Intelligent operation regulation and control system and method for packaging bag production line

Abstract

The invention relates to the technical field of production line intelligent regulation and control, and particularly discloses an intelligent operation regulation and control system and method for a packaging bag production line. Performing time sequence correlation and condition implicit feature interaction of environmental parameter data of each transmission sub-region by adopting a data analysis and processing technology based on deep learning, then constructing a transmission sub-region spatial topological matrix of the plurality of transmission sub-regions and performing topological feature extraction; therefore, whether the environmental condition of each detection point is abnormal or not is automatically judged according to the global interactive codes between the plurality of temperature-humidity interactive fusion representations and the spatial topological features of the transmission sub-regions. In this way, the threshold value is automatically adjusted according to the historical trend, parameter changes are intelligently recognized, potential problems are warned in advance, and flexible and accurate environment monitoring is achieved.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent control of production lines, and more specifically, to an intelligent operation control system and method for a packaging bag production line. Background Art

[0002] In the packaging industry, especially in the production line of packaging bags, the degree of automation is getting higher and higher, and each link from raw material supply to finished product packaging requires a high degree of coordination and control. In the production process, environmental conditions (such as temperature and humidity) have a direct impact on product quality. For example, in the production process of plastic film packaging, changes in temperature and humidity may cause problems such as material deformation or adhesion, which in turn affects the final quality of the product. Therefore, controlling production environmental conditions has become one of the key links to ensure product quality.

[0003] Chinese patent CN113589772B proposes an intelligent operation control and management system for a packaging bag graphic printing production line. It first detects surface defects of the packaging bags, then monitors the temperature and humidity of the conveying area to ensure a suitable conveying environment, and at the same time establishes a management database to store standard information to support intelligent control. Then, the flow direction of the packaging bags is determined based on the collected information. Finally, the printing template is adjusted in the printing area according to the characteristics of the packaging bags to achieve precise printing.

[0004] In the above patent, the abnormal detection of environmental conditions at each detection point is to compare the actual environmental parameters with the standard range. If a certain environmental parameter of the detection point exceeds the standard range, the detection point is regarded as an abnormal detection. However, the method mentioned in the patent relies on a preset standard range, which limits its flexibility and adaptability to a certain extent. That is, static standards are difficult to cope with changes in environmental parameters over time and fluctuations in production conditions. Especially under variable or extreme conditions, fixed thresholds may no longer be applicable, resulting in misjudgments or missed judgments. In addition, due to the lack of time series analysis of historical data, this method cannot capture the trends and potential patterns of environmental parameters evolving over time, thereby losing the ability to predict and prevent future anomalies. For example, in the process of gradual increase in humidity, although the preset threshold has not yet been reached, this trend itself may have hinted at an impending quality problem.

[0005] Therefore, an optimized intelligent operation control system for the packaging bag production line is desired. Summary of the invention

[0006] The present application provides a packaging bag production line intelligent operation control system and method, which can automatically adjust the threshold value according to the historical trend of the current environmental parameters, rather than a fixed static standard. And it can identify the trend of these parameters changing over time. Even if the current environmental parameters have not reached the preset threshold, their development trend can also prompt potential problems of the system, thereby achieving more flexible, accurate and intelligent monitoring of production environment conditions.

[0007] In a first aspect, an intelligent operation and control system for a packaging bag production line is provided, comprising: a packaging bag defect detection module, used to perform surface defect detection on packaging bags to be printed at the inlet end of the packaging bag production line and screen out packaging bags to be printed with qualified quality; a packaging bag conveying module, used to convey the packaging bags to be printed with qualified quality to a conveying area; a conveying area division module, used to divide the conveying area into sub-areas to obtain a plurality of conveying sub-areas; a conveying area environment detection module, used to perform environmental condition analysis and abnormality identification on the plurality of conveying sub-areas to obtain identification results; wherein the conveying area environment detection module comprises: an environmental parameter acquisition encoding unit, used to obtain a time queue of environmental parameter data of a plurality of conveying sub-areas collected by an environmental parameter acquisition terminal deployed in the plurality of conveying sub-areas, wherein the environmental parameter data comprises temperature value and humidity value.

[0008] The environmental parameter fusion unit is used to perform conditional feature fusion based on time series features on the time queues of multiple environmental parameter data to obtain significant interactive fusion representations of multiple temperature-humidity time series features.

[0009] The spatial topology extraction unit is used to construct a transmission sub-region spatial topology matrix of the multiple transmission sub-regions and perform topological feature extraction to obtain the transmission sub-region spatial topology features.

[0010] The global time series interaction unit is used to perform global transmission area temperature-humidity time series feature interaction encoding on the multiple temperature-humidity time series feature significant interaction fusion representations and the transmission sub-area spatial topological features to obtain a sequence of temperature-humidity time series feature interaction representations under the global transmission area.

[0011] The recognition result generating unit is used to obtain the recognition results of each transmission sub-area based on the sequence of temperature-humidity time series characteristics interactively represented under the global transmission area.

[0012] In a second aspect, a method for intelligent operation and control of a packaging bag production line is provided, comprising: performing surface defect detection on packaging bags to be printed at the inlet end of the packaging bag production line and screening out packaging bags to be printed with qualified quality; transmitting the packaging bags to be printed with qualified quality to a conveying area; dividing the conveying area into sub-areas to obtain a plurality of conveying sub-areas; performing environmental condition analysis and abnormality identification on the plurality of conveying sub-areas to obtain identification results; wherein, performing environmental condition analysis and abnormality identification on the plurality of conveying sub-areas to obtain identification results comprises: obtaining a time queue of environmental parameter data of the plurality of conveying sub-areas collected by an environmental parameter collection terminal deployed in the plurality of conveying sub-areas, wherein the environmental parameter data The data include temperature values ​​and humidity values; conditional feature fusion based on time series features is performed on the time queues of multiple environmental parameter data respectively to obtain multiple temperature-humidity time series feature significant interactive fusion representations; a transmission sub-region spatial topology matrix of the multiple transmission sub-regions is constructed and topological feature extraction is performed to obtain the transmission sub-region spatial topological features; global transmission region temperature-humidity time series feature interactive encoding is performed on the multiple temperature-humidity time series feature significant interactive fusion representations and the transmission sub-region spatial topological features to obtain a sequence of temperature-humidity time series feature interactive representations under the global transmission region; based on the sequence of temperature-humidity time series feature interactive representations under the global transmission region, the recognition results of each transmission sub-region are obtained.

[0013] The present application provides an intelligent operation control system and method for a packaging bag production line, which uses a time queue of environmental parameter data (temperature and humidity values) of multiple transmission sub-areas collected by environmental parameter collection terminals deployed at the multiple environmental detection points, and uses data analysis and processing technology based on deep learning to perform temporal association and conditional implicit feature interaction of environmental parameter data of each transmission sub-area, then constructs a transmission sub-area spatial topological matrix of the multiple transmission sub-areas and extracts topological features, so as to automatically judge whether there is an abnormality in the environmental conditions of each detection point according to the global interaction encoding between multiple temperature-humidity interactive fusion representations and transmission sub-area spatial topological features. In this way, the threshold can be automatically adjusted according to the historical trend of the current environmental parameters, rather than a fixed static standard. And the trend of these parameters changing over time can be identified. Even if the current environmental parameters have not reached the preset threshold, their development trend can also prompt potential problems in the system, thereby achieving more flexible, accurate and intelligent monitoring of production environmental conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings of the embodiments of the present application are briefly introduced below. Obviously, the drawings described below only relate to some embodiments of the present application, and are not intended to limit the present application.

[0015] Figure 1 This is a schematic block diagram of the intelligent operation and control system of the packaging bag production line according to an embodiment of the present application.

[0016] Figure 2 This is a schematic block diagram of a conveying area environment detection module in the intelligent operation and control system of a packaging bag production line according to an embodiment of the present application.

[0017] Figure 3 This is a schematic diagram of the data flow of the conveying area environment detection module in the intelligent operation and control system of the packaging bag production line in an embodiment of the present application.

[0018] Figure 4 This is a schematic block diagram of an environmental parameter fusion unit in the intelligent operation and control system of a packaging bag production line according to an embodiment of the present application.

[0019] Figure 5 This is a schematic flow chart of the intelligent operation control method of the packaging bag production line according to an embodiment of the present application. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present application.

[0021] Traditional environmental condition anomaly detection methods usually judge whether the detection point is abnormal based on a fixed standard range, that is, when the actual environmental parameters of a detection point exceed the preset normal range, the detection point will be marked as abnormal. Although this method is simple and direct, it has obvious limitations. First, it relies too much on pre-set static thresholds, which makes it difficult for the system to adapt to the changes of environmental parameters over time and production conditions. Especially in the case of variable or extreme environmental conditions, the fixed threshold standards may no longer apply, which is easy to produce wrong judgments or miss abnormal conditions. Secondly, such methods often ignore the in-depth analysis of historical data, especially the lack of grasp of the trend of environmental parameters over time, which makes the system unable to effectively predict future abnormal conditions. For example, even if the humidity level has not yet reached the set warning line, if it shows a continuous upward trend, this may indicate an impending quality problem, and traditional methods cannot warn of this. Therefore, traditional anomaly detection methods are insufficient in flexibility, adaptability and foresight, and urgently need to be improved to improve the efficiency of environmental monitoring in the production process.

[0022] Based on this, this application proposes an intelligent operation and control system for a packaging bag production line. Figure 1 Schematic diagram of the intelligent operation control system of the packaging bag production line of the embodiment of the present application. Figure 1 As shown, the packaging bag production line intelligent operation control system 100 includes: a packaging bag defect detection module 110, which is used to perform surface defect detection on the packaging bags to be printed at the entrance end of the packaging bag production line and screen out the packaging bags to be printed with qualified quality; a packaging bag conveying module 120, which is used to convey the packaging bags to be printed with qualified quality to the conveying area; a conveying area division module 130, which is used to divide the conveying area into sub-areas to obtain multiple conveying sub-areas; a conveying area environment detection module 140, which is used to perform environmental condition analysis and abnormality recognition on the multiple conveying sub-areas to obtain recognition results.

[0023] In the above-mentioned intelligent operation and control system of the packaging bag production line, the packaging bag defect detection module 110 is used to perform surface defect detection on the packaging bags to be printed at the entrance end of the packaging bag production line and screen out the packaging bags to be printed with qualified quality. It should be understood that by performing surface defect detection on the packaging bags, defective packaging bags can be discovered and removed in time to prevent these unqualified products from entering the subsequent production process, thereby ensuring the quality of the final product. If unqualified packaging bags enter the production line without screening, unnecessary waste of resources, such as printing ink, electricity, etc., may be caused in the subsequent processing process. These wastes can be effectively reduced through early detection and screening. If unqualified products are discovered in the later stage of the production process, not only do they need to be re-produced, but they may also cause the production line to stop, affecting production efficiency. Early detection can avoid these problems and make the production process smoother.

[0024] Optionally, in one embodiment of the present application, a visual inspection system can be used to scan the surface of the packaging bag using a high-definition camera and image processing technology. These systems can quickly and accurately identify defects such as scratches, stains, color deviations, etc. At the same time, a robotic arm or automated transmission system can be combined to immediately remove unqualified packaging bags from the production line once they are detected, ensuring that only packaging bags that meet the standards can continue to enter the subsequent process.

[0025] In the above-mentioned intelligent operation and control system of the packaging bag production line, the packaging bag conveying module 120 is used to convey the qualified packaging bags to be printed to the conveying area. It should be understood that the conveying area is an important link connecting different production processes. Transmitting qualified packaging bags to the conveying area can ensure the continuity and efficiency of the production process. If qualified packaging bags cannot be transmitted to the conveying area in time, the production line may be interrupted, affecting production efficiency. Automated transmission can reduce human intervention, reduce the risk of errors and damage that may occur during manual handling, and improve the safety and reliability of production. Through automated transmission, production resources can be better managed and scheduled to ensure that each process can operate efficiently and avoid waste of resources.

[0026] In the above-mentioned intelligent operation and control system of the packaging bag production line, the conveying area division module 130 is used to divide the conveying area into sub-areas to obtain multiple conveying sub-areas. It should be understood that by dividing the conveying area into multiple sub-areas, refined management of each sub-area can be achieved. Each sub-area can be monitored and adjusted independently, thereby improving the control accuracy and efficiency of the overall production process. The environmental conditions (such as temperature and humidity) of different sub-areas may be different. Through sub-area division, the environmental parameters of each sub-area can be more accurately monitored and analyzed, abnormal situations can be discovered and handled in time, and the stability of the production environment can be ensured. If a sub-area fails or is abnormal, the problem can be quickly located through sub-area division to reduce the impact on the entire production line. In this way, measures can be taken more quickly to resume production. Sub-area division helps to reasonably allocate production resources, such as energy, manpower and equipment. Each sub-area can optimize resource allocation according to its own needs to improve resource utilization.

[0027] In the above-mentioned intelligent operation and control system of the packaging bag production line, the transmission area environment detection module 140 is used to analyze the environmental conditions and identify abnormalities of the multiple transmission sub-areas to obtain identification results. It should be understood that by analyzing the environmental parameters such as temperature and humidity of each sub-area, the stability of the production environment can be monitored in real time. A stable production environment is a prerequisite for ensuring product quality and production efficiency. By timely discovering and identifying abnormal conditions in environmental conditions, measures can be taken quickly to make adjustments to avoid negative impacts of environmental changes on the production process.

[0028] Correspondingly, in the transmission area environment detection module, the technical concept of the present application is to collect the environmental parameter data (temperature and humidity values) of multiple transmission sub-areas through the time queue of the environmental parameter collection terminal deployed at the multiple environmental detection points, and use the data analysis and processing technology based on deep learning to perform the temporal association and conditional implicit feature interaction of the environmental parameter data of each transmission sub-area, and then construct the transmission sub-area spatial topological matrix of the multiple transmission sub-areas and extract the topological features, so as to automatically judge whether there is an abnormality in the environmental conditions of each detection point according to the global interaction encoding between the multiple temperature-humidity interaction fusion representations and the transmission sub-area spatial topological features. In this way, the threshold can be automatically adjusted according to the historical trend of the current environmental parameters, rather than a fixed static standard. And the trend of these parameters changing over time can be identified. Even if the current environmental parameters have not reached the preset threshold, their development trend can also prompt the potential problems of the system, thereby realizing more flexible, accurate and intelligent monitoring of production environment conditions.

[0029] Figure 2 This is a schematic block diagram of a conveying area environment detection module in the intelligent operation and control system of a packaging bag production line according to an embodiment of the present application. Figure 3This is a schematic diagram of the data flow of the transmission zone environment detection module in the intelligent operation and control system of the packaging bag production line of the embodiment of the present application. Figure 2 As shown, the transmission area environment detection module 140 includes: an environmental parameter acquisition encoding unit 141, which is used to obtain a time queue of environmental parameter data of multiple transmission sub-areas collected by environmental parameter acquisition terminals deployed in the multiple transmission sub-areas, wherein the environmental parameter data includes temperature values ​​and humidity values; an environmental parameter fusion unit 142, which is used to perform conditional feature fusion based on time series features on the time queues of multiple environmental parameter data to obtain multiple temperature-humidity time series feature significant interactive fusion representations; a spatial topology extraction unit 143, which is used to construct a transmission sub-area spatial topology matrix of the multiple transmission sub-areas and perform topological feature extraction to obtain transmission sub-area spatial topology features; a global time series interaction unit 144, which is used to perform global transmission area temperature-humidity time series feature interactive encoding on the multiple temperature-humidity time series feature significant interactive fusion representations and the transmission sub-area spatial topology features to obtain a sequence of temperature-humidity time series feature interactive representations under the global transmission area; and a recognition result generation unit 145, which is used to obtain the recognition results of each transmission sub-area based on the sequence of temperature-humidity time series feature interactive representations under the global transmission area.

[0030] In the above-mentioned intelligent operation and control system of the packaging bag production line, the environmental parameter acquisition encoding unit 141 is used to obtain the time queue of environmental parameter data of multiple transmission sub-areas collected by the environmental parameter acquisition terminals deployed in the multiple transmission sub-areas, wherein the environmental parameter data include temperature values ​​and humidity values. It should be understood that by collecting environmental parameters such as temperature and humidity in real time, the environmental conditions of each transmission sub-area can be timely understood to ensure the stability and consistency of the production environment. Real-time monitoring can quickly discover abnormal changes in environmental parameters. The time queue records the changing trends of environmental parameters over time. By analyzing these historical data, the long-term changing laws of environmental parameters can be captured. Based on the trend analysis of historical data, future environmental changes can be predicted, potential problems can be warned in advance, and preventive measures can be taken. At the same time, according to the historical trends of environmental parameters, the system can automatically adjust the threshold value to achieve adaptive control and improve the flexibility and adaptability of the production process.

[0031] In the above-mentioned intelligent operation and control system of the packaging bag production line, the environmental parameter fusion unit 142 is used to perform conditional feature fusion based on time series features on the time queues of multiple environmental parameter data to obtain multiple temperature-humidity time series feature significant interactive fusion representations. It should be understood that temperature and humidity are two important environmental parameters, and there is a complex interaction between them. Analyzing each parameter separately may not fully reflect the changes in environmental conditions, but through feature fusion, the interaction effects between these parameters can be captured. The changes of environmental parameters over time have time series characteristics, and the laws and trends of these changes can be more accurately captured through the fusion of time series features. The significant interaction features between temperature and humidity can provide more information to help the system identify abnormalities in environmental conditions more accurately. The abnormality of a single parameter may not be obvious, but the interaction effects of multiple parameters may reveal potential problems. Through feature fusion, the robustness of the system can be enhanced and misjudgments caused by accidental fluctuations in a single parameter can be reduced.

[0032] Optionally, in one embodiment of the present application, Figure 4 Schematic block diagram of the environmental parameter fusion unit in the intelligent operation control system of the packaging bag production line of the embodiment of the present application. Figure 4 As shown, the environmental parameter fusion unit 142 includes: a temperature and humidity time series encoding subunit 1421, which is used to arrange the time queue of the environmental parameter data into a temperature time series input vector and a humidity time series input vector according to the time dimension, and then pass the temperature time series input vector and the humidity time series input vector through a sequence encoder based on a 1D-CNN model to obtain a temperature time series implicit correlation feature vector and a humidity time series implicit correlation feature vector; a temperature and humidity conditional interaction fusion subunit 1422, which is used to perform conditional implicit feature significant interaction fusion on the temperature time series implicit correlation feature vector and the humidity time series implicit correlation feature vector to obtain the temperature-humidity time series feature significant interaction fusion representation.

[0033] In the above-mentioned intelligent operation and control system of the packaging bag production line, the temperature and humidity time series encoding subunit 1421 is used to arrange the time queue of the environmental parameter data into a temperature time series input vector and a humidity time series input vector according to the time dimension, and then pass the temperature time series input vector and the humidity time series input vector through a sequence encoder based on the 1D-CNN model to obtain a temperature time series implicit correlation feature vector and a humidity time series implicit correlation feature vector. It should be understood that considering that the temperature and humidity in the time queue of each environmental parameter data are constantly changing with the change of time, therefore, the time queue of the environmental parameter data is arranged into a temperature time series input vector and a humidity time series input vector according to the time dimension to provide a basis for subsequent time series feature capture. Then, considering that the time series characteristics of temperature and humidity are different in different local time intervals. Based on this, in order to better capture and extract the local time series feature information of temperature and humidity, in the technical solution of the present application, the temperature time series input vector and the humidity time series input vector are passed through a sequence encoder based on a 1D-CNN model with a medical one-dimensional convolution sliding window (i.e., convolution kernel) to capture local features and obtain the temperature time series implicit correlation feature vector and the humidity time series implicit correlation feature vector.

[0034] In the above-mentioned intelligent operation and control system of the packaging bag production line, the temperature and humidity conditional interaction fusion subunit 1422 is used to perform conditional implicit feature significant interaction fusion on the temperature time series implicit association feature vector and the humidity time series implicit association feature vector to obtain the temperature-humidity time series feature significant interaction fusion representation. It should be understood that considering that the temperature time series implicit association feature vector and the humidity time series implicit association feature vector each carry local time series association information of different physical contents, and there is a mutual dependence between temperature and humidity, a simple independent analysis may ignore some important interactions. A mechanism is needed to integrate these two features so that they can express the mutual influence between each other at a higher level, so that the model can better focus on the key significant time series interaction part. In the technical solution of the present application, the temperature time series implicit association feature vector and the humidity time series implicit association feature vector are subjected to conditional implicit feature significant interaction fusion to obtain the temperature-humidity time series feature significant interaction fusion representation vector as the temperature-humidity time series feature significant interaction fusion representation.

[0035] Optionally, in one embodiment of the present application, the temperature and humidity condition interactive fusion subunit is used to include: a temperature-humidity joint implicit capture secondary subunit, which is used to input the temperature time series implicit association feature vector and the humidity time series implicit association feature vector into a joint implicit feature capture network to obtain a temperature-humidity joint implicit feature vector; a temperature-humidity joint implicit feature activation secondary subunit, which is used to perform feature activation based on a Sigmoid function on the temperature-humidity joint implicit feature vector to obtain a temperature-humidity conditional feature vector; a temperature-humidity time series feature interactive secondary subunit, which is used to perform conditional dependent interactive fusion on the temperature time series implicit association feature vector and the humidity time series implicit association feature vector based on the temperature-humidity conditional feature vector to obtain a temperature-humidity time series feature significant interactive fusion representation vector as the temperature-humidity time series feature significant interactive fusion representation.

[0036] Optionally, in one embodiment of the present application, the temperature-humidity joint implicit capture secondary sub-unit is used to: after adding the temperature time series implicit correlation feature vector and the humidity time series implicit correlation feature vector by position, multiply the obtained temperature-humidity time series sum vector with the weight matrix and then add it by position with the bias vector to obtain the temperature-humidity time series joint interaction vector; use the tanh function to process the temperature-humidity time series joint interaction vector to obtain the temperature-humidity joint implicit feature vector.

[0037] Optionally, in one embodiment of the present application, the temperature-humidity time series feature interaction secondary subunit includes: a temperature semantic contribution calculation tertiary subunit, which is used to calculate the temperature semantic contribution of the temperature time series implicit association feature vector relative to the temperature-humidity condition feature vector; a humidity semantic contribution calculation tertiary subunit, which is used to calculate the humidity semantic contribution of the humidity time series implicit association feature vector relative to the temperature-humidity condition feature vector; a semantic contribution normalization modulation tertiary subunit, which is used to normalize the temperature semantic contribution and the humidity semantic contribution, and use the normalized temperature semantic contribution and the normalized humidity semantic contribution to modulate the temperature time series implicit association feature vector. The quantity and the humidity time series implicit association feature vector are weighted modulated to obtain the modulated temperature time series implicit association feature vector and the modulated humidity time series implicit association feature vector; the temperature and humidity time series feature guided interaction three-level sub-unit is used to use the modulated temperature time series implicit association feature vector as the query vector, the modulated humidity time series implicit association feature vector as the key vector and the temperature-humidity condition feature vector as the value vector, and input the modulated temperature time series implicit association feature vector, the modulated humidity time series implicit association feature vector and the temperature-humidity condition feature vector into the feature significant guided interaction module based on the converter structure to obtain the temperature-humidity time series feature significant interaction fusion representation vector.

[0038] Optionally, in one embodiment of the present application, the semantic contribution normalization modulation three-level sub-unit is used to: calculate the sum of the temperature semantic contribution and the humidity semantic contribution to obtain the temperature-humidity semantic contribution sum; respectively divide the temperature semantic contribution and the humidity semantic contribution by the temperature-humidity semantic contribution sum to obtain the normalized temperature semantic contribution and the normalized humidity semantic contribution; perform position point multiplication on the temperature time series implicit association feature vector and the normalized temperature semantic contribution to obtain the modulated temperature time series implicit association feature vector; perform position point multiplication on the humidity time series implicit association feature vector and the normalized humidity semantic contribution to obtain the modulated humidity time series implicit association feature vector.

[0039] Optionally, in one embodiment of the present application, the temperature semantic contribution calculation three-level sub-unit is used to: calculate the positional division of the corresponding positions of the temperature time series implicit association feature vector and the temperature-humidity condition feature vector to obtain a temperature semantic contribution vector; calculate the logarithmic function value with base two of the absolute value of each eigenvalue of the temperature semantic contribution vector to obtain a temperature semantic contribution logarithmic vector; calculate the positional dot product of the temperature semantic contribution logarithmic vector and the temperature semantic contribution logarithmic vector, and perform positional multiplication on the obtained dot product vector to obtain the temperature semantic contribution value; calculate an exponential function with the natural constant e as the base, with the temperature semantic contribution value as the exponent, to obtain the temperature semantic contribution.

[0040] In summary, in the embodiment of the present application, the temperature time series implicit correlation feature vector and the humidity time series implicit correlation feature vector are subjected to conditional implicit feature significant interactive fusion, which can be expressed by the formula: ;in, and are the temperature time series implicit correlation feature vector and the humidity time series implicit correlation feature vector, respectively, To add by location, and are the weight matrix and bias vector respectively, for function, is the temperature-humidity joint implicit feature vector, for function, represents the temperature-humidity condition feature vector, is the eigenvalue of each position in the temperature time series implicitly associated eigenvector, is the eigenvalue of each position in the implicitly associated eigenvector of the humidity time series, is the eigenvalue of each position in the temperature-humidity condition eigenvector, is the number of eigenvalues ​​in the temperature time series implicit correlation eigenvector and the humidity time series implicit correlation eigenvector, represents the logarithmic function value with base 2, represents an exponential function with the natural constant e as the base, is the temperature semantic contribution, is the humidity semantic contribution, and are the normalized temperature semantic contribution and the normalized humidity semantic contribution, respectively. and They are the implicit correlation feature vector of the temperature time series after modulation and the implicit correlation feature vector of the humidity time series after modulation, yes The transposed vector of yes Length, represents vector multiplication, yes function, It is the representation vector of the significant interaction fusion of the temperature-humidity time series features.

[0041] Specifically, first, the shared implicit structure and original features between the temperature time series implicit association feature vector and the humidity time series implicit association feature vector are captured and extracted through a joint implicit feature capture network to obtain a temperature-humidity joint implicit feature vector, and then the Sigmoid function is used to perform feature activation on the temperature-humidity joint implicit feature vector to extract key and important feature information to obtain a temperature-humidity conditional feature vector. Next, the semantic contribution between the temperature time series implicit association feature vector and the humidity time series implicit association feature vector relative to the temperature-humidity conditional feature vector is calculated to determine the temporal similarity between different features and conditional features. In other words, this evaluation helps to determine the contribution of each feature to the final result, thereby obtaining the temperature semantic contribution and humidity semantic contribution. Then, in order to make all its weights on the same scale, which is helpful for subsequent weighted modulation processing, the temperature semantic contribution and humidity semantic contribution are normalized, and the normalized temperature semantic contribution and humidity semantic contribution are used to modulate the temperature time series implicit association feature vector and the humidity time series implicit association feature vector to highlight and strengthen the important time series information in the feature, and obtain the modulated temperature time series implicit association feature vector and the modulated humidity time series implicit association feature vector. Finally, the modulated temperature time series implicit association feature vector is used as the query vector, the modulated humidity time series implicit association feature vector is used as the key vector, and the temperature-humidity condition feature vector is used as the value vector, which is input into the feature-to-feature significant guided interaction module based on the converter structure to promote the interaction and integration between features, so that the model can more effectively understand the contextual temporal relationship between features, and can automatically focus on key features, and finally form a temperature-humidity time series feature significant interaction fusion representation vector. In particular, in a specific implementation of the present application, the feature-to-feature significant guided interaction module based on the converter structure processes and analyzes these three vectors by using the transformer mechanism.

[0042] In the above-mentioned intelligent operation and control system of the packaging bag production line, the spatial topology extraction unit 143 is used to construct the transmission sub-region spatial topology matrix of the multiple transmission sub-regions and perform topological feature extraction to obtain the transmission sub-region spatial topology features. It should be understood that in order to more comprehensively understand and describe the spatial relationship between the transmission sub-regions and extract features useful for environmental monitoring and anomaly detection, in the technical solution of the present application, the transmission sub-region spatial topology matrix of the multiple transmission sub-regions is constructed, and the transmission sub-region spatial topology matrix is ​​passed through a topological feature extractor based on a hole convolutional neural network to obtain a transmission sub-region spatial topology feature matrix. It should be understood that the hole convolutional neural network increases the receptive field by inserting zero padding inside the convolution kernel, thereby capturing greater contextual information without increasing the number of parameters. In the present application, the use of a hole convolutional neural network to process the spatial topology matrix can effectively capture spatial information of different scales, which is very important for understanding the propagation and influence of environmental conditions between different sub-regions.

[0043] Optionally, in one embodiment of the present application, the transmission sub-area spatial topology matrix of the multiple transmission sub-areas can be constructed in the following manner. First, the transmission area is divided into multiple sub-areas according to the layout and production requirements of the production line. Each sub-area can be an independent physical area or a logical division. A unique number or identifier is assigned to each sub-area for subsequent management and analysis. The physical distance between each sub-area is measured, which can be a straight-line distance or a path distance. The spatial topology matrix is ​​a two-dimensional matrix, and the rows and columns correspond to each sub-area, respectively. The elements in the matrix represent the relationship between the sub-areas. More specifically, in this embodiment, the elements in the matrix can be the physical distance between the corresponding two sub-areas. In other embodiments of the present application, the elements in the matrix can be the adjacency relationship between the corresponding two sub-areas, for example, sub-area A is adjacent to sub-area B, sub-area B is adjacent to sub-area C, etc.

[0044] Optionally, in one embodiment of the present application, the spatial topology extraction unit is used to: construct a transmission sub-region spatial topology matrix of the multiple transmission sub-regions, and pass the transmission sub-region spatial topology matrix through a topological feature extractor based on a void convolutional neural network to obtain a transmission sub-region spatial topology feature matrix as the transmission sub-region spatial topology feature.

[0045] In the above-mentioned intelligent operation and control system of the packaging bag production line, the global time series interaction unit 144 is used to perform global transmission area temperature-humidity time series feature interaction encoding on the multiple temperature-humidity time series feature significant interaction fusion representations and the transmission sub-area spatial topological features to obtain a sequence of temperature-humidity time series feature interaction representations under the global transmission area. It should be understood that considering that the multiple temperature-humidity time series feature significant interaction fusion representation vectors express the significant interaction time series information between the temperature and humidity environments of multiple detection points. The transmission sub-area spatial topological feature matrix describes the spatial topological structure relationship between multiple sub-areas. Based on this, in order to integrate the environmental characteristics of each sub-area and understand the interaction between these features in a global scope, so as to better understand the global distribution and interaction of temperature and humidity characteristics in the entire transmission area. This helps to identify environmental patterns and abnormal conditions in the entire area. In the technical solution of the present application, the multiple temperature-humidity time series feature significant interaction fusion representation vectors and the transmission sub-area spatial topological feature matrix are passed through the global transmission area temperature-humidity time series feature interaction encoder based on the graph convolutional neural network model to obtain a sequence of temperature-humidity time series feature interaction representation vectors under the global transmission area.

[0046] Optionally, in one embodiment of the present application, the global timing interaction unit is used to: pass the multiple temperature-humidity timing feature significant interaction fusion representation vectors and the transmission sub-area spatial topological feature matrix through a global transmission area temperature-humidity timing feature interaction encoder based on a graph convolutional neural network model to obtain a sequence of temperature-humidity timing feature interaction representation vectors under the global transmission area as a sequence of temperature-humidity timing feature interaction representations under the global transmission area.

[0047] In the above-mentioned intelligent operation and control system of the packaging bag production line, the recognition result generation unit 145 is used to obtain the recognition results of each transmission sub-area based on the sequence of the temperature-humidity time series feature interaction representation under the global transmission area. Optionally, in one embodiment of the present application, the recognition result generation unit is used to: extract the temperature-humidity time series feature interaction representation vector under the global transmission area corresponding to the detected transmission sub-area from the sequence of the temperature-humidity time series feature interaction representation vector under the global transmission area, and pass the temperature-humidity time series feature interaction representation vector under the global transmission area through the classifier-based environmental anomaly identifier to obtain the recognition result, and the recognition result is used to indicate whether the environmental conditions of the detected transmission sub-area are abnormal. That is, the sequence of the temperature-humidity time series feature interaction representation under the global transmission area obtained by global interaction using multiple temperature-humidity time series feature significant interaction fusion representation vectors and the transmission sub-area spatial topological feature matrix is ​​classified and processed, so as to automatically determine whether the environmental conditions of each detection point are abnormal. In this way, the threshold can be automatically adjusted according to the historical trend of the current environmental parameters, rather than a fixed static standard. It can also identify the trends of these parameters changing over time. Even if the current environmental parameters have not yet reached the preset threshold, their development trends can indicate potential problems in the system, thereby achieving more flexible, accurate and intelligent monitoring of environmental conditions.

[0048] Preferably, when the temperature-humidity time series feature significant interaction fusion representation vector represents the joint implicit condition-guided interaction feature of the temperature-humidity time series feature distribution, when it is topologically associated based on the sub-region spatial topology through a graph convolutional neural network model together with the transmission sub-region spatial topology feature matrix, the obtained temperature-humidity time series feature interaction representation vector under the global transmission area will also have semantic representation diversity due to the difference in feature expression in the time-space dimension, thereby affecting the classification mapping regression offset and reducing the accuracy of the recognition results obtained by the classifier-based environmental anomaly identifier.

[0049] Based on this, when the temperature-humidity time series feature interaction representation vector under the global transmission area is subjected to classification probability regression through a classifier-based environmental anomaly identifier, the temperature-humidity time series feature interaction representation vector under the global transmission area is first optimized. The optimization process specifically includes: determining the mean and variance of the feature set composed of the feature values ​​of the temperature-humidity time series feature interaction representation vector under the global transmission area, and dividing the mean by the variance to obtain a probability weight ,in, and denote the mean and variance respectively, Represents the probability weight.

[0050] The first intermediate eigenvector is obtained by multiplying the temperature-humidity time series feature interaction representation vector under the global transmission area by the inverse of its maximum eigenvalue. ,in, represents the temperature-humidity time series feature interaction representation vector in the global transmission area, represents the maximum eigenvalue in the temperature-humidity time series feature interaction representation vector under the global transmission area, It means point multiplication by position. represents the first intermediate eigenvector.

[0051] The first intermediate feature vector is added with the probability weight and then the logarithm with base 2 is calculated to obtain the second intermediate feature vector ;in, represents the first intermediate eigenvector, represents the probability weight, Indicates adding by position point. represents the second intermediate eigenvector.

[0052] After subtracting the first intermediate eigenvector from the probability weight point, the inverse of each eigenvalue is calculated to obtain a third intermediate eigenvector ,in, represents the first intermediate eigenvector, represents the probability weight, It means to subtract by position point. represents the third intermediate eigenvector.

[0053] Finally, the second intermediate feature vector is added to the third intermediate feature vector to obtain the optimized temperature-humidity time series feature interaction representation vector under the global transmission area: ;in, represents the second intermediate eigenvector, Indicates adding by position point. represents the third intermediate eigenvector, Represents the temperature-humidity time series feature interaction representation vector under the optimized global transmission area.

[0054] That is, considering that the weight matrix during classification probability regression acts on the temperature-humidity time series feature interaction representation vector under the global transmission area, the classification regression offset of the temperature-humidity time series feature interaction representation vector under the global transmission area causes the weight stimulus parameter uncertainty of the weight matrix, so that the posterior probability density inference of the temperature-humidity time series feature interaction representation vector under the global transmission area through the weight matrix is ​​missing, which will affect the accuracy of the probability regression result. The above optimization process simulates the mesoscale interaction structure under probability constraints between the eigenvalue scale and the eigenvector scale of the temperature-humidity time series feature interaction representation vector under the global transmission area through the probability statistical characteristics of the temperature-humidity time series feature interaction representation vector under the global transmission area, so as to construct a bidirectional hidden variable motif based on the mesoscale short sequence relative to the probability statistical value to perform mesoscale bidirectional migration, and perform posterior recovery based on the sequence constraint on the interaction information, so as to improve the convergence effect of the temperature-humidity time series feature interaction representation vector under the global transmission area in the probability density domain, thereby improving the accuracy of the recognition result obtained by the temperature-humidity time series feature interaction representation vector under the global transmission area through the classifier-based environmental anomaly identifier. In this way, the thresholds can be automatically adjusted based on the historical trends of the current environmental parameters, rather than fixed static standards. And the trend of these parameters changing over time can be identified. Even if the current environmental parameters have not yet reached the preset thresholds, their development trends can also prompt potential problems in the system, thus achieving more flexible, accurate and intelligent monitoring of environmental conditions.

[0055] In summary, the intelligent operation and control system of the packaging bag production line according to the embodiment of the present application is explained, which uses the time queue of the environmental parameter data (temperature value and humidity value) of multiple transmission sub-areas collected by the environmental parameter collection terminal deployed at the multiple environmental detection points, and adopts the data analysis and processing technology based on deep learning to perform the temporal association and conditional implicit feature interaction of the environmental parameter data of each transmission sub-area, and then constructs the transmission sub-area spatial topological matrix of the multiple transmission sub-areas and extracts the topological features, so as to automatically judge whether there is an abnormality in the environmental conditions of each detection point according to the global interaction encoding between the multiple temperature-humidity interaction fusion representations and the transmission sub-area spatial topological features. In this way, the threshold can be automatically adjusted according to the historical trend of the current environmental parameters, rather than a fixed static standard. And the trend of these parameters changing over time can be identified. Even if the current environmental parameters have not reached the preset threshold, their development trend can also prompt the potential problems of the system, thereby realizing more flexible, accurate and intelligent monitoring of production environment conditions.

[0056] Figure 5 Schematic flow chart of the intelligent operation control method of the packaging bag production line of the embodiment of the present application. Figure 5As shown, the intelligent operation and control method of the packaging bag production line includes: S1, performing surface defect detection on the packaging bags to be printed at the entrance end of the packaging bag production line and screening out the packaging bags to be printed with qualified quality; S2, transmitting the packaging bags to be printed with qualified quality to the conveying area; S3, dividing the conveying area into sub-areas to obtain a plurality of conveying sub-areas; S4, performing environmental condition analysis and abnormality recognition on the plurality of conveying sub-areas to obtain recognition results.

[0057] Optionally, in one embodiment of the present application, environmental condition analysis and anomaly identification are performed on the multiple transmission sub-areas to obtain identification results, including: obtaining a time queue of environmental parameter data of multiple transmission sub-areas collected by an environmental parameter collection terminal deployed in the multiple transmission sub-areas, wherein the environmental parameter data includes temperature values ​​and humidity values; performing conditional feature fusion based on time series features on the time queues of the multiple environmental parameter data to obtain multiple temperature-humidity time series feature significant interactive fusion representations; constructing a transmission sub-area spatial topology matrix of the multiple transmission sub-areas and performing topological feature extraction to obtain transmission sub-area spatial topological features; performing global transmission area temperature-humidity time series feature interactive encoding on the multiple temperature-humidity time series feature significant interactive fusion representations and the transmission sub-area spatial topological features to obtain a sequence of temperature-humidity time series feature interactive representations under the global transmission area; based on the sequence of temperature-humidity time series feature interactive representations under the global transmission area, obtaining identification results of each transmission sub-area.

[0058] Here, those skilled in the art can understand that the specific operations of each step in the above-mentioned packaging bag production line intelligent operation control method have been referred to above. Figures 1 to 4 The intelligent operation and control system of the packaging bag production line has been introduced in detail, and therefore, its repeated description will be omitted.

[0059] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

[0060] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the module division is only a logical function division, and there may be other division methods in actual implementation. The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0061] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0062] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference to a figure in a claim should not be considered as limiting the claim to which it relates.

[0063] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units stated in the device claim can also be implemented by one unit through software or hardware.

[0064] Finally, it should be noted that the above description has been given for the purpose of illustration and description. In addition, the above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. An intelligent operation control system for a packaging bag production line, characterized in that: include: The packaging bag defect detection module is used to detect surface defects of the packaging bags to be printed at the inlet of the packaging bag production line and select the packaging bags to be printed with qualified quality; A packaging bag conveying module, used for conveying the packaging bags to be printed of qualified quality to a conveying area; A transmission area division module, used for dividing the transmission area into sub-areas to obtain a plurality of transmission sub-areas; A transmission area environment detection module is used to perform environmental condition analysis and abnormality identification on the multiple transmission sub-areas to obtain identification results; wherein the transmission area environment detection module includes: an environmental parameter acquisition encoding unit, used to obtain a time queue of environmental parameter data of multiple transmission sub-areas collected by environmental parameter acquisition terminals deployed in the multiple transmission sub-areas, wherein the environmental parameter data includes temperature values ​​and humidity values; an environmental parameter fusion unit, used to perform conditional feature fusion based on time series features on the time queues of multiple environmental parameter data to obtain multiple temperature-humidity time series feature significant interactive fusion representations; a spatial topology extraction unit, used to construct a transmission sub-area spatial topology matrix of the multiple transmission sub-areas and perform topological feature extraction to obtain transmission sub-area spatial topology features; a global time series interaction unit, used to perform global transmission area temperature-humidity time series feature interactive encoding on the multiple temperature-humidity time series feature significant interactive fusion representations and the transmission sub-area spatial topology features to obtain a sequence of temperature-humidity time series feature interactive representations under the global transmission area; and a recognition result generation unit, used to obtain the recognition results of each transmission sub-area based on the sequence of temperature-humidity time series feature interactive representations under the global transmission area.

2. The intelligent operation and control system for packaging bag production line according to claim 1 is characterized in that: The environmental parameter fusion unit includes: a temperature and humidity time series encoding subunit, which is used to arrange the time queue of the environmental parameter data into a temperature time series input vector and a humidity time series input vector according to the time dimension, and then pass the temperature time series input vector and the humidity time series input vector through a sequence encoder based on a 1D-CNN model to obtain a temperature time series implicit correlation feature vector and a humidity time series implicit correlation feature vector; a temperature and humidity conditional interaction fusion subunit, which is used to perform conditional implicit feature significant interaction fusion on the temperature time series implicit correlation feature vector and the humidity time series implicit correlation feature vector to obtain the temperature-humidity time series feature significant interaction fusion representation.

3. The intelligent operation control system of the packaging bag production line according to claim 2 is characterized in that: The temperature-humidity condition interactive fusion subunit is used to include: a temperature-humidity joint implicit capture secondary subunit, which is used to input the temperature time series implicit association feature vector and the humidity time series implicit association feature vector into a joint implicit feature capture network to obtain a temperature-humidity joint implicit feature vector; a temperature-humidity joint implicit feature activation secondary subunit, which is used to perform feature activation based on the Sigmoid function on the temperature-humidity joint implicit feature vector to obtain a temperature-humidity condition feature vector; a temperature-humidity time series feature interactive secondary subunit, which is used to perform conditional dependence interactive fusion on the temperature time series implicit association feature vector and the humidity time series implicit association feature vector based on the temperature-humidity condition feature vector to obtain a temperature-humidity time series feature significant interactive fusion representation vector as the temperature-humidity time series feature significant interactive fusion representation.

4. The intelligent operation control system of the packaging bag production line according to claim 3 is characterized in that: The temperature-humidity joint implicit capture secondary subunit is used to: add the temperature time series implicit correlation feature vector and the humidity time series implicit correlation feature vector by position, multiply the obtained temperature-humidity time series sum vector by the weight matrix, and then add it by position with the bias vector to obtain the temperature-humidity time series joint interaction vector; use the tanh function to process the temperature-humidity time series joint interaction vector to obtain the temperature-humidity joint implicit feature vector.

5. The intelligent operation control system of the packaging bag production line according to claim 4 is characterized in that: The temperature-humidity time series feature interaction secondary subunit includes: a temperature semantic contribution calculation tertiary subunit, which is used to calculate the temperature semantic contribution of the temperature time series implicit association feature vector relative to the temperature-humidity condition feature vector; a humidity semantic contribution calculation tertiary subunit, which is used to calculate the humidity semantic contribution of the humidity time series implicit association feature vector relative to the temperature-humidity condition feature vector; a semantic contribution normalization modulation tertiary subunit, which is used to normalize the temperature semantic contribution and the humidity semantic contribution, and use the normalized temperature semantic contribution and the normalized humidity semantic contribution to weighted modulate the temperature time series implicit association feature vector and the humidity time series implicit association feature vector to obtain a modulated temperature time series implicit association feature vector and a modulated humidity time series implicit association feature vector; a temperature-humidity time series feature guided interaction tertiary subunit, which is used to use the modulated temperature time series implicit association feature vector as a query vector, the modulated humidity time series implicit association feature vector as a key vector and the temperature-humidity condition feature vector as a query vector. The conditional feature vector is used as the value vector, and the modulated temperature time series implicit association feature vector, the modulated humidity time series implicit association feature vector and the temperature-humidity conditional feature vector are input into the feature-to-feature significant guided interaction module based on the converter structure to obtain the temperature-humidity time series feature significant interaction fusion representation vector; wherein the semantic contribution normalization modulation three-level sub-unit is used to: calculate the sum of the temperature semantic contribution and the humidity semantic contribution to obtain the temperature-humidity semantic contribution sum; divide the temperature semantic contribution and the humidity semantic contribution by the temperature-humidity semantic contribution sum respectively to obtain the normalized temperature semantic contribution and the normalized humidity semantic contribution; perform position point multiplication of the temperature time series implicit association feature vector and the normalized temperature semantic contribution to obtain the modulated temperature time series implicit association feature vector; perform position point multiplication of the humidity time series implicit association feature vector and the normalized humidity semantic contribution to obtain the modulated humidity time series implicit association feature vector.

6. The intelligent operation control system of the packaging bag production line according to claim 5 is characterized in that: The temperature semantic contribution calculation three-level sub-unit is used to: calculate the positional division of the corresponding positions of the temperature time series implicit association feature vector and the temperature-humidity condition feature vector to obtain a temperature semantic contribution vector; calculate the logarithmic function value with base two of the absolute value of each eigenvalue of the temperature semantic contribution vector to obtain a temperature semantic contribution logarithmic vector; calculate the positional dot product of the temperature semantic contribution logarithmic vector and the temperature semantic contribution logarithmic vector, and perform positional multiplication on the obtained dot product vector to obtain the temperature semantic contribution value; calculate an exponential function with the natural constant e as the base and the temperature semantic contribution value as the exponent to obtain the temperature semantic contribution.

7. The intelligent operation control system of the packaging bag production line according to claim 6 is characterized in that: The spatial topology extraction unit is used to: construct a transmission sub-area spatial topology matrix of the multiple transmission sub-areas, and pass the transmission sub-area spatial topology matrix through a topological feature extractor based on a void convolutional neural network to obtain a transmission sub-area spatial topology feature matrix as the transmission sub-area spatial topology feature.

8. The intelligent operation control system of the packaging bag production line according to claim 7 is characterized in that: The global timing interaction unit is used to: pass multiple temperature-humidity timing feature significant interaction fusion representation vectors and the transmission sub-area spatial topological feature matrix through a global transmission area temperature-humidity timing feature interaction encoder based on a graph convolutional neural network model to obtain a sequence of temperature-humidity timing feature interaction representation vectors under the global transmission area as a sequence of temperature-humidity timing feature interaction representation under the global transmission area.

9. The intelligent operation control system of the packaging bag production line according to claim 8 is characterized in that: The recognition result generating unit is used to extract the temperature-humidity time series feature interaction representation vector of the global transmission area corresponding to the detected transmission sub-area from the sequence of the temperature-humidity time series feature interaction representation vector of the global transmission area, and pass the temperature-humidity time series feature interaction representation vector of the global transmission area through a classifier-based environmental anomaly identifier to obtain a recognition result, and the recognition result is used to indicate whether there is an abnormality in the environmental conditions of the detected transmission sub-area.

10. A method for intelligent operation and control of a packaging bag production line, characterized in that: include: At the entrance of the packaging bag production line, the surface defects of the packaging bags to be printed are inspected and qualified packaging bags to be printed are selected; Transmitting the packaging bags to be printed with qualified quality to a conveying area; The transmission area is divided into sub-areas to obtain multiple transmission sub-areas; environmental condition analysis and anomaly identification are performed on the multiple transmission sub-areas to obtain identification results; wherein, the environmental condition analysis and anomaly identification are performed on the multiple transmission sub-areas to obtain identification results, including: obtaining a time queue of environmental parameter data of multiple transmission sub-areas collected by environmental parameter collection terminals deployed in the multiple transmission sub-areas, wherein the environmental parameter data include temperature values ​​and humidity values; performing conditional feature fusion based on time series features on the time queues of multiple environmental parameter data to obtain multiple temperature-humidity time series feature significant interactive fusion representations; constructing a transmission sub-area spatial topology matrix of the multiple transmission sub-areas and performing topological feature extraction to obtain transmission sub-area spatial topological features; performing global transmission area temperature-humidity time series feature interactive encoding on the multiple temperature-humidity time series feature significant interactive fusion representations and the transmission sub-area spatial topological features to obtain a sequence of temperature-humidity time series feature interactive representations under the global transmission area; based on the sequence of temperature-humidity time series feature interactive representations under the global transmission area, obtaining identification results of each transmission sub-area.

Images