A method for real-time acquisition and analysis of printing production data

Abstract

The application particularly relates to the technical field of big data analysis, and discloses a printing production data real-time collection and analysis method, which comprises the following steps: S1, multi-source data collection: collecting multi-dimensional printing production data; S2, edge real-time processing: obtaining a multi-dimensional printing production data set and performing real-time analysis; S3, data depth analysis: obtaining a printing production quality evaluation index and predicting a printing equipment failure probability; S4, optimization strategy generation: formulating a printing production optimization strategy; S5, automatic control execution: automatically adjusting parameters of a printing production equipment according to an equipment adjustment instruction; and S6, execution result feedback. The multi-dimensional printing production data is collected through a data collection terminal, data processing and analysis are performed through an edge computing node in combination with a cloud server, a multi-dimensional analysis model and a prediction model are constructed through the cloud server, and an equipment adjustment instruction is executed through an automatic execution terminal, so that the product qualification rate and the production stability are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of big data analytics, and more specifically, to a method for real-time acquisition and analysis of printing production data. Background Technology

[0002] With industrial development, global manufacturing has entered a new stage of development, shifting from processing-oriented industries to high-tech industries. As society progresses, consumers' demand for personalized printed products is increasing, requiring printing companies to collect and analyze data in real time to quickly respond to market changes and achieve large-scale personalized customization production. The maturity of Industrial Internet of Things (IIoT), big data, and artificial intelligence technologies provides the technical feasibility for real-time data collection and analysis. Existing printing data collection and analysis methods specifically include printing data acquisition, printing data preprocessing, data storage and transmission, data analysis and modeling, and decision support and feedback. Through end-to-end data collection, standardized data processing, data analysis, and closed-loop decision-making, real-time acquisition of printing data is achieved, helping companies improve efficiency, optimize quality, and reduce printing costs.

[0003] However, it still has some shortcomings in actual use. First, the data collection dimension is singular. Existing printing data collection and analysis methods mainly focus on the basic operating parameters and production progress parameters of printing equipment, while the collection of implicit key data is insufficient, making it difficult to accurately locate the root cause and increasing the cost of quality control.

[0004] Second, the real-time analysis capability is insufficient. Existing printing data acquisition and analysis methods mostly adopt traditional batch processing modes, which cannot meet the requirements of real-time analysis. Furthermore, there is a lack of efficient flow framework deployment plans, making it difficult to quickly process and analyze large amounts of data, resulting in the inability to detect abnormal situations in the printing process in a timely manner.

[0005] Third, data analysis capabilities are limited. Existing printing data collection and analysis methods only perform statistical analysis, lacking the ability to deeply mine and model data, making it difficult to achieve scientific decision-making and production optimization. Summary of the Invention

[0006] In view of this, embodiments of the present invention provide a method for real-time acquisition and analysis of printing production data. The method collects multi-dimensional printing production data through a data acquisition terminal, performs data processing and analysis through edge computing nodes combined with a cloud server, and constructs multi-dimensional analysis and prediction models through the cloud server. This effectively solves the problems of single data acquisition dimensions, insufficient real-time analysis capabilities, and limited data analysis capabilities mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for real-time acquisition and analysis of printing production data, comprising a data acquisition terminal, an edge computing node, a cloud server, an automatic control terminal, and a mobile interactive terminal, the steps of which are as follows:

[0008] S1: Multi-source data acquisition: Real-time acquisition of multi-dimensional printing production data through data acquisition terminals, including printing quality data, printing efficiency data and equipment status data, and transmission of the real-time acquired data to edge computing nodes;

[0009] S2: Real-time edge processing: Edge computing nodes perform data cleaning and feature extraction on multi-dimensional printing production data to obtain a multi-dimensional printing production dataset. They then perform real-time analysis using preset thresholds and transmit the multi-dimensional printing production dataset to the cloud server.

[0010] S3: In-depth data analysis: The cloud server receives and stores the printing production dataset, builds a printing production data analysis model, calculates the printing production quality assessment index, and builds a prediction model to predict the probability of printing equipment failure.

[0011] S4: Optimization Strategy Generation: Based on the printing production quality assessment index, the printing production quality is judged, and the printing production optimization strategy is formulated by combining the failure probability of printing equipment and the printing production quality, and then distributed to the edge computing nodes.

[0012] S5: Automatic control execution: The edge computing node receives the printing production optimization strategy, generates equipment adjustment instructions, sends them to the automatic control terminal, and automatically adjusts the parameters of the printing production equipment according to the equipment adjustment instructions;

[0013] S6: Execution result feedback: The adjusted multi-dimensional printing production data is collected in real time using the data acquisition terminal and fed back to the cloud server. The data in the printing production process is displayed in real time through the human-computer interaction terminal.

[0014] The technical effects and advantages of this invention are as follows:

[0015] This invention collects multi-dimensional printing production data through a data acquisition terminal, including printing quality data, printing efficiency data, and equipment status data, forming a complete data link between equipment, materials, and quality. This breaks through the limitations of traditional methods that only collect basic equipment parameters, improves the accuracy of printing production quality control, avoids repeated trial and error caused by missing data in a single dimension, and increases the product qualification rate.

[0016] This invention uses edge computing nodes combined with cloud servers for data processing and analysis. It adopts an architecture of edge computing nodes + cloud collaboration. The edge computing nodes perform real-time cleaning and threshold judgment at the data source and use the threshold judgment to trigger device protection commands. The cloud devices are responsible for complex model analysis, which meets the real-time monitoring needs of high-speed printing production and improves production stability.

[0017] This invention constructs multi-dimensional analysis and prediction models through cloud servers, calculates the printing production quality assessment index through a weighted average model, and achieves a leap from basic statistics to in-depth prediction. By quantitatively assessing production quality and equipment status, it formulates precise preventive maintenance plans, reduces unplanned downtime, and improves production efficiency and automation. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the method steps of the present invention.

[0019] Figure 2 This is a schematic diagram of the overall structure of the present invention.

[0020] Figure 3 This is a schematic diagram of the production status level determination steps of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] As attached Figure 1 The method for real-time acquisition and analysis of printing production data shown includes a data acquisition terminal, an edge computing node, a cloud server, an automatic control terminal, and a mobile interactive terminal.

[0023] In a more specific application of this invention, the data acquisition terminal is used to acquire physical data, image data, material information, and equipment status in real time during the printing production process, covering multi-dimensional data. Specifically, it includes acquisition devices of sensor type, visual inspection type, label recognition type, and device interface type. The sensor type can be a vibration sensor and a pressure sensor, the visual inspection type can be an industrial camera and a spectrometer, the label recognition type can be an RFID tag reader and a barcode scanner, and the device interface type can be a PLC data acquisition module. It connects to the edge computing node and the automatic control terminal through an industrial Ethernet to ensure the stability of data transmission.

[0024] Edge computing nodes are used to preprocess the raw data uploaded by data acquisition terminals, perform local real-time analysis, and unify the heterogeneous protocols of different data acquisition terminals. Specific devices include edge computing gateways and industrial-grade switches. The industrial-grade switches are used to realize the aggregation and distribution of data from multiple terminals. Edge computing nodes connect to cloud servers through 5G networks to transmit preprocessed structured data, and connect data acquisition terminals and automatic control terminals through industrial buses to form a localized data closed loop.

[0025] Cloud servers are used for data storage and management, in-depth data analysis and modeling, and remote monitoring and collaboration. Specifically, they include cloud server clusters and data platforms. Cloud server clusters are used to support elastic scaling of computing power, while data platforms are used to integrate production data and enterprise management data to build a unified database. Cloud servers are connected to mobile interactive terminals via 5G networks.

[0026] The automatic control terminal is used to automatically adjust printing production parameters according to the instructions transmitted by the edge computing node, including triggering equipment start and stop, parameter switching, and alarm shutdown. Specifically, it includes a PLC controller, an industrial robot, and intelligent actuators. The PLC controller is used to receive control instructions from the edge computing node and drive the equipment to execute them. The industrial robot is used to automatically complete the operations of material loading and unloading and defective product sorting. The intelligent actuators include electric regulating valves and servo motors. The automatic control terminal is connected to the edge computing node through an industrial bus to receive real-time control signals.

[0027] Mobile interactive terminals are used for on-site data entry and query, early warning and collaborative response, and visual interaction. Specifically, they include industrial PDAs or tablets and smart wearable devices, which connect to cloud servers via 5G networks, support offline mode to download cached data and synchronize it after connecting to the network.

[0028] For the connection methods of the aforementioned data acquisition terminals, edge computing nodes, cloud servers, automatic control terminals, and mobile interactive terminals, please refer to [link / reference needed]. Figure 2 .

[0029] The specific embodiments of the present invention include the following steps:

[0030] S1: Multi-source data acquisition: Real-time acquisition of multi-dimensional printing production data through data acquisition terminals, including printing quality data, printing efficiency data and equipment status data, and transmission of the real-time acquired data to edge computing nodes;

[0031] Furthermore, printing quality data includes color compliance rate, registration error, number of misprinted prints, total number of prints, dot gain, and number of defective prints. Printing efficiency data includes effective production time, number of prints produced, and plate changeover efficiency. Equipment status data includes vibration speed, maximum cylinder temperature, ambient temperature, cylinder length, running time, and number of malfunctions.

[0032] In this embodiment, it should be specifically explained that the detection of color qualification rate requires installing an intelligent industrial camera at the paper delivery end of the printing press to capture images of the printed products in real time, compare them with standard color values ​​to calculate the color difference, and calculate the ratio of the number of qualified products to the total number of products detected to obtain the color qualification rate; the detection of registration error requires deploying a laser displacement sensor at the paper delivery end to measure the physical offset of the registration position; the detection of the number of misprinted products and the number of defective products requires deploying an industrial camera at the end of the printing press to continuously scan the printed products, identify misprinted areas and other defects, and count the number of defective products; the measurement of dot gain requires deploying a high-resolution industrial camera at the sampling inspection station to capture the dot area, and the software analyzes the degree of dot edge diffusion to calculate the percentage difference between the actual dot area and the theoretical value.

[0033] Effective production time and running time need to be recorded in real time by PLC, including power-on, shutdown, and standby status of the equipment. Non-production time such as plate change, failure, and manual intervention is excluded. The effective production time is the time of continuous operation and production of qualified products. The total running time includes all start-up and shutdown cycles. The number of printed products produced needs to be monitored in real time by deploying a photoelectric counter at the end of the production line. Plate change efficiency is determined by scanning the RFID tag of the plate roller at the start of plate change and stopping the timing after plate change is completed and calibrated. The ratio of the standard time to the actual plate change time is the plate change efficiency.

[0034] Vibration frequency detection requires deploying an acceleration sensor on the printing press roller bearing housing to collect vibration acceleration signals in real time and convert them into vibration velocity; roller temperature and ambient temperature are detected by deploying infrared temperature sensors on the roller surface and workshop walls, with the sensors transmitting temperature data in real time; the number of faults is detected by counting when the PLC receives abnormal signals from the equipment.

[0035] S2: Real-time edge processing: Edge computing nodes perform data cleaning and feature extraction on multi-dimensional printing production data to obtain a multi-dimensional printing production dataset. They then perform real-time analysis using preset thresholds and transmit the multi-dimensional printing production dataset to the cloud server.

[0036] Furthermore, data cleaning specifically includes outlier removal, missing value processing, and data standardization; feature extraction specifically includes time-domain feature extraction, frequency-domain feature extraction, and image feature extraction; thresholds specifically include color deviation threshold, registration error threshold, vibration speed threshold, and roller temperature threshold; and edge computing nodes compare the real-time received color difference pass rate, registration error, vibration speed, and roller temperature with preset thresholds to generate equipment adjustment instructions and transmit them to the automatic control terminal to adjust the production parameters of the printing equipment in real time.

[0037] In this embodiment, it is necessary to specifically explain that outlier removal, missing value processing, and data standardization unify the physical quantities of different sensors into dimensionless values ​​and map multi-dimensional printing production data to the range [0,1] for easier subsequent analysis and processing.

[0038] Real-time analysis is performed using thresholds. For example, when the real-time detected registration error is greater than the threshold of 0.1mm, the equipment fine-tuning program is automatically triggered, and a correction command of 0.5mm is sent to the printing press PLC via the OPC UA protocol to increase or decrease the error. When the roller temperature is greater than 85 degrees Celsius, the edge computing node immediately reduces the printing speed and sends an emergency notification to the cloud server, which is then transmitted to the human-machine interface terminal to notify the relevant staff.

[0039] S3: In-depth data analysis: The cloud server receives and stores the printing production dataset, builds a printing production data analysis model, calculates the printing production quality assessment index, and builds a prediction model to predict the probability of printing equipment failure.

[0040] Furthermore, calculating the printing production quality assessment index requires obtaining the printing quality assessment index y1, the printing efficiency assessment index y2, and the equipment condition assessment index y3. These indices are then combined using the formula... The printing production quality assessment index y is calculated, and d1, d2 and d3 represent the weight coefficients of the printing quality assessment index, printing efficiency assessment index and equipment status assessment index, respectively, with a total of 1.

[0041] Furthermore, obtaining the print quality assessment index requires setting a time window and obtaining the color pass rate η of individual printed products in the print production dataset within that time window. 1i overprinting error d i and the network expansion rate r i The average color qualification rate η1 and the average registration error d are obtained. The ratio of the difference between the single dot gain rate and the planned dot gain rate r to the planned dot gain rate is calculated using the formula. Obtain the dot gain deviation r ai And obtain the mean deviation r of the dot expansion rate. aObtain the number of misprinted products (n1), the total number of prints (n), and the number of defective products (n2) from the print production dataset within the specified time window. Calculate the ratio of the number of misprinted products and the number of defective products to the total number of prints to obtain the misprinting defect rate (r). h and total defect rate r b The color pass rate, registration error, dot gain deviation, smear printing defect rate, and total defect rate are calculated using the following formula:

[0042] ,

[0043] The printing quality evaluation index y1,r is obtained. 标 Indicates the standard dot gain deviation, Δr max Indicates the maximum permissible dot gain deviation, d max Indicates the maximum allowable overprinting error, r hmax This represents the maximum permissible rate of stencil printing defects, r. bmax This represents the maximum permissible total defect rate. a1, a2, a3, a4, and a5 represent the corresponding weighting coefficients, and their sum is 1.

[0044] Obtaining the printing efficiency evaluation index requires acquiring the effective production time t, the number of printed products N, and the plate change efficiency η2 from the printing production dataset within the time window. The time utilization rate η is then calculated by dividing the effective production time by the maximum theoretical production time. t The performance utilization rate η is obtained by calculating the ratio of the number of printed products produced to the maximum theoretical number of printed products. N The formulas utilize time utilization, performance utilization, and changeover efficiency:

[0045] ,

[0046] The printing efficiency evaluation index y2 is obtained, where η 2min and η 2max b1, b2, and b3 represent the minimum and maximum values ​​of the page change efficiency, respectively, and the weighting coefficients of time utilization, performance utilization, and page change efficiency are respectively, with a total of 1.

[0047] Obtaining the equipment condition assessment index requires the vibration velocity V and the maximum roller temperature T from the printing production data within the time window. 1m The ambient temperature T2, drum length l, running time t1, and number of failures n3 are used to calculate the maximum drum temperature, ambient temperature, and drum length using the formula. The temperature gradient D is calculated, and the mean time between failures (MTBF) is obtained by calculating the ratio of operating time to the number of failures. w Using vibration velocity, temperature gradient, and mean time between failures (MTBF) as an example, the formula is:

[0048] ,

[0049] The equipment condition assessment index y3 is obtained, where exp represents the exponential decay model, and V 临界 The critical value of the vibration frequency is represented by G1, which is a piecewise linear model, t wmin and t wmax c1, c2, and c3 represent the minimum and maximum mean time between failures (MTBF), respectively. c1, c2, and c3 represent the weighting coefficients of vibration velocity, temperature gradient, and MTBF, respectively, and their sum is 1.

[0050] In this embodiment, it should be specifically noted that all weighting coefficients in the above calculation steps are obtained based on the analysis of a large amount of printing production data and are set by professional staff.

[0051] Piecewise linear model G 临界 and G 理想 These represent the critical temperature gradient and the ideal temperature gradient, respectively.

[0052] The construction of the prediction model requires dividing the multi-dimensional printing production dataset into training, validation, and test sets, handling undersampling and oversampling class imbalance, performing cross-validation, parameter tuning on the validation set, and evaluation on the test set. Evaluation metrics include classification metrics and time-series prediction metrics. Hyperparameter tuning, ensemble learning, and transfer learning are then performed, and the failure probability of the printing equipment is obtained by using the output of the deployed prediction model.

[0053] S4: Optimization Strategy Generation: Based on the printing production quality assessment index, the printing production quality is judged, and the printing production optimization strategy is formulated by combining the failure probability of printing equipment and the printing production quality, and then distributed to the edge computing nodes.

[0054] Furthermore, judging the quality of printing production requires the construction of a production quality assessment threshold. The printing production quality assessment index calculated in real time is compared with the production quality assessment threshold to classify the production status into four levels, from high to low: Level 1, Level 2, Level 3, and Level 4. Production optimization strategies are then formulated by combining the production status level with the failure probability of printing equipment.

[0055] Furthermore, such as Figure 3 As shown, the steps for determining the production status level are as follows:

[0056] A1: When the printing production quality assessment index is greater than 0.8 and less than 1, it indicates that the production status belongs to the first level;

[0057] This indicates that the production status is at an excellent level. If the current process is maintained, there is room to explore ways to improve efficiency.

[0058] A2: When the printing production quality assessment index is greater than 0.6 and less than 0.8, it indicates that the production status belongs to the second level;

[0059] This indicates that the production status is at a good level, but local optimization is needed, such as fine-tuning parameters or implementing preventative measures.

[0060] A3: When the printing production quality assessment index is greater than 0.6 and less than 0.4, it indicates that the production status belongs to the third level;

[0061] This indicates that the production status is at a normal level, but there are obvious shortcomings. A detailed analysis is needed to determine if there are any problems with the printing quality, printing efficiency, and equipment status.

[0062] A4: When the printing production quality assessment index is less than 0.4, it indicates that the production status belongs to the fourth level.

[0063] This indicates that the production status is poor and requires immediate intervention to avoid systemic risks.

[0064] In this embodiment, it is necessary to specifically explain the production optimization strategy. If quality defects occur frequently, such as a smearing defect rate >1% and a color pass rate <80%, the ink viscosity and printing pressure need to be automatically adjusted, the scanning frequency increased, the current batch of products marked, and the batch traceability process initiated. If an efficiency bottleneck occurs, such as a time utilization rate <60% and a plate change efficiency <0.5, a standardized video of the plate change operation needs to be pushed to the human-machine interaction terminal to guide the production line staff, analyze similar historical production data, and recommend the optimal combination of process parameters. When equipment failure warnings and a sudden drop in quality index occur simultaneously, equipment shutdown protection should be prioritized, production should be suspended, and an emergency maintenance process should be initiated to avoid further losses.

[0065] If the predicted short-term probability of printing equipment failure is greater than 80%, it is necessary to reduce the equipment operating speed, reduce the mechanical load, notify maintenance personnel to prepare spare parts, adjust the production schedule, and transfer subsequent orders to standby equipment.

[0066] S5: Automatic control execution: The edge computing node receives the printing production optimization strategy, generates equipment adjustment instructions, sends them to the automatic control terminal, and automatically adjusts the parameters of the printing production equipment according to the equipment adjustment instructions;

[0067] In this embodiment, it should be specifically explained that the instructions generated by the edge computing node include parameter adjustment type and device action type. The standardized instructions are converted into the device's native protocol, and parameter boundary checks, device status checks, and security protection mechanisms are performed on the instructions. The automatic control terminal performs analog quantity control, digital quantity control, and multi-device collaborative control to achieve closed-loop control.

[0068] S6: Execution result feedback: The adjusted multi-dimensional printing production data is collected in real time using the data acquisition terminal and fed back to the cloud server. The data in the printing production process is displayed in real time through the human-computer interaction terminal.

[0069] In this embodiment, it is necessary to specifically explain that the adjusted multi-dimensional printing production data is monitored in real time and the data is synchronized. The adjusted printing production quality evaluation index is calculated using a cloud server to determine whether the adjustment effect is effective.

[0070] The human-machine interface includes a main monitoring screen and a mobile app. The main monitoring screen includes an equipment status dashboard, a quality trend chart, and an efficiency data wall. The equipment status dashboard displays the vibration speed, temperature, and operating status of each printing press in real time. The quality trend chart dynamically plots the registration error and warning threshold lines in real time. The efficiency data wall displays the calculated data in a scrolling manner. The mobile app includes a personalized dashboard, anomaly push notifications, and historical data queries. The personalized dashboard displays data of interest based on roles, such as quality engineers focusing on defect rates and equipment managers focusing on vibration data. Anomaly push notifications deliver warning information in the form of images, text, and voice. Historical data queries support searching historical production data by time, equipment, and order number, and generating PDF reports.

[0071] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0072] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for real-time acquisition and analysis of print production data, characterized in that, The system comprises a data acquisition terminal, an edge computing node, a cloud server, an automatic control terminal and a mobile interactive terminal, and the steps are as follows: S1: Multi-source data acquisition: The multi-dimensional printing production data, including printing quality data, printing efficiency data and equipment state data, are collected in real time by the data acquisition terminal, and the collected data are transmitted to the edge computing node; S2: Edge real-time processing: The edge computing node performs data cleaning and feature extraction on the multi-dimensional printing production data to obtain a multi-dimensional printing production data set, performs real-time analysis by using a preset threshold, and transmits the multi-dimensional printing production data set to the cloud server; S3: Data deep analysis: The cloud server receives and stores the printing production data set, constructs a printing production data analysis model, calculates a printing production quality evaluation index, and constructs a prediction model to predict the printing equipment failure probability; The calculation of the printing production quality evaluation index requires obtaining a printing quality evaluation index y1, a printing efficiency evaluation index y2 and an equipment state evaluation index y3; The acquisition of the printing quality evaluation index requires setting a time window, obtaining the color pass rate η of a single printed product in the printing production data set in the time window 1i , the overprint error d i , and the dot gain rate r i , obtaining the average color pass rate η1 and the average overprint error d, and calculating the difference between the single dot gain rate and the planned dot gain rate r and the ratio of the planned dot gain rate by the formula to obtain the dot gain deviation r ai , and obtaining the average dot gain deviation r a , obtaining the number of blocked plate printed products n1, the total number of printed products n, and the number of defective printed products n2 in the printing production data set in the time window, respectively calculating the ratio of the number of blocked plate printed products and the number of defective printed products to the total number of printed products to obtain the blocked plate printing defect rate r h and the total defect rate r b , and obtaining the average color pass rate, the average overprint error, the average dot gain deviation, the blocked plate printing defect rate, and the total defect rate by the formula: ， a printing quality evaluation index y1, r is obtained 标 denotes a standard dot gain deviation, Δr max denotes a maximum dot gain deviation allowed, d max denotes a maximum trap error allowed, r hmax denotes a maximum scumming defect rate allowed, r bmax denotes a maximum total defect rate allowed, a1, a2, a3, a4 and a5 respectively denote corresponding weight coefficients, and the sum is 1; S4: Optimization strategy generation: The printing production quality is judged according to the printing production quality evaluation index, and the printing production optimization strategy is formulated in combination with the printing equipment failure probability and the printing production quality, and is issued to the edge computing node; S5: Automatic control execution: The edge computing node receives the printing production optimization strategy, generates a device adjustment instruction, and issues it to the automatic control terminal, which automatically adjusts the parameters of the printing production equipment according to the device adjustment instruction; S6: Execution result feedback: The adjusted multi-dimensional printing production data are collected in real time by the data acquisition terminal, fed back to the cloud server, and displayed in real time in the printing production process through the human-computer interaction terminal.

2. The method for real-time acquisition and analysis of printing production data according to claim 1, characterized in that: The printing quality data include color qualification rate, overprint error, number of blurring printed products, total number of printed products, dot gain rate and number of defective printed products, the printing efficiency data include effective production time, number of printed products and plate changing efficiency, and the equipment state data include vibration speed, maximum cylinder temperature, environment temperature, cylinder length, running time and failure frequency.

3. The method of real-time acquisition and analysis of print production data according to claim 1, characterized in that: The data cleaning specifically includes outlier rejection, missing value processing and data standardization, the feature extraction specifically includes time domain feature extraction, frequency domain feature extraction and image feature extraction, the threshold specifically includes color deviation threshold, overprint error threshold, vibration speed threshold and cylinder temperature threshold, and the edge computing node compares the real-time received color difference qualification rate, overprint error, vibration speed and cylinder temperature with the preset threshold, generates a device adjustment instruction and transmits it to the automatic control terminal to adjust the production parameters of the printing equipment in real time.

4. The method for real-time acquisition and analysis of printing production data according to claim 1, characterized in that: The acquisition of the printing production quality evaluation index requires using a formula to combine the printing quality evaluation index, the printing efficiency evaluation index, and the equipment state evaluation index The printing production quality evaluation index y is calculated, where d1, d2, and d3 represent the weight coefficients of the printing quality evaluation index, the printing efficiency evaluation index, and the equipment state evaluation index, respectively, and the sum is 1.

5. The method for real-time acquisition and analysis of printing production data according to claim 1, characterized in that: The obtaining of the printing efficiency evaluation index requires obtaining effective production time t, production print quantity N and change edition efficiency η2 in a time window, calculating the ratio of the effective production time and the maximum theoretical production time to obtain time utilization efficiency η t , calculating the ratio of the production print quantity and the maximum theoretical production print quantity to obtain performance utilization efficiency η N , using the time utilization efficiency, performance utilization efficiency and change edition efficiency to use the formula: ， an evaluation index y2 of printing efficiency is obtained, where η 2min and η 2max respectively represent the minimum and maximum values of the changeover efficiency, b1, b2 and b3 respectively represent the weight coefficients of the time utilization rate, the performance utilization rate and the changeover efficiency, and the sum is 1.

6. The method for real-time acquisition and analysis of printing production data according to claim 1, characterized in that: The acquisition of the device state evaluation index requires vibration velocity V, maximum cylinder temperature T, ambient temperature T2, cylinder length l, running time t1, and failure number n3 in the printing production data set within a time window 1m The temperature gradient G is calculated using the maximum cylinder temperature, ambient temperature, and cylinder length using the formula The average failure-free time t is calculated by calculating the ratio of running time and failure number w The vibration velocity, temperature gradient, and average failure-free time are used to calculate the device state evaluation index by the formula: ， obtaining a device state evaluation index y3, where exp represents an exponential decay model, V 临界 represents a critical value of the vibration velocity, G1 is a piecewise linear model, t wmin and t wmax respectively represent minimum and maximum values of the mean failure-free time, c1, c2 and c3 respectively represent weight coefficients of the vibration velocity, the temperature gradient and the mean failure-free time, and the sum is 1; Piecewise linear model where G 临界 and G 理想 represent the critical and ideal temperature gradients, respectively.

7. The method of real-time acquisition and analysis of print production data according to claim 1, wherein: The judgment of the printing production quality requires constructing a production quality evaluation threshold, comparing the calculated printing production quality evaluation index with the production quality evaluation threshold, and dividing the production state level, which is ranked from high to low as first level, second level, third level and fourth level, and the production optimization strategy is formulated in combination with the production state level and the printing equipment failure probability.

8. The method of real-time acquisition and analysis of print production data according to claim 7, characterized in that: The judgment process of the production state level is as follows: A1: When the printing production quality evaluation index is greater than 0.8 and less than 1, the production state belongs to the first level; A2: When the printing production quality evaluation index is greater than 0.6 and less than 0.8, it means that the production state belongs to the second grade; A3: When the printing production quality evaluation index is greater than 0.4 and less than 0.6, it means that the production state belongs to the third grade; A4: When the printing production quality evaluation index is less than 0.4, it means that the production state belongs to the fourth grade.

Images