at the i-th moment; the preset number of historical moments N=600, and in other embodiments, the implementer may preset the value of the number of historical moments according to the specific implementation method; Get the corrected change in the melt temperature of layer a at the i-th moment: ; Where, Represents the degree of correction change of the melt temperature of layer a at the i-th moment; Represents the reliability of the change of the melt temperature of layer a at the i-th moment; Represents the degree of change in the melt temperature of layer a at the i-th moment; represents the stability index value of the melt temperature change of layer a at the i-th moment; When , it means that the change degree of the melt temperature of the a-th layer at the i-th moment is stable and does not need to be adjusted. The change degree of the melt temperature of the a-th layer at the i-th moment can be used as the corrected change degree of the melt temperature of the a-th layer at the i-th moment; When , it means that the degree of change of the melt temperature of layer a at the i-th moment deviates from the stability index value, which means that there may be abnormal changes in the melt temperature of layer a at the i-th moment. At this time, if the reliability of the change of the melt temperature of layer a at the i-th moment is higher, it means that the change of the melt temperature of layer a at the i-th moment is more likely to be caused by the change of material purity. It is necessary to significantly increase the degree of change of the melt temperature of layer a at the i-th moment to ensure that the abnormal section is outstanding; on the contrary, if the reliability of the change of the melt temperature of layer a at the i-th moment is lower, it means that the change of the melt temperature of layer a at the i-th moment is more likely to be a pseudo-abnormality caused by occasional mutation. At this time, it is necessary to appropriately increase the degree of change of the melt temperature of layer a at the i-th moment.
[0039] It should be noted that in order to ensure the efficient operation of the temperature control system, the temperature of each layer of melt needs to be accurately predicted in real time so as to quickly adjust the power and ensure that the temperature can respond quickly to changes to avoid temperature lag and fluctuation. When the traditional exponential smoothing algorithm is used to predict the melt temperature of each layer, the future melt temperature will be smoothed based on the past melt temperature of each layer, and a fixed smoothing coefficient will usually be used; however, due to the material switching during the actual production of flexible films, the temperature of the melt may change significantly. At this time, the fixed smoothing parameters set cannot adapt to the characteristics of local temperature changes, which may cause prediction lag and temperature control response delay. Therefore, the present invention adaptively obtains the final smoothing parameters of the melt temperature of each layer at each moment according to the degree of correction change of the melt temperature of each layer at each moment.
[0040] In this embodiment of the present invention, the final smoothing parameter of the melt temperature of the a-th layer at the i-th moment is obtained: ; Where, represents the final smoothing parameter of the melt temperature of layer a at the i-th moment; Represents the degree of correction change of the melt temperature of layer a at the i-th moment; Represents the degree of change in the melt temperature of layer a at the i-th moment; Represents the preset smoothing parameter value. In the embodiment of the present invention, the preset =0.05; The larger the value of , the greater the degree of correction of the melt temperature of layer a at the i-th moment is, the greater the degree of change before correction. This indicates that the change of the melt temperature of layer a at all local moments of the i-th moment is due to continuous anomalies caused by changes in material purity. It is necessary to significantly increase the smoothing parameter to ensure that the recent moment data changes are of high concern, thereby improving the prediction accuracy. On the contrary, The smaller the value of , the more likely it is that the change in the melt temperature of the a-th layer in all local moments at the i-th moment may be caused by occasional mutations. In this case, it is maintained near the preset smoothing parameter to avoid excessive attention resulting in low prediction accuracy.
[0041] According to the final smoothing parameter of the melt temperature of the a-th layer at the i-th moment, the exponential smoothing method is used to smooth the melt temperatures of all a-th layers at the i-th moment and its local moments to obtain the predicted melt temperature of the a-th layer at the i+1-th moment.
[0042] The temperature control module 104 is used to adjust the operating power of the temperature control equipment of each layer according to the predicted melt temperature of each layer at the next moment at each moment, so that the melt temperature of each layer approaches the target temperature value.
[0043] In an embodiment of the present invention, the melt temperature of the a-th layer at all historical moments at the current moment and the power of the temperature control equipment of the a-th layer are input into the PID control algorithm to construct a control model, and the predicted value of the melt temperature of the a-th layer at the next moment after the current moment is input into the PID control model to obtain the predicted power of the temperature control equipment of the a-th layer at the next moment after the current moment, and adjust it so that the melt temperature of the a-th layer at the next moment after the current moment approaches the predetermined target value.
[0044] It should be noted that inputting historical moment data into a PID control algorithm to construct a control model is an existing technology and will not be described in detail.
[0045] The above are all preferred embodiments of the present invention, and are not intended to limit the scope of protection 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 scope of protection of the present invention.
Claims
1. A temperature control system for flexible film production, characterized in that: Includes the following modules: Data acquisition module, used to collect the melt temperature of each layer at each moment; A change degree acquisition module is used to obtain the local moment of each moment, and obtain the change degree of each layer melt temperature at each moment based on the difference between the melt temperature of each layer at each moment and the melt temperature of each layer at its local moment; A final smoothing parameter acquisition module is used to construct a random cut tree for each layer of melt temperature at each moment and its local moments, obtain the change reliability of each layer of melt temperature at each moment based on the tree depth of the random cut tree and the standard deviation of the timestamps corresponding to the data in all single subsets of the random cut tree; adjust the degree of change of each layer of melt temperature at each moment based on the change reliability to obtain the corrected degree of change of each layer of melt temperature at each moment; obtain the final smoothing parameter of each layer of melt temperature at each moment based on the corrected degree of change; and obtain the predicted melt temperature of each layer at the next moment based on the final smoothing parameter; The temperature control module is used to adjust the operating power of the temperature control equipment on each layer based on the predicted melt temperature of each layer at the next moment at each moment, so that the melt temperature of each layer approaches the target temperature.
2. A temperature control system for flexible film production according to claim 1, characterized in that: The step of obtaining the degree of change in the melt temperature of each layer at each moment includes: ； Where, Represents the degree of change in the melt temperature of layer a at the i-th moment; Represents the number of local moments at the i-th moment; Represents the timestamp corresponding to the i-th moment; Represents the timestamp corresponding to the jth local moment at the i-th moment; Represents the melt temperature value of layer a at the i-th moment; Represents the melt temperature value of layer a at the jth local moment at the i-th moment; exp() represents the exponential function with natural constants as the base; norm() represents the normalization function.
3. A temperature control system for flexible film production according to claim 1, characterized in that: The random cut tree is constructed for each layer melt temperature at each moment and its local moment, including: The melt temperature of the a-th layer at the i-th moment and all its local moments is taken as the reference data set, the mean of the maximum and minimum values in the reference data set is taken as the segmentation threshold of the reference data set, and the reference data set is segmented to obtain the first layer of the random cut tree; the mean of the maximum and minimum values of each subset of the first layer is taken as the segmentation threshold of each subset of the first layer, and each subset of the first layer is segmented to obtain the second layer of the random cut tree; and so on, until each data of the reference data set is segmented out and the random cut tree of the melt temperature of the a-th layer at the i-th moment is obtained.
4. A temperature control system for flexible film production according to claim 1, characterized in that: The obtaining of the reliability of the change of the melt temperature of each layer at each moment includes: Get the depth of the random cut tree of the melt temperature of the a-th layer at the i-th moment and the single subset of the random cut tree of the melt temperature of the a-th layer at the i-th moment; ； Where, Represents the reliability of the change of the melt temperature of layer a at the i-th moment; The depth of the random cut tree representing the melt temperature of layer a at the i-th moment; The maximum value among the depths of the random cut tree representing the melt temperature of layer a at all local moments at moment i; The standard deviation of the timestamps corresponding to the data in all single subsets of the random cut tree representing the melt temperature of the a-th layer at the i-th moment; norm() represents the normalization function.
5. A temperature control system for flexible film production according to claim 4, characterized in that: The method of obtaining the depth of the random cut tree of the melt temperature of the a-th layer at the i-th moment and the single subset of the random cut tree of the melt temperature of the a-th layer at the i-th moment includes: The number of layers of the random cut tree of the melt temperature of the a-th layer at the i-th moment is used as the depth of the random cut tree of the melt temperature of the a-th layer at the i-th moment; The subset containing only one data in the random cut tree of the melt temperature of the a-th layer at the i-th moment is recorded as a single subset of the random cut tree of the melt temperature of the a-th layer at the i-th moment.
6. A temperature control system for flexible film production according to claim 1, characterized in that: The obtaining of the corrected change degree of the melt temperature of each layer at each moment includes: ； Where, Represents the degree of correction change of the melt temperature of layer a at the i-th moment; Represents the reliability of the change of the melt temperature of layer a at the i-th moment; Represents the degree of change in the melt temperature of layer a at the i-th moment; Represents the stability index value of the melt temperature change in layer a at the i-th moment.
7. A temperature control system for flexible film production according to claim 1, characterized in that: The method of obtaining the final smoothing parameter of the melt temperature of each layer at each moment includes: ； Where, represents the final smoothing parameter of the melt temperature of layer a at the i-th moment; Represents the degree of correction change of the melt temperature of layer a at the i-th moment; Represents the degree of change in the melt temperature of layer a at the i-th moment; Represents the preset smoothing parameter value.
8. The temperature control system for flexible film production according to claim 1, characterized in that: The step of obtaining the predicted melt temperature of each layer at the next moment at each moment includes: According to the final smoothing parameter of the melt temperature of the a-th layer at the i-th moment, the exponential smoothing method is used to smooth the melt temperatures of all a-th layers at the i-th moment and its local moments to obtain the predicted melt temperature of the a-th layer at the i+1-th moment.
9. The temperature control system for flexible film production according to claim 1, characterized in that: The operation power of the temperature control equipment of each layer is adjusted according to the predicted melt temperature of each layer at the next moment at each moment, including: The melt temperature of the a-th layer at all historical moments at the current moment and the power of the temperature control equipment of the a-th layer are input into the PID control algorithm to construct a control model. The predicted value of the melt temperature of the a-th layer at the next moment after the current moment is input into the control model to obtain the predicted power of the temperature control equipment of the a-th layer at the next moment after the current moment, and the power is adjusted so that the melt temperature of the a-th layer at the next moment after the current moment approaches the predetermined target value.
10. The temperature control system for flexible film production according to claim 1, characterized in that: The collecting of the melt temperature of each layer at each moment includes: The temperature is collected every 1 second by default, and high-precision melting temperature sensors are built into the flow channel areas of each layer. During the film production process, the melt temperature of each layer at each moment is collected in real time.