Deep learning-based intelligent control method and system for hot melt coating production line

Abstract

The invention discloses a hot melt coating production line intelligent control method and system based on deep learning, and relates to the technical field of production line control. Comprising the steps that based on basic information, a flow rate obtaining formula is set to obtain hot-melt coating flow rate data corresponding to the basic information, and flow data items are obtained; based on the operation information, a utilization rate acquisition formula is set to acquire hot-melt coating utilization rate data corresponding to the operation information, and a utilization data item is obtained; according to the method, the comparison result of the flow data item and the use data item is used as training data, the corresponding relation between the flow data item and the use data item is captured through deep learning, the use rate model of the hot-melt coating is generated, the preset use position of the target hot-melt coating is obtained, the demand data is obtained, and the demand data is used as input data. And data feedback information corresponding to the demand data is obtained, reverse adjustment is performed on the production line parameters according to a deep learning result, and intelligent control is performed on the production line.

Description

Technical Field

[0001] The present invention relates to the technical field of production line control, and specifically to an intelligent control method and system for a hot-melt coating production line based on deep learning. Background Art

[0002] Hot-melt coating is a kind of coating that can be melted at high temperature and coated on the surface of various substrates. This kind of coating usually consists of resin, filler and other additives, and has good adhesion, wear resistance and corrosion resistance. The coating melts after being heated to a certain temperature and solidifies after being coated and cooled. When producing hot-melt coating, a production line is usually used for production. The production line of hot-melt coating generally consists of multiple steps and equipment to ensure the high quality and consistency of the coating, including raw material preparation. Raw materials such as resin, filler, pigment and additive need to be prepared and weighed according to the formula, mixed, melted and homogenized, cooled and solidified, crushed, screened and packaged. In the actual production process, each link in the production needs to be controlled.

[0003] A production control method and system for powder coating with the patent publication number CN116719295A obtains powder coating processing material information; matches parameters from a production control parameter library according to the powder coating processing material information to determine matching control parameters and a production process flow, wherein the production process flow has a mapping relationship with the matching control parameters; based on the matching control parameters, the production process flow, and the powder coating processing material information, conducts production simulation through a digital factory and monitors the processing quality of the simulated production process to obtain a production process quality monitoring set; determines whether the production process quality in the production process quality monitoring set meets the process quality requirements; adjusts the process control parameters that do not meet the process quality requirements until the processing result prediction results obtained from all processes meet the quality requirements of each process, and then determines the control parameters corresponding to all processes to obtain the entire production control parameters, generates corresponding control information, and provides reference opinions for production control, achieving the technical effects of improving the accuracy of production processing control and ensuring the quality of powder coating processing results.

[0004] When the above-mentioned and similar technical solutions melt and spray the produced hot-melt coating on the object to be sprayed, due to the influence of environmental factors, the hot-melt coating produced by different production processes during the production process cannot fully achieve the maximum utilization rate in actual use, and it needs to be adjusted according to the actual use scenario to prevent the actual use effect of the hot-melt coating from not being maximized, resulting in a poor use effect. Summary of the Invention

[0005] The object of the present invention is to provide an intelligent control method and system for a hot-melt coating production line based on deep learning to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: An intelligent control method for a hot-melt coating production line based on deep learning:

[0007] Obtain the basic information of the target hot-melt coating production line to obtain calibration information items, and the calibration information items are used to represent the marker information corresponding to the production line;

[0008] Based on the calibration information items, obtain the basic information and operation information of the hot-melt coating to obtain a comparison relationship set, and the comparison relationship set includes at least one set of corresponding relationships between the basic information and the operation information;

[0009] Including:

[0010] Based on the basic information, set a flow rate acquisition formula to obtain the hot-melt coating flow rate data corresponding to the basic information, and obtain a flow data item, and the flow data item is used to represent the attribute information of the hot-melt coating obtained under different production processes;

[0011] Based on the operation information, set a utilization rate acquisition formula to obtain the hot-melt coating utilization rate data corresponding to the operation information, and obtain a usage data item, and the usage data item is used to represent the usage feedback information of the hot-melt coating under different operation environments;

[0012] Use the comparison result of the flow data item and the usage data item as training data, and capture the corresponding relationship and corresponding method between the flow data item and the usage data item through deep learning to generate a utilization rate model of the hot-melt coating;

[0013] Obtain the predetermined usage position of the target hot-melt coating to obtain a predetermined usage item, obtain demand data based on the predetermined usage item, input the demand data as input data into the utilization rate model, obtain data feedback information corresponding to the demand data, the data feedback information corresponds to the basic information, and reversely adjust the production line parameters according to the deep learning result.

[0014] Furthermore, the basic information of the hot-melt coating includes resin content, melting temperature, stirring time, total amount of glass beads added, plasticizer ratio, and curing time, and the flow rate acquisition formula is:

[0015] ;

[0016] Where is the flow rate, is the equivalent flow channel radius of the coating in the molten state, is the melting temperature, is the mass ratio of the plasticizer, is the mass ratio of the resin, is the addition amount of glass beads, is the stirring time, is the curing time, is the reference viscosity, and and are experimental fitting coefficients, and and and are reference values.

[0017] Furthermore, the operation information of the hot-melt coating includes the average temperature, average humidity, coating thickness, and usage duration. The formula for obtaining the usage rate is:

[0018] ;

[0019] where is the coating consumption per unit area, i.e., the usage rate, is the initial set thickness of the coating, is the real-time thickness of the coating, is the average temperature of the spraying environment, is the average relative humidity of the environment, is the usage duration, and are material characteristic coefficients, is a minimum constant.

[0020] Furthermore, the method for obtaining the equivalent flow channel radius of the coating in the molten state is:

[0021] Create a formula for obtaining the equivalent flow channel radius of the coating in the molten state:

[0022] ;

[0023] where is the reference temperature.

[0024] Furthermore, the method for obtaining the average temperature of the spraying environment and the average relative humidity of the environment includes:

[0025] Based on the usage duration, obtain the target deadline value. Based on the target deadline value, obtain the environmental temperature and environmental humidity data to obtain the environmental temperature set and the environmental humidity set;

[0026] Based on the requested deadline, respectively obtain the average values of the environmental temperature set and the environmental humidity set to obtain the average temperature of the spraying environment and the average relative humidity of the environment.

[0027] Furthermore, the hot melt paint production line stores initial information, and the method for obtaining the calibration information item includes: the initial information includes an initial code of the hot melt paint production line, and the hot melt paint production line is distinguished based on the initial code to obtain the calibration information item.

[0028] Furthermore, the method for obtaining the control relationship set includes:

[0029] Based on the basic information, the hot melt coatings are arranged in sequence numbers to obtain sequence distribution items;

[0030] Based on the corresponding information of the sequence number distribution items stored in the job information, the corresponding relationship between the basic information and the job information is obtained, and then the comparison relationship set is obtained.

[0031] Furthermore, the method for obtaining the usage rate model of the hot melt coating includes:

[0032] Based on the correspondence between the flow data item, the basic information and the operation information, the input attribute item is obtained, and the output attribute item is used based on the use data item;

[0033] Create a data classification table, classify the input attribute items and the output attribute items, and obtain the corresponding result set;

[0034] Select the target deep training model, set the split ratio, and split the corresponding result set based on the split ratio to obtain the training set and the test set;

[0035] The training set is used as the training data to train the target deep training model and obtain the usage rate model;

[0036] The usage model is tested and optimized using the test set.

[0037] Furthermore, the method for obtaining the demand data includes:

[0038] Set the request threshold to obtain the request time item, where the request threshold is a fixed time value;

[0039] Based on the request threshold, the temperature information and humidity information of the scheduled usage item are obtained through the target device to obtain the demand data.

[0040] Furthermore, a hot melt coating production line intelligent control system based on deep learning uses the above-mentioned hot melt coating production line intelligent control method based on deep learning, including:

[0041] Production marking module: obtains the basic information of the target hot-melt coating production line and obtains the calibration information item, which is used to represent the marking information corresponding to the production line;

[0042] Information acquisition module: obtains basic information and operation information of hot-melt coatings and obtains a reference relationship set;

[0043] Information processing module: Based on the basic information, set a flow rate acquisition formula to obtain the hot melt coating flow rate data corresponding to the basic information, and obtain the flow data items. Based on the operation information, set a utilization rate acquisition formula to obtain the hot melt coating utilization rate data corresponding to the operation information, and obtain the usage data items. Use the comparison result of the flow data items and the usage data items as the training data, and capture the corresponding relationship and corresponding method between the flow data items and the usage data items through deep learning to generate a utilization rate model for the hot melt coating;

[0044] Feedback adjustment module: Obtain the predetermined usage position of the target hot melt coating to obtain the predetermined usage items. Based on the predetermined usage items, obtain the demand data. Use the demand data as the input data and input it into the utilization rate model to obtain the data feedback information corresponding to the demand data. The data feedback information corresponds to the basic information, and the production line parameters are adjusted reversely according to the deep learning result.

[0045] Compared with the prior art, the beneficial effects of the present invention are:

[0046] This intelligent control method and system for a hot melt coating production line based on deep learning respectively set a flow rate acquisition formula to obtain the hot melt coating flow rate data corresponding to the basic information, and obtain the flow data items, and set a utilization rate acquisition formula to obtain the hot melt coating utilization rate data corresponding to the operation information, and obtain the usage data items. Use the comparison result of the flow data items and the usage data items as the training data, and capture the corresponding relationship and corresponding method between the flow data items and the usage data items through deep learning to generate a utilization rate model for the hot melt coating. Obtain the predetermined usage position of the target hot melt coating to obtain the predetermined usage items. Based on the predetermined usage items, obtain the demand data. Use the demand data as the input data and input it into the utilization rate model to obtain the data feedback information corresponding to the demand data, so as to adjust the production line parameters reversely according to the deep learning result, realizing the intelligent control effect of the production line. Description of the Drawings

[0047] Figure 1 It is a schematic diagram of the overall process of the present invention;

[0048] Figure 2 It is a schematic diagram of the comparison relationship set of the present invention;

[0049] Figure 3 It is a schematic diagram of the initial set thickness and real-time thickness of the coating of the present invention;

[0050] Figure 4 It is a schematic diagram of the training set and test set of the present invention. Detailed Embodiments

[0051] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0052] In the actual application process, due to the influence of many environmental factors such as environmental temperature and humidity, the hot-melt coatings produced by different production processes often cannot achieve the theoretically optimal use efficiency, thereby affecting the final spraying effect. The environmental temperature is a key factor affecting the viscosity, fluidity, and curing speed of the hot-melt coating. At a lower environmental temperature, the viscosity of the hot-melt coating will increase significantly, and the fluidity will decrease, resulting in a poor atomization effect during the spraying process, uneven coating thickness, and even nozzle clogging. On the other hand, the lower temperature will also prolong the curing time of the coating, making the coating more vulnerable to contamination or scratching. On the contrary, although a too high environmental temperature is beneficial to the fluidity of the coating, it may cause the coating to cure too quickly, affecting its adhesion to the substrate, and even problems such as coating blistering and cracking may occur. Humidity is another important factor affecting the curing quality of the hot-melt coating. A high-humidity environment will delay the curing speed of the hot-melt coating. Especially for some water-based hot-melt coatings, too high humidity may cause phenomena such as coating whitening and adhesion. In addition, high humidity may also promote the accumulation of water vapor inside the coating, reducing the water resistance and durability of the coating. Therefore, the use effects of the molten coatings produced by different production processes are also different in the actual use process and need to be adjusted according to the actual use scenario. And the technical solution provided in this period obtains the hot-melt coating flow rate data corresponding to the basic information by respectively setting the flow rate acquisition formula to obtain the flow data item, and sets the usage rate acquisition formula to obtain the hot-melt coating usage rate data corresponding to the operation information to obtain the usage data item. Using the comparison result of the flow data item and the usage data item as the training data, capturing the corresponding relationship and corresponding method between the flow data item and the usage data item through deep learning, generating the usage rate model of the hot-melt coating, obtaining the predetermined use position of the target hot-melt coating to obtain the predetermined use item, obtaining the demand data based on the predetermined use item, using the demand data as the input data and inputting it into the usage rate model to obtain the data feedback information corresponding to the demand data, thereby reversely adjusting the production line parameters according to the deep learning result, achieving the intelligent control effect of the production line, as Figure 1 - Figure 2 shown, including steps S100 - S700.

[0053] Step S100: Obtain the basic information of the target hot-melt coating production line to obtain the calibration information item.

[0054] It should be noted that the calibration information item is used to represent the marking information corresponding to the production line. Initial information is stored on the hot-melt coating production line. The method for obtaining the calibration information item includes: The initial information includes the initial code of the hot-melt coating production line. The hot-melt coating production lines are differentiated based on the initial code to obtain the calibration information item.

[0055] Specifically, in a production site, there is more than one production line, and each production line produces hot-melt coating. Therefore, it is necessary to classify and mark each production line. The production lines are differentiated through the initial code set on the production line, and then the calibration information item is obtained. The calibration information item is only for differentiating the production lines.

[0056] Step S200: Based on the calibration information item, obtain the basic information and operation information of the hot-melt coating to obtain a set of control relationships.

[0057] It should be noted that as Figure 2 shown, the set of control relationships includes at least one set of corresponding relationships between basic information and operation information. The method for obtaining the set of control relationships includes: Based on the basic information, arrange the serial numbers of the hot-melt coating to obtain a serial number distribution item; Based on the information corresponding to the serial number distribution item stored in the operation information, obtain the corresponding relationship between the basic information and the operation information, and then obtain the set of control relationships.

[0058] Specifically, a production line produces different batches of hot-melt coating, and different batches of hot-melt coating are respectively used in different situations. Therefore, it is necessary to first arrange the serial numbers of the produced hot-melt coating according to the basic information during the production process and correspond to the subsequent actual operation information.

[0059] Step S300: Based on the basic information, set a flow rate acquisition formula to obtain the hot-melt coating flow rate data corresponding to the basic information, and obtain a flow data item.

[0060] It should be noted that as Figure 2 shown, the flow data item is used to represent the attribute information of the hot-melt coating obtained under different production processes. The basic information of the hot-melt coating includes resin content, melting temperature, stirring time, total glass bead addition amount, plasticizer ratio, and curing time. The flow rate acquisition formula is:

[0061] ;

[0062] Where is the flow rate, is the equivalent flow channel radius of the coating in the molten state, is the melting temperature, is the mass ratio of the plasticizer, is the mass ratio of the resin, is the glass bead addition amount, is the stirring time, is the curing time, is the reference viscosity, , , are the experimental fitting coefficients, , , , are the reference values;

[0063] The method for obtaining the equivalent flow channel radius of the coating in the molten state is as follows:

[0064] Create a formula for obtaining the equivalent flow channel radius of the coating in the molten state:

[0065] ;

[0066] where is the reference temperature.

[0067] Specifically, through the set flow rate acquisition formula, the finished product quality of the hot-melt coating under different production processes is reflected. The unit of the flow rate is g / 10min, the unit of the equivalent flow channel radius of the coating in the molten state is mm, and it is related to the temperature. Among them, the experimental fitting coefficients , , are 0.12, 0.08, and -0.15 respectively. Through orthogonal experiments, in the flow rate acquisition formula, as the positive driving factor in the numerator term, when the melting temperature increases, the flow channel radius increases significantly through , and the linear term strengthens the thermal motion. The increase in the proportion of plasticizer directly reduces the intermolecular force, showing a linear gain . Among the resistance factors in the denominator term, the increase in the resin proportion leads to an increase in viscosity, showing a linear relationship . The increase in the glass bead addition amount forms a flow barrier, and the resistance is linearly amplified through . The extension of the curing time triggers a pre-crosslinking reaction, reflecting a non-linear increase in viscosity. The increase in the stirring time improves the dispersion uniformity, but the marginal effect decreases, conforming to the logarithmic function law. The reference values , , , are 35%, 300 kg / m³, 30 s, and 5 min respectively. The reference temperature is 190 °C. Among them, all the numerical values are the conventional values of the hot-melt coating in the production process and can be obtained according to the past production process of the hot-melt coating.

[0068] Example 1

[0069] In a specific implementation process, when a certain production line A produces hot-melt coatings, the set production process parameters are: the melting temperature is set at 200 °C, the mass ratio of the plasticizer is 5%, the mass ratio of the resin is 40%, the addition amount of glass beads is 350 kg / m³, the stirring time is 8 min, and the curing time is 25 s. At this time, according to the formula for obtaining the equivalent flow channel radius of the coating in the molten state:

[0070] ;

[0071] Calculation gives:

[0072] ;

[0073] Substitute it into the formula:

[0074] ;

[0075] After calculation:

[0076] ;

[0077] Finally, the flow rate Q≈85.3 g / 10 min is obtained.

[0078] Step S400: Based on the operation information, set a usage rate acquisition formula to obtain the hot-melt coating usage rate data corresponding to the operation information, and obtain usage data items.

[0079] It should be noted that, as Figure 2 - Figure 3 shown, the usage data items are used to represent the usage feedback information of the hot-melt coating in different operation environments. The operation information of the hot-melt coating includes the average temperature, average humidity, coating thickness, and usage duration. The usage rate acquisition formula is:

[0080] ;

[0081] Where is the coating consumption per unit area, that is, the usage rate, is the initial set thickness of the coating, is the real-time thickness of the coating, is the average temperature of the spraying environment, is the average relative humidity of the environment, is the usage duration, and are the material characteristic coefficients, is a very small constant.

[0082] Specifically, and as the material characteristic coefficients are 0.15 and 0.12 respectively, obtained based on experimental calibration. As the very small constant This is to prevent the denominator from being zero, and its value is 0.01.

[0083] Example Two

[0084] In the specific implementation process, relevant data of the hot-melt coating sprayed on a certain road is obtained. Among them, the initial set thickness of the coating is 2.0 mm, the real-time thickness of the coating detected in real time is 1.9 mm, the average spraying environment temperature is 25 °C, the average environmental relative humidity is 60%, and the usage duration is 10 min. At this time, according to the usage rate acquisition formula:

[0085] ;

[0086] Calculated as:

[0087] ;

[0088] That is, the usage rate is 4.89 kg / ㎡.

[0089] Step S500: Use the comparison result of the flow data item and the usage data item as training data, capture the corresponding relationship and corresponding method between the flow data item and the usage data item through deep learning, and generate a usage rate model of the hot-melt coating.

[0090] It should be noted that the flow rate data calculated for the hot-melt coatings produced under different production processes are different, and the usage rate data calculated for the hot-melt coatings used in different working environments are also different. When the hot-melt coating produced under a certain production process acts on a certain working environment, the flow rate data and the usage rate data are also correlated. And since the quantity of the hot-melt coating produced by a production line is definitely not 1, the quantity of the flow rate data is multiple, and the quantity of the usage rate data is also multiple. Therefore, the comparison result of the flow data item and the usage data item can be used as training data, and frameworks such as TensorFlow or PyTorch are used to build and train the model, and the comparison result is used as training data for training, and the dataset is iteratively trained multiple times to optimize the model parameters.

[0091] Specifically, such as Figure 4As shown in the figure, the method for obtaining the utilization rate model of the hot-melt coating includes: obtaining the input attribute items based on the correspondence relationship of the flow data items, basic information, and operation information, and using the usage data items as the output attribute items; creating a data classification table, classifying and corresponding the input attribute items and the output attribute items to obtain the corresponding result set; selecting a target deep training model, where the target deep training model is a feedforward neural network, setting a splitting ratio, the splitting ratio is 70%, splitting the corresponding result set based on the splitting ratio to obtain a training set and a test set, the training set is 70% of it, and the test set is the other 30%; using the training set as the training data to train the target deep training model to obtain the utilization rate model; using the test set to test and optimize the utilization rate model.

[0092] Embodiment III

[0093] In the specific implementation process, when a production line A produces hot-melt coating, a total of ten groups of hot-melt coatings are produced, numbered 1-10 respectively. The production process parameters set for No. 1 are: the melting temperature is set at 200 °C, the mass ratio of the plasticizer is 5%, the mass ratio of the resin is 40%, the addition amount of glass beads is 350 kg / m³, the stirring time is 8 min, and the curing time is 25 s. At this time, after calculation, the flow rate Q≈85.3 g / 10 min. Since there are a total of ten groups, the final flow rate is also 10 in total, as shown in Table 1;

[0094] Table 1

[0095]

[0096] At the same time, the relevant data of the hot-melt coating sprayed on a certain road is obtained. When No. 1 is in use, the initial set thickness of the coating is 2.0 mm, the real-time thickness of the coating detected in real time is 1.9 mm, the average spraying environment temperature is 25 °C, the average environmental relative humidity is 60%, and the usage time is 10 min. At this time, after calculation, the utilization rate is 4.89 kg / ㎡. Since there are a total of ten groups, the final utilization rate is also 10 in total, as shown in Table 2;

[0097] Table 2

[0098]

[0099] These ten groups correspond one by one. At this time, based on the correspondence relationship of the flow data items, basic information, and operation information, the input attribute items are obtained, and the usage data items are used as the output attribute items. According to the set splitting ratio, that is, the flow rate, production process parameters, and relevant data of the hot-melt coating numbered 1-10 are used as the input data, and the utilization rates numbered 1-10 are used as the output data. However, among them, numbers 1-7 are used as the training set, and 8-10 are used as the test set.

[0100] Step S600: Acquire the scheduled use location of the target hot melt coating, obtain the scheduled use items, and acquire demand data based on the scheduled use items.

[0101] It should be noted that the data feedback information corresponds to the basic information, and the production line parameters are reversely adjusted according to the deep learning results. The method for obtaining demand data includes: setting a request threshold to obtain a request time item, the request threshold is a fixed time value, and the request threshold is the previous 24 hours; based on the request threshold, the temperature information and humidity information of the predetermined usage items are obtained through the target device, the target device is a temperature sensor and a humidity sensor, and the demand data is obtained.

[0102] Step S700: input the demand data as input data into the usage rate model to obtain data feedback information corresponding to the demand data.

[0103] It should be noted that after obtaining the temperature information and humidity information, the production line parameters are reversely adjusted according to the deep learning results. Due to the correspondence between the flow data items, basic information and operation information, the input attribute items are obtained, and the data items are used as the output attribute items. The flow data items are the results calculated using the basic information. The basic information includes resin content, melting temperature, stirring time, total amount of glass beads added, plasticizer ratio and curing time. The operation information includes average temperature, average humidity, coating thickness and usage time. Therefore, after obtaining the demand data, that is, understanding the temperature information and humidity information of the scheduled usage items, the coating is now The layer thickness and usage time are both estimated items, that is, when the user sprays hot-melt coating, the user estimates the spraying thickness and the estimated usage time, and then the usage rate data can be calculated to obtain the usage data item. Because the usage data item will also change when the demand data changes, the flow data item corresponding to the usage data item can be reversely obtained according to the usage rate model, and then the resin content, melting temperature, stirring time, total glass bead addition, plasticizer ratio and curing time in the basic information are adjusted to make the flow data item match the changed usage data item, and then the intelligent adjustment effect of the production line can be completed based on the results of deep learning.

[0104] Embodiment 4

[0105] In the specific implementation process, the target hot-melt coating is scheduled to be used in a section of a highway in a certain area, and the scheduled use item is obtained. According to the threshold of the previous 24 hours, the temperature sensor and the humidity sensor are used to obtain the demand data of the scheduled use item. The average ambient temperature is 25°C, and the average relative humidity is 60%. At this time, the expected spraying thickness is 2mm. In order to keep the thickness of the hot-melt coating at 1.9mm within 10 minutes, the formula for obtaining the usage rate is:

[0106] ;

[0107] It is calculated that:

[0108] ;

[0109] That is, the usage rate is 4.89 kg / ㎡. At the same time, according to the basic information corresponding to this usage rate value in the usage rate model, the flow data items corresponding to the usage data items are obtained in reverse, and then the resin content, melting temperature, stirring time, total glass bead addition amount, plasticizer ratio, and curing time in the basic information are adjusted to make the flow data items match the changed usage data items. Furthermore, the intelligent adjustment effect of the production line can be completed based on the results of deep learning.

[0110] An intelligent control system for a hot-melt coating production line based on deep learning uses the above-mentioned intelligent control method for a hot-melt coating production line based on deep learning, including: a production marking module: obtaining the basic information of the target hot-melt coating production line to obtain calibration information items, and the calibration information items are used to represent the marking information corresponding to the production line; an information acquisition module: obtaining the basic information and operation information of the hot-melt coating to obtain a control relationship set; an information processing module: based on the basic information, setting a flow rate acquisition formula to obtain the hot-melt coating flow rate data corresponding to the basic information, obtaining flow data items, and based on the operation information, setting a usage rate acquisition formula to obtain the hot-melt coating usage rate data corresponding to the operation information, obtaining usage data items, using the comparison result of the flow data items and the usage data items as training data, capturing the corresponding relationship and corresponding method between the flow data items and the usage data items through deep learning, and generating a usage rate model of the hot-melt coating; a feedback adjustment module: obtaining the predetermined usage position of the target hot-melt coating to obtain a predetermined usage item, obtaining demand data based on the predetermined usage item, using the demand data as input data and inputting it into the usage rate model, obtaining data feedback information corresponding to the demand data, the data feedback information corresponding to the basic information, and reversely adjusting the production line parameters according to the deep learning results.

[0111] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended embodiments and their equivalents.

Claims

1. An intelligent control method for a hot-melt coating production line based on deep learning: Obtain the basic information of the target hot-melt coating production line to obtain calibration information items, where the calibration information items are used to represent the marker information corresponding to the production line; Based on the calibration information items, obtain the basic information and operation information of the hot-melt coating to obtain a control relation set, where the control relation set includes at least one set of corresponding relations between the basic information and the operation information; It is characterized by including: Based on the basic information, set a flow rate acquisition formula to obtain the hot-melt coating flow rate data corresponding to the basic information, and obtain a flow data item, where the flow data item is used to represent the attribute information of the hot-melt coating obtained under different production processes; Based on the operation information, set a usage rate acquisition formula to obtain the hot-melt coating usage rate data corresponding to the operation information, and obtain a usage data item, where the usage data item is used to represent the usage feedback information of the hot-melt coating under different operation environments; Use the comparison result of the flow data item and the usage data item as training data, and capture the corresponding relationship and corresponding method between the flow data item and the usage data item through deep learning to generate a usage rate model for the hot-melt coating; Obtain the predetermined use position of the target hot-melt coating to obtain a predetermined use item, obtain demand data based on the predetermined use item, input the demand data as input data into the usage rate model, obtain data feedback information corresponding to the demand data, where the data feedback information corresponds to the basic information, and reverse-adjust the production line parameters according to the deep learning result.

2. The intelligent control method for a hot-melt coating production line based on deep learning according to claim 1, wherein: The basic information of the hot-melt coating includes resin content, melting temperature, stirring time, total glass bead addition amount, plasticizer ratio, and curing time, and the flow rate acquisition formula is: ； Among them is the flow rate, is the equivalent flow channel radius of the coating in the molten state, is the melting temperature, is the mass ratio of the plasticizer, is the mass ratio of the resin, is the addition amount of glass beads, is the stirring time, is the curing time, is the reference viscosity, 、 、 are experimental fitting coefficients, 、 、 、 are reference values.

3. The intelligent control method for a hot-melt coating production line based on deep learning according to claim 1, characterized in that: The operation information of the hot-melt coating includes average temperature, average humidity, coating thickness, and usage duration, and the usage rate acquisition formula is: ； in is the paint consumption per unit area, i.e., the utilization rate, Set the initial thickness of the coating. is the real-time thickness of the coating, is the average temperature of the spraying environment, is the average relative humidity of the environment, For the duration of use, and is the material characteristic coefficient, is a very small constant.

4. The intelligent control method for a hot melt coating production line based on deep learning according to claim 2, characterized in that: The method for obtaining the equivalent flow channel radius of the coating in the molten state is: Create a formula for obtaining the equivalent flow channel radius of the coating in the molten state: ; wherein is the reference temperature.

5. The intelligent control method for a hot-melt coating production line based on deep learning according to claim 3, wherein: The method for obtaining the average spraying environment temperature and the average environmental relative humidity includes: Based on the usage duration, obtain a target time limit value, and based on the target time limit value, obtain environmental temperature and environmental humidity data to obtain an environmental temperature set and an environmental humidity set; Based on the requested time limit, respectively obtain the average values of the environmental temperature set and the environmental humidity set to obtain the average spraying environment temperature and the average environmental relative humidity.

6. The intelligent control method for a hot-melt coating production line based on deep learning according to claim 1, characterized in that: Initial information is stored on the hot-melt coating production line. The method for obtaining the calibration information items includes: the initial information includes the initial code of the hot-melt coating production line, and the hot-melt coating production line is distinguished based on the initial code to obtain the calibration information items.

7. The intelligent control method for a hot-melt coating production line based on deep learning according to claim 1, wherein: The method for obtaining the control relation set includes: Based on the basic information, arrange the serial numbers of the hot-melt coating to obtain a serial number distribution item; Based on the corresponding information of the serial number distribution item stored in the operation information, obtain the corresponding relationship between the basic information and the operation information, and then obtain the control relation set.

8. The intelligent control method for a hot-melt coating production line based on deep learning according to claim 1, wherein: The method for obtaining the usage rate model of the hot-melt coating includes: Based on the corresponding relationship between the flow data item, the basic information, and the operation information, obtain an input attribute item, and use the usage data item as an output attribute item; Create a data classification table, classify the input attribute items and the output attribute items, and obtain the corresponding result set; Select the target deep training model, set the split ratio, and split the corresponding result set based on the split ratio to obtain the training set and the test set; The training set is used as the training data to train the target deep training model and obtain the usage rate model; The usage model is tested and optimized using the test set.

9. The intelligent control method for a hot-melt coating production line based on deep learning according to claim 1, characterized in that: The method for obtaining the demand data includes: Set the request threshold to obtain the request time item, where the request threshold is a fixed time value; Based on the request threshold, the temperature information and humidity information of the scheduled usage item are obtained through the target device to obtain the demand data.

10. An intelligent control system for a heat-fused coating production line based on deep learning, characterized in that: A hot melt coating production line intelligent control method based on deep learning according to any one of claims 1 to 9 is used, comprising: Production marking module: obtains the basic information of the target hot-melt coating production line and obtains the calibration information item, which is used to represent the marking information corresponding to the production line; Information acquisition module: obtains basic information and operation information of hot-melt coatings and obtains a reference relationship set; Information processing module: Based on the basic information, a flow rate acquisition formula is set to obtain the hot melt coating flow rate data corresponding to the basic information, and a flow data item is obtained; based on the operation information, a usage rate acquisition formula is set to obtain the hot melt coating usage rate data corresponding to the operation information, and a usage data item is obtained; the comparison results of the flow data item and the usage data item are used as training data, and the corresponding relationship and corresponding method between the flow data item and the usage data item are captured through deep learning, so as to generate a usage rate model of the hot melt coating; Feedback adjustment module: obtain the scheduled usage location of the target hot-melt coating, get the scheduled usage items, obtain the demand data based on the scheduled usage items, input the demand data as input data into the usage rate model, obtain the data feedback information corresponding to the demand data, the data feedback information corresponds to the basic information, and reversely adjust the production line parameters according to the deep learning results.

Images