Full-process management system and method for product printing

Abstract

The invention discloses a full-process management system and method for product printing, and relates to the technical field of printing, and the method comprises the following steps: carrying out the continuous collection of multi-dimensional real-time control parameter data generated during the operation of ink-jet printing equipment through a process data collection mechanism in a printing operation process, the whole-process sensing and dynamic monitoring of the printing state are realized; the obtained original real-time control parameter data is preprocessed, and a structured standard data set is established, so that the accuracy, consistency and analysis effectiveness of the data are improved. According to the method, the process data is collected in real time, the ink diffusion trend is recognized in combination with feature engineering and machine learning, ink jet parameters are dynamically regulated and controlled through adaptive pulse width modulation, ink droplet deposition in a non-target area is effectively inhibited, and the recognition rate of image areas such as two-dimensional codes is increased. The method does not need additional hardware, is high in real-time performance and stability, and remarkably improves the quality consistency and yield of printed matters.

Description

Technical Field

[0001] The present invention relates to the field of printing technology, and in particular to a full-process management system and method for product printing. Background Art

[0002] Full-process management of product printing refers to the systematic, standardized, and intelligent collaborative management of every key link in the printing production process. This includes information collection, data analysis, and dynamic optimization scheduling throughout the entire process, from pre-press design document receipt and review, printing task scheduling, equipment parameter setting, material preparation and dispatching, printing execution process monitoring, quality inspection and defect tracing, to post-press processing, finished product warehousing, and delivery management. By establishing a multi-dimensional data interaction mechanism and a closed-loop process control system, this management approach enables real-time perception and closed-loop response to printing production progress, resource allocation, quality status, and abnormal events, thereby improving printing efficiency, stability, and product consistency. It is widely applicable to various application scenarios such as label printing, packaging printing, and color publishing printing.

[0003] The existing technology has the following deficiencies: In the existing technology, inkjet or offset printing is usually used to print QR codes, barcodes and variable data areas during the printing production process. However, due to insufficient control of ink dispersion in the printing environment, especially under high-speed operation or high-air turbulence conditions, ink mist diffusion is very likely to occur. This phenomenon refers to the drift and deposition of ink particles in non-target areas, forming a low-contrast interference pattern or fuzzy band on the printed surface that is difficult to detect with the naked eye but has slight fluctuations in grayscale distribution. This type of pseudo-image is usually located around the QR code, barcode or other variable data area that relies on image recognition. Although it does not constitute obvious contamination, it will seriously interfere with the subsequent machine recognition and scanning and parsing process, causing variable data recognition failure or misreading. The existing technology has failed to effectively identify and suppress such low-contrast grayscale interference. In particular, in online quality inspection systems, traditional image processing algorithms based on edge enhancement or high-contrast extraction often have difficulty perceiving such pseudo-patterns, resulting in missed defects, which in turn affects the usability and pass rate of printed products.

[0004] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a full-process management system and method for product printing. By collecting process data in real time, combining feature engineering and machine learning to identify ink diffusion trends, and using adaptive pulse-width modulation to dynamically control inkjet parameters, this system effectively suppresses ink droplet deposition in non-target areas and improves the recognition rate of image regions such as QR codes. This method requires no additional hardware, exhibits strong real-time performance, and is highly stable, significantly improving the quality consistency and yield of printed products, thereby addressing the aforementioned issues in the background art.

[0006] In order to achieve the above object, the present invention provides the following technical solution: a full-process management method for product printing, comprising the following steps:

[0007] During the printing process, the process data collection mechanism is used to continuously collect real-time control parameter data of multiple dimensions generated during the operation of the inkjet printing equipment, realizing full-process perception and dynamic monitoring of the printing status;

[0008] Pre-process the acquired raw real-time control parameter data and establish a structured standard data set to improve data accuracy, consistency and analysis effectiveness;

[0009] Through feature engineering methods, key features reflecting the ink mist diffusion state are accurately extracted from the data set, and the extracted key features are comprehensively analyzed to quantify the ink mist diffusion degree;

[0010] The key features that have undergone comprehensive analysis and quantification are constructed into feature vectors and input into a pre-trained machine learning model. The machine learning model then quickly and accurately identifies and analyzes the feature vectors, intelligently assessing whether there is ink mist diffusion in the current printing process.

[0011] When the evaluation results of the machine learning model indicate that ink mist diffusion occurs in the current printing process, the width of the inkjet pulse is dynamically shortened through the adaptive pulse width modulation method to reduce the initial kinetic energy of ink droplets generated in each injection unit, quickly control the ink droplet particle size and speed, and quickly suppress the aerial diffusion of ink particles from a physical level, reducing pollution and grayscale disturbances in non-target areas.

[0012] Preferably, during the printing operation, a process data collection mechanism is used to continuously collect real-time control parameter data of multiple dimensions generated during the operation of the inkjet printing device. The specific steps are as follows:

[0013] First, define the acquisition parameter range and clarify the key control parameter dimensions that need to be monitored;

[0014] Secondly, deploy collection interfaces and perception nodes to achieve seamless access to various parameter sources based on existing equipment control system interfaces or embedded sensor modules;

[0015] Then, the data sampling frequency and synchronization mechanism are set to ensure that the parameters of each dimension are continuously collected with a millisecond sampling period under a unified time base, thus constructing a high-precision, multi-channel time series data stream;

[0016] Finally, a data cache and upload channel is established to transmit the real-time collected data to the edge controller or central processing platform, forming a continuous and stable data input stream, providing a data basis for subsequent process status analysis, intelligent identification and dynamic control.

[0017] Preferably, key features reflecting the diffusion state of ink micro-mist are accurately extracted from the data set through feature engineering methods. The extracted features include the spatial migration trend of particles in the ink under the action of the thermal flow field and the fluctuation of the initial kinetic energy of the ink droplets driven by the inkjet pulse. The spatial migration trend of particles in the ink under the action of the thermal flow field and the fluctuation of the initial kinetic energy of the ink droplets driven by the inkjet pulse are comprehensively analyzed under the detection window to generate ink particle migration sensitive reference values and pulse kinetic energy fluctuation reference values, respectively. The ink micro-mist diffusion degree is quantified by the ink particle migration sensitive reference values and the pulse kinetic energy fluctuation reference values.

[0018] Preferably, the specific steps of comprehensively analyzing the spatial migration trend of the particles in the ink under the action of the thermal flow field in the detection window to generate the ink particle migration sensitivity reference value are as follows:

[0019] Based on the actual deposition trajectory of ink droplets during inkjet imaging, a droplet spatial offset vector field is constructed and superimposed with the corresponding thermal disturbance vector field to characterize the disturbance effect of the thermal flow field on the droplet trajectory. The thermal response offset rate is obtained by calculating the weighted projection relationship between the droplet offset vector and the thermal disturbance direction. The calculation formula is as follows:

[0020]

[0021] Where, is the thermal response offset rate, is the ink droplet spatial offset vector field, which represents the spatial offset vector of the ink droplet at the coordinate point (x, y). is the local thermal disturbance vector field, which represents the direction and intensity of the thermal flow disturbance at the coordinate point (x, y), Ω is the detection window range, and λ is the trend enhancement factor, which is used to apply nonlinear weight adjustment to the dot product result and the thermal disturbance modulus. is the modulus of the thermal disturbance vector;

[0022] On the basis of obtaining the thermal response offset rate, the inkjet direction offset dispersion is further introduced to reflect the overall discrete distribution of the ink drop offset angle relative to the theoretical inkjet direction. The calculation formula of the inkjet direction offset dispersion is as follows:

[0023]

[0024] Where, δ θ is the directional discreteness, which is used to quantify the overall discreteness of the inkjet offset direction within the window. θ(x, y) is the offset direction angle, which indicates the spatial offset vector of the ink droplet at the coordinate point (x, y). The angle formed with the ideal inkjet direction, θ ref is the reference inkjet direction, η is the offset enhancement factor, |Ω| is the total number of valid coordinate points contained in the detection window area Ω;

[0025] Finally, the thermal response offset rate and the directional offset discreteness are combined to generate the ink particle migration sensitivity reference value, which is defined as follows:

[0026]

[0027] Where PMSR is the ink particle migration sensitivity reference value.

[0028] Preferably, the specific steps of comprehensively analyzing the fluctuation of the initial kinetic energy of the ink droplet under the inkjet pulse drive in the detection window to generate the pulse kinetic energy fluctuation reference value are as follows:

[0029] By performing multi-dimensional fusion processing on the driving parameters collected by the printing equipment during the inkjet operation, a response factor that characterizes the initial kinetic energy state of the ink droplet is formed based on the inkjet head driving voltage, pulse width, jetting frequency, and nozzle flow resistance disturbance factor. The formula for the inkjet pulse kinetic energy response factor is as follows:

[0030]

[0031] Where V i is the inkjet drive voltage, which represents the instantaneous voltage applied by the inkjet drive control unit to the inkjet head during the i-th inkjet pulse, τ i is the pulse width, which indicates the length of time the inkjet pulse signal remains at a high level in the i-th inkjet pulse, ζ i is the nozzle flow resistance disturbance factor, f i is the injection frequency, tanh is the hyperbolic tangent function, Λ i is the inkjet pulse kinetic energy response factor, which represents the actual kinetic energy response level obtained by the ink droplet under the i-th inkjet pulse;

[0032] The kinetic energy response factors of multiple inkjet pulses during the continuous inkjet cycle are obtained. i Afterwards, a dynamic analysis of the sequence change trend within the detection window is performed to quantify the continuous stability of the kinetic energy output process. The following trend function is used to define the pulse kinetic energy fluctuation reference value. The formula is as follows:

[0033]

[0034] Where, Λ i-1is the inkjet pulse kinetic energy response factor under the previous inkjet pulse, n is the total number of inkjet pulses, ω is the trend enhancement coefficient, and PKFR is the pulse kinetic energy fluctuation reference value.

[0035] Preferably, the ink particle migration sensitivity reference value and pulse kinetic energy fluctuation reference value that have been comprehensively analyzed and quantified are constructed into feature vectors and input into a pre-trained machine learning model. The ink micro-mist diffusion risk coefficient is generated by the model, and the ink micro-mist diffusion risk coefficient is used to intelligently evaluate whether there is ink micro-mist diffusion in the current printing process.

[0036] Preferably, the ink mist diffusion risk coefficient is compared with a preset ink mist diffusion risk coefficient reference threshold value to evaluate whether there is ink mist diffusion in the current printing process. The specific steps are as follows:

[0037] If the ink micro-mist diffusion risk coefficient is greater than the pre-set ink micro-mist diffusion risk coefficient reference threshold, it indicates that ink micro-mist diffusion occurs in the current printing process; if the ink micro-mist diffusion risk coefficient is less than or equal to the pre-set ink micro-mist diffusion risk coefficient reference threshold, it indicates that ink micro-mist diffusion does not occur in the current printing process.

[0038] Preferably, when the machine learning model evaluation results indicate that ink mist diffusion occurs during the current printing process, the adaptive pulse width modulation method is used to dynamically shorten the inkjet pulse width to reduce the initial kinetic energy of ink droplets generated in each ejection unit, and the specific steps for quickly controlling the ink droplet size and speed are as follows:

[0039] When the machine learning model evaluation results indicate that ink mist diffusion occurs during the current printing process, the inkjet pulse width is slightly shortened based on pulse width modulation technology to reduce the initial kinetic energy of the ink droplets and suppress the airborne diffusion trend of the ink particles. The inkjet pulse width adjustment formula is as follows:

[0040]

[0041] Where, IMDRC is the ink mist diffusion risk coefficient, IMDRC ref is the reference threshold of the ink mist diffusion risk coefficient, α is the adaptive adjustment sensitivity coefficient, which is used to control the proportional factor of the inkjet pulse width shortening amplitude, and its value range is 0<α<1. std is the standard inkjet pulse width, PW new is the adjusted inkjet pulse width;

[0042] After dynamic pulse width adjustment, the real-time feedback evaluation of the ink droplet initial kinetic energy control effect is conducted to evaluate the change in the ink droplet initial kinetic energy before and after the adjustment to verify the diffusion suppression effect brought about by the shortened pulse width. The formula for calculating the initial kinetic energy change ratio is as follows:

[0043]

[0044] Where, ΔE inkdrop is the change ratio of the initial kinetic energy of the ink droplet, and β is the nonlinear response coefficient.

[0045] Preferably, a full-process management system for product printing includes a printing process data acquisition module, a data preprocessing and standardization module, a feature extraction and diffusion quantification module, an ink diffusion intelligent evaluation module, and an adaptive inkjet control and adjustment module.

[0046] The printing process data acquisition module uses the process data acquisition mechanism to continuously collect real-time control parameter data of multiple dimensions generated during the operation of the inkjet printing equipment during the printing process, realizing the full process perception and dynamic monitoring of the printing status;

[0047] The data preprocessing and standardization module preprocesses the acquired raw real-time control parameter data and establishes a structured standard data set to improve data accuracy, consistency, and analysis effectiveness;

[0048] The feature extraction and diffusion quantification module uses feature engineering methods to accurately extract key features that reflect the diffusion state of ink micro-mist from the data set, and conducts comprehensive analysis on the extracted key features to quantify the diffusion degree of ink micro-mist;

[0049] The intelligent ink diffusion assessment module constructs feature vectors from key features that have been comprehensively analyzed and quantified, and inputs them into a pre-trained machine learning model. The machine learning model then quickly and accurately identifies and analyzes the feature vectors, intelligently assessing whether there is ink mist diffusion during the current printing process.

[0050] The adaptive inkjet control and adjustment module, when the machine learning model evaluation results indicate that ink micro-mist diffusion occurs in the current printing process, dynamically shortens the width of the inkjet pulse through the adaptive pulse width modulation (PWM) method to reduce the initial kinetic energy of ink droplets generated in each injection unit, quickly control the ink droplet particle size and speed, and quickly suppress the aerial diffusion of ink particles from a physical level, reducing pollution and grayscale disturbances in non-target areas.

[0051] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0052] This method, based on the real-time collection and standardized processing of process data, combines feature engineering with machine learning models to accurately identify ink particle migration trends and inkjet kinetic energy fluctuations. This allows for rapid identification of abnormal ink diffusion conditions at the initial stage. Adaptive pulse width modulation is then used to timely adjust inkjet parameters, suppressing ink droplet deposition in non-target areas and significantly reducing interference and misjudgment risks in variable image areas such as QR codes and barcodes. This solution, which does not rely on additional hardware modifications, offers high real-time performance, strong compatibility, and excellent process stability, significantly improving the overall quality consistency and production yield of printed products, meeting the stringent requirements for image readability and finished product quality in high-precision, high-reliability printing scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction to the drawings required for use in the embodiments will be given below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0054] Figure 1 The present invention is a flowchart of a method for managing the entire process of product printing.

[0055] Figure 2 This is a module diagram of a full-process management system for product printing according to the present invention. DETAILED DESCRIPTION

[0056] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of this disclosure will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.

[0057] The present invention provides Figure 1 The full-process management method of a product printing shown includes the following steps:

[0058] During the printing process, the process data collection mechanism is used to continuously collect real-time control parameter data of multiple dimensions generated during the operation of the inkjet printing equipment, realizing full-process perception and dynamic monitoring of the printing status;

[0059] During the printing process, a process data acquisition mechanism is used to continuously collect real-time control parameter data in multiple dimensions generated during the operation of the inkjet printing equipment. This refers to the real-time acquisition of key operating parameters of the printing equipment during operation through the system's built-in data monitoring module, such as inkjet pulse width, nozzle voltage and current, jet frequency, ink supply pressure, ink temperature, ambient humidity, and other multi-dimensional data. The purpose of this process is to comprehensively and dynamically perceive the operating status and changing trends of the printing process, forming a real-time information map covering inkjet control, ink behavior, and environmental factors, thereby providing basic support for subsequent process optimization, abnormality warning, and intelligent regulation. Through this full-process perception and dynamic monitoring mechanism, the controllability of printing quality and system responsiveness can be significantly improved, potential defects can be avoided, and the stability and consistency of the production process can be ensured.

[0060] During the printing process, the process data acquisition mechanism is used to continuously collect real-time control parameter data of multiple dimensions generated during the operation of the inkjet printing equipment. The specific steps are as follows: First, define the acquisition parameter range and clarify the key control parameter dimensions that need to be monitored, such as inkjet pulse width, voltage, current, ink supply pressure, ink temperature, nozzle ejection frequency, and ambient temperature and humidity; second, deploy acquisition interfaces and perception nodes, and achieve seamless access to various parameter sources based on existing equipment control system interfaces or embedded sensor modules; then, set the data sampling frequency and synchronization mechanism to ensure that parameters of each dimension are continuously collected with a millisecond sampling period under a unified time base, and build a high-precision, multi-channel time series data stream; finally, establish data caching and upload channels to transmit the real-time collected data to the edge controller or central processing platform, forming a continuous and stable data input stream, providing a data basis for subsequent process status analysis, intelligent identification and dynamic control.

[0061] Pre-process the acquired raw real-time control parameter data and establish a structured standard data set to improve data accuracy, consistency and analysis effectiveness;

[0062] Preprocessing the acquired raw real-time control parameter data and establishing a structured standard data set means that after collecting the raw multi-dimensional control data from the printing equipment's operation, the data is first cleaned, including removing missing values, eliminating outliers, correcting timestamp errors, and unifying the unit format. Subsequently, data from different sources and dimensions are standardized and normalized, converting them into a unified data set with a consistent structure, standardized format, and time-series alignment. This process aims to improve data accuracy and consistency, eliminate interference caused by data noise or chaotic formatting, and lay a solid foundation for subsequent feature extraction, model training, and intelligent analysis, thereby ensuring that the data analysis results of the entire printing process monitoring and control system are more reliable, stable, and efficient.

[0063] Through feature engineering methods, key features reflecting the ink mist diffusion state are accurately extracted from the data set, and the extracted key features are comprehensively analyzed to quantify the ink mist diffusion degree;

[0064] Through feature engineering methods, key features reflecting the diffusion state of ink micro-mist are accurately extracted from the data set. The extracted features include the spatial migration trend of particles in the ink under the action of the thermal flow field and the fluctuation of the initial kinetic energy of the ink droplets driven by the inkjet pulse. The spatial migration trend of particles in the ink under the action of the thermal flow field and the fluctuation of the initial kinetic energy of the ink droplets driven by the inkjet pulse are comprehensively analyzed under the detection window to generate ink particle migration sensitive reference values and pulse kinetic energy fluctuation reference values, respectively. The ink micro-mist diffusion degree is quantified by the ink particle migration sensitive reference values and the pulse kinetic energy fluctuation reference values.

[0065] The ink micro-mist diffusion degree is quantified by the ink particle migration sensitivity reference value and the pulse kinetic energy fluctuation reference value. Its core role is to convert the microscopic ink atomization behavior that is difficult to directly observe and determine during the printing process into a measurable, traceable and controllable numerical evaluation result. The ink particle migration sensitivity reference value reflects the tendency of ink particles to deviate from the target path and drift to non-target areas under thermal flow disturbances, while the pulse kinetic energy fluctuation reference value describes the stability of the inkjet pulse per unit time and its influence on the kinetic energy of the ink droplets. The two together act on the formation and distribution of atomized particles. By comprehensively analyzing these two indicators within the detection window, it is possible to accurately identify whether the ink atomization is in an abnormally active state, thereby realizing real-time identification and trend warning of the ink micro-mist diffusion phenomenon, providing a reliable decision-making basis for the intelligent control system, and fundamentally improving the stability of the printing process and the imaging quality and readability of image recognition areas (such as QR codes and barcodes).

[0066] If the spatial migration of ink particles under the influence of thermal flow fields deviates from their originally intended jetting path, it generally indicates that the ink particles are drifting and depositing in non-target areas, representing a state of ink mist diffusion. From a microscopic physical perspective, during the ink jetting process, ink droplets should be deposited along a predetermined trajectory in the target printing area, driven by controlled kinetic energy. However, when thermal flow fields generated by equipment operation (such as UV light source heating, equipment heat dissipation, and localized convective airflow) create temperature gradients and airflow disturbances at the nozzle outlet or image formation area, ink droplets, especially smaller mist particles, are susceptible to buoyancy, thermal convection, or vortex disturbances, causing them to spatially drift, deviating from the originally intended ink trajectory and depositing at the edges of the image or in the background. This deviation results in deposition, rather than concentrated ink deposition, forming uneven grayscale, low-contrast, and blurred areas, known as ink mist diffusion bands. These diffusion bands often appear around high-resolution areas such as QR codes and barcodes, interfering with image recognition. Therefore, deviations in spatial migration behavior are a direct manifestation of ink particle diffusion and deposition in non-target areas and serve as an important physical indicator of ink mist diffusion.

[0067] The specific steps for comprehensively analyzing the spatial migration trend of ink particles under the action of the thermal flow field within the detection window to generate the ink particle migration sensitivity reference value are as follows:

[0068] Based on the actual deposition trajectory of ink droplets during inkjet imaging, a droplet spatial offset vector field is constructed and superimposed with the corresponding thermal disturbance vector field to characterize the disturbance effect of the thermal flow field on the droplet trajectory. The thermal response offset rate is obtained by calculating the weighted projection relationship between the droplet offset vector and the thermal disturbance direction. The calculation formula is as follows:

[0069]

[0070] Where, is the thermal response deviation rate, which is used to quantify the degree of driving consistency of the thermal flow disturbance on the spatial deviation behavior of the ink droplet. It is the ink droplet spatial offset vector field, which represents the spatial offset vector of the ink droplet at the coordinate point (x, y). It is used to describe whether the ink droplet deviates from the ideal deposition path under the influence of thermal disturbance and other interference factors. It is the core vector basis for measuring "non-target deposition". is the local thermal disturbance vector field, which represents the direction and intensity of the thermal flow disturbance at the coordinate point (x, y). It simulates the directional factors of thermal convection or temperature difference driving during the flight of the ink droplet and is the vector driving source of the thermal field on the migration of the ink droplet. Ω is the detection window range, which represents the range of the local spatial area selected for analysis. λ is the trend enhancement factor, which is used to apply nonlinear weight adjustment to the dot product result and the thermal disturbance modulus. The value range is 0.5-1.5. If λ is set to <1, the response to small offset trends is enhanced, which is suitable for early diffusion identification. If λ>1, the weight of drastic offset trends is enhanced, which is more suitable for obvious offset analysis. is the modulus of the thermal disturbance vector, the Euclidean norm of the thermal disturbance vector, which is used to reflect the intensity of the thermal disturbance in the local area;

[0071] By constructing a spatial offset vector field for ink droplets and a thermal disturbance vector field, we quantify the actual interference effect of the thermal flow field on the ink droplet trajectory, thereby capturing the trend and intensity of ink particle drift in non-target areas. This process accurately extracts the migration and offset behavior caused by thermal flow, providing the core driving characteristics for determining whether the ink is in a micro-mist diffusion state.

[0072] On the basis of obtaining the thermal response offset rate, the inkjet direction offset dispersion is further introduced to reflect the overall discrete distribution of the ink drop offset angle relative to the theoretical inkjet direction. The calculation formula of the inkjet direction offset dispersion is as follows:

[0073]

[0074] Where, δ θ is the directional discreteness, which is used to quantify the overall discreteness of the inkjet offset direction within the window. The larger the value, the more obvious the fluctuation in the offset direction, the more unstable the ink droplet, and the higher the diffusion probability. θ(x, y) is the offset direction angle, which means the spatial offset vector of the ink droplet at the coordinate point (x, y). The angle formed with the ideal inkjet direction, θ ref is the reference inkjet direction, η is the offset enhancement factor, which is used to nonlinearly amplify the weight of large-angle offset in the overall discreteness calculation, with a value range of 1-2, and |Ω| is the total number of valid coordinate points contained in the detection window area Ω;

[0075] The above steps measure the angular distribution of ink particle migration relative to the theoretical jetting direction during the inkjet process, thereby determining whether the migration behavior exhibits systematic deviation or random diffusion. Directional dispersion can be used to identify the consistency of particle migration directions, assisting in determining whether the ink is drifting and depositing in non-target areas under thermal disturbances, thereby improving the spatial resolution and accuracy of diffusion risk assessment.

[0076] Finally, the thermal response offset rate and the directional offset discreteness are combined to generate the ink particle migration sensitivity reference value, which is defined as follows:

[0077]

[0078] Where PMSR is the ink particle migration sensitivity reference value.

[0079] This reference value is used to comprehensively evaluate the spatial drift trend and directional fluctuation of ink particles under thermal disturbance conditions, and can accurately quantify the risk level of ink mist diffusion under the current operating state.

[0080] A larger ink particle migration sensitivity reference value, generated by comprehensively analyzing the spatial migration trends of ink particles under the influence of the thermal flow field within the detection window, generally indicates a greater tendency for ink particles to deviate from the target injection path. This indicates that ink droplets are significantly affected by disturbances such as thermal convection, buoyancy, or local turbulence in the air, making them prone to drift and deposition in non-target areas, thus forming micro-mist diffusion. Therefore, a higher reference value indicates that the system is at risk of ink micro-mist diffusion or that its trend is increasing. Conversely, a low ink particle migration sensitivity reference value indicates that the ink droplet migration trajectory is stable, thermal field disturbances have little impact on their movement path, ink particles can be accurately deposited in the target area, and the printing process does not exhibit micro-mist diffusion characteristics.

[0081] Abnormal fluctuations in the initial kinetic energy of ink droplets driven by an inkjet pulse over a short period of time can indicate that ink particles are drifting to non-target areas and depositing, resulting in a state of micro-mist diffusion. This is because the initial kinetic energy of an ink droplet is determined by the voltage, current, and pulse width of the inkjet pulse. Drastic fluctuations in these driving parameters within a short period can directly lead to instability in the droplet's velocity and size, disrupting the droplet's inertial path through the air. On the one hand, ink droplets with insufficient kinetic energy are susceptible to minor environmental disturbances (such as local airflow and temperature gradients), causing them to deviate from their intended landing point. On the other hand, ink droplets with excessive kinetic energy may undergo secondary atomization before contacting the substrate, forming a large number of suspended particles. These non-targeted ink particles can deposit at image edges, background, or blank areas, resulting in low-contrast fuzzy bands or grayscale disturbances, a typical phenomenon of micro-mist diffusion. Therefore, fluctuations in the initial kinetic energy can serve as a precursor to non-target ink diffusion and an important technical indicator for evaluating inkjet stability and atomization trends.

[0082] The specific steps for comprehensively analyzing the fluctuation of the initial kinetic energy of the ink droplet driven by the inkjet pulse in the detection window to generate the reference value of the pulse kinetic energy fluctuation are as follows:

[0083] By performing multi-dimensional fusion processing on the driving parameters collected by the printing equipment during the inkjet operation, a response factor that characterizes the initial kinetic energy state of the ink droplet is formed based on the inkjet head driving voltage, pulse width, jetting frequency, and nozzle flow resistance disturbance factor. The formula for the inkjet pulse kinetic energy response factor is as follows:

[0084]

[0085] Where V i is the inkjet drive voltage, which represents the instantaneous voltage applied by the inkjet drive control unit to the inkjet head during the i-th inkjet pulse, τ i is the pulse width, which indicates the length of time the inkjet pulse signal remains at a high level in the i-th inkjet pulse, ζ i is the nozzle flow resistance disturbance factor, which indicates the degree of change in fluid resistance in the nozzle channel caused by factors such as contamination, ink viscosity change, and bubble retention during the i-th inkjet pulse. i is the injection frequency, which indicates the number of injections performed by the nozzle per unit time (such as 1 second). Tanh is the hyperbolic tangent function, a typical nonlinear compression function. Its value range is (-1,1) and has a linear response characteristic for small input values, while it has a saturation suppression effect for large input values. i is the inkjet pulse kinetic energy response factor, which represents the actual kinetic energy response level obtained by the ink droplet under the i-th inkjet pulse;

[0086] The main purpose of introducing the hyperbolic tangent function tanh in this formula is to perform nonlinear compression and numerical balance processing on the inkjet pulse kinetic energy input, so as to improve the adaptability and robustness of the model to actual working condition fluctuations. Specifically, since the inkjet driving voltage V i , pulse width τ i and flow resistance factor ζ i The combination of may produce excessively large kinetic energy calculation values in some extreme cases. If the original multiplication and division relationship is directly used as the kinetic energy response output, it is easy to cause model overfitting, feature dominance imbalance, or the algorithm is highly sensitive to individual outliers. The tanh function has the characteristics of "intermediate linearity and boundary suppression", that is, when the input value is small, it responds approximately linearly and retains subtle change characteristics; when the input value is too large, its output approaches ±1, effectively suppressing the impact of the surge value on the overall evaluation result and preventing the model from being skewed. Therefore, the use of the hyperbolic tangent function can ensure that the system responds sensitively to slight fluctuations while constructing a kinetic energy characterization mechanism with the ability to limit high-amplitude disturbances, thereby improving the stability and reliability of ink diffusion trend identification.

[0087] The above steps integrate the key driving parameters in the inkjet pulse process to construct a response factor Λ that can quantify the actual kinetic energy state of the ink droplet. iThis factor provides a core characteristic basis with physical significance and model versatility for the subsequent identification of inkjet stability fluctuations and ink mist diffusion trends.

[0088] The kinetic energy response factors of multiple inkjet pulses during the continuous inkjet cycle are obtained. i Afterwards, a dynamic analysis of the sequence change trend within the detection window is performed to quantify the continuous stability of the kinetic energy output process. The following trend function is used to define the pulse kinetic energy fluctuation reference value. The formula is as follows:

[0089]

[0090] Where, Λ i-1 It is the inkjet pulse kinetic energy response factor under the previous inkjet pulse, n is the total number of inkjet pulses, ω is the trend enhancement coefficient, which is used to enhance the impact of small kinetic energy jumps on the fluctuation index and give the change trend a higher sensitivity. The value range is 3-5. PKFR is the reference value of pulse kinetic energy fluctuation.

[0091] The core function of the hyperbolic tangent function tanh in the above steps is to perform nonlinear compression and smoothing on the amplitude of the inkjet pulse kinetic energy response, thereby enhancing the model's sensitivity to medium-amplitude jumps and suppressing the disturbance of abnormal extreme values on the overall evaluation results. In the actual inkjet process, control parameters such as pulse width, voltage or injection frequency may occasionally fluctuate sharply. If they are directly used for the calculation of the kinetic energy fluctuation index, it is very easy to amplify the overall fluctuation judgment due to a single abnormal jump, resulting in false triggering of the control logic. The tanh function has a natural "saturation zone" characteristic - it amplifies smaller changes approximately linearly to maintain accuracy; it automatically compresses drastic changes that exceed the threshold to the range limit (approaching ±1) to avoid abnormal values dominating the judgment result. Therefore, the use of tanh can achieve a balance between sensitive detection and stable output, effectively improving the accuracy of the fluctuation reference value in identifying the ink mist diffusion trend and the robustness of the system, and is particularly suitable for application in high-speed and frequently disturbed inkjet scenarios.

[0092] The above steps analyze the fluctuation range of the kinetic energy response factor during the continuous inkjet cycle to accurately identify whether there is a short-term unstable transition during the inkjet process, thereby determining whether the ink droplet ejection state is abnormally fluctuating. The pulse kinetic energy fluctuation reference value generated in this step can be used as an important quantitative indicator to measure the risk of ink micro-mist diffusion, providing an intelligent judgment basis for the system to promptly start the adjustment of inkjet control parameters.

[0093] A larger pulse kinetic energy fluctuation reference value, generated by comprehensively analyzing the fluctuations in the initial kinetic energy of ink droplets driven by the inkjet pulse within the detection window, indicates that the initial kinetic energy of the ink droplets fluctuates violently within a short period of time. This indicates significant instability in the droplet size, velocity, and ejection direction. This instability increases the probability of ink droplets drifting in the air and makes them susceptible to environmental disturbances or thermal airflow, causing the droplets to deviate from the target path and form deposits in non-target areas, manifesting as a micro-mist diffusion phenomenon. Conversely, a lower pulse kinetic energy fluctuation reference value indicates good inkjet kinetic energy continuity, stable droplet trajectories, less prone to diffusion drift, precise ink deposition, and clear image boundaries.

[0094] The key features that have undergone comprehensive analysis and quantification are constructed into feature vectors and input into a pre-trained machine learning model. The machine learning model then quickly and accurately identifies and analyzes the feature vectors, intelligently assessing whether there is ink mist diffusion in the current printing process.

[0095] The ink particle migration sensitivity reference value and pulse kinetic energy fluctuation reference value that have been comprehensively analyzed and quantified are constructed into feature vectors and input into a pre-trained machine learning model. The ink micro-mist diffusion risk coefficient is generated by the model, and the ink micro-mist diffusion risk coefficient is used to intelligently evaluate whether there is ink micro-mist diffusion in the current printing process.

[0096] A "pre-trained machine learning model" refers to an intelligent model with prediction and discrimination capabilities obtained through offline training based on a large amount of historical process data and its corresponding ink mist diffusion state before the system is put into actual printing operation. During the training stage, the system first collects data such as inkjet control parameters, thermal flow field characteristics, environmental variables and visual recognition results from multiple printing batches, and a manual or automatic labeling system classifies each set of data as to whether it is accompanied by ink mist diffusion. Then, feature learning is performed on the labeled data through machine learning algorithms (such as random forests, support vector machines, decision trees, lightweight neural networks, etc.) to find the deep laws between "ink mist diffusion" and changes in related parameters, and train generalizable model parameters. When the model shows high accuracy, recall rate and stability on the validation set, it can be solidified into a deployable "pre-trained model" and embedded in the printing quality management system.

[0097] During actual printing operations, the system uses the real-time calculated ink particle migration sensitivity reference value (PMSR) and pulse kinetic energy fluctuation reference value (PKFR) as input features to construct a feature vector that is then fed into the pre-trained model. Without requiring relearning, the model can quickly determine whether ink micromist diffusion is currently occurring based on the existing training weights and further generate a continuous quantitative indicator—the ink micromist diffusion risk coefficient (IMDRC). This coefficient can be used as a basis for process control. For example, when the IMDRC exceeds a set threshold, the system can intelligently determine that diffusion risk is occurring and automatically trigger inkjet pulse width adjustment or process compensation mechanisms. Compared to traditional rule-based or experience-based judgment methods, this approach offers greater robustness, adaptability, and real-time responsiveness, significantly improving the accuracy of identifying micromist diffusion issues and the efficiency of handling them. As the core of data-driven intelligent decision-making, pre-trained models hold irreplaceable and critical value in modern intelligent printing systems.

[0098] The machine learning model is not limited here. Any machine learning model that can perform a comprehensive analysis of the ink particle migration sensitivity reference value PMSR and the pulse kinetic energy fluctuation reference value PKFR to generate the ink micro mist diffusion risk coefficient IMDRC can be used. To implement the technical solution of the present invention, the present invention provides a specific implementation method:

[0099] The formula for generating the ink micro-mist diffusion risk coefficient IMDRC is as follows: IMDRC = h1·PMSR+h2·PKFR, where h1 and h2 are the preset proportional coefficients of the ink particle migration sensitivity reference value PMSR and the pulse kinetic energy fluctuation reference value PKFR, respectively, and both h1 and h2 are greater than 0.

[0100] "Preset proportionality coefficients" refer to two weighting factors, h1 and h2, that are manually or empirically set in advance to linearly weight the two key reference values—the PMSR and PKFR—in generating the IMDRC (Ink Micromist Diffusion Risk Factor). These factors are used to adjust the influence or relative importance of each reference value in the final risk factor calculation.

[0101] Specifically, a larger h1 indicates that the system prioritizes the impact of particle migration on diffusion risk; conversely, a larger h2 indicates that the system prioritizes the effects of inkjet kinetic energy fluctuations on diffusion behavior. The preset proportionality factor, determined based on historical data fitting, expert experience, or optimization algorithms, is a key parameter in the model structure, used to balance the risk contributions of different physical mechanisms. This ensures that the generated IMDRC is more scientific, controllable, and adaptable to actual operating conditions.

[0102] It can be seen from the ink micro-mist diffusion risk coefficient that the larger the ink particle migration sensitivity reference value generated after comprehensive analysis of the spatial migration trend of particles in the ink under the action of the thermal flow field in the detection window, and the larger the pulse kinetic energy fluctuation reference value generated after comprehensive analysis of the fluctuation of the initial kinetic energy of the ink droplet driven by the inkjet pulse in the detection window, the larger the generated ink micro-mist diffusion risk coefficient is, indicating that the probability of ink micro-mist diffusion phenomenon in the current printing process is greater, and vice versa, it indicates that the probability of ink micro-mist diffusion phenomenon in the current printing process is smaller.

[0103] Compare and analyze the ink mist diffusion risk coefficient with the pre-set ink mist diffusion risk coefficient reference threshold to assess whether there is ink mist diffusion in the current printing process. The specific steps are as follows:

[0104] If the ink micro-mist diffusion risk coefficient is greater than the pre-set ink micro-mist diffusion risk coefficient reference threshold, it indicates that ink micro-mist diffusion occurs in the current printing process; if the ink micro-mist diffusion risk coefficient is less than or equal to the pre-set ink micro-mist diffusion risk coefficient reference threshold, it indicates that ink micro-mist diffusion does not occur in the current printing process.

[0105] When the machine learning model evaluation results indicate that ink mist diffusion is occurring during the current printing process, the adaptive pulse width modulation (PWM) method dynamically shortens the inkjet pulse width to reduce the initial kinetic energy of each ink droplet generated within the jetting unit. This rapidly controls the droplet size and velocity, physically suppressing the aerial diffusion of ink particles and reducing contamination and grayscale disturbances in non-target areas.

[0106] The purpose of the above steps is to perform active and precise dynamic control based on the physical inkjet mechanism after identifying the risk of ink micro-mist diffusion, so as to suppress the abnormal diffusion behavior of ink droplets from the source and prevent them from deviating from the original trajectory in the air and depositing in non-target areas. During the inkjet process, the initial kinetic energy of the ink droplets is closely related to their particle size and speed, and these parameters are directly controlled by the width of the inkjet drive pulse. When the pulse width is large, the ejected ink droplets gain higher kinetic energy and are more likely to form wakes in the air or drift due to airflow interference, causing the ink particles to form low-contrast artifacts around highly sensitive areas such as QR codes and barcodes, thereby interfering with subsequent image recognition and information analysis. By introducing an adaptive pulse width modulation (PWM) method, the system can slightly shorten the pulse width in real time when micro-mist diffusion is detected, reducing the ejection kinetic energy and flight distance of the ink droplets, allowing the ink droplets to adhere to the target surface more quickly, reducing their drift opportunities and diffusion radius in the air, and thus effectively curbing non-target deposition. At the same time, this control action can take effect quickly while maintaining printing quality, without interrupting production or replacing equipment. It is immediate, accurate and highly industrially adaptable. It is one of the key links in achieving closed-loop control of atomization diffusion, and provides technical support for ensuring the clarity of the printing area and the accuracy of functional pattern recognition.

[0107] When the machine learning model evaluation results indicate that ink mist diffusion is occurring during the current printing process, the adaptive pulse width modulation (PWM) method is used to dynamically shorten the inkjet pulse width to reduce the initial kinetic energy of each ink droplet generated within the ejection unit. The specific steps for quickly controlling the ink droplet size and velocity are as follows:

[0108] When the machine learning model evaluation results indicate that ink mist diffusion occurs during the current printing process, the inkjet pulse width is slightly shortened and adjusted based on pulse width modulation (PWM) technology to reduce the initial kinetic energy of the ink droplets and suppress the tendency of ink particles to diffuse in the air. The inkjet pulse width adjustment formula is as follows:

[0109]

[0110] Where, IMDRC is the ink mist diffusion risk coefficient, IMDRC ref is the reference threshold of the ink mist diffusion risk coefficient, α is the adaptive adjustment sensitivity coefficient, which is used to control the proportional factor of the inkjet pulse width shortening amplitude. The value range is 0<α<1, which controls the "reaction speed" of the adjustment amplitude, that is, when the risk index increases, whether the system can quickly and significantly reduce the pulse width. PW std It is the standard inkjet pulse width, the reference pulse width set under normal working conditions, representing the inkjet control parameters under ideal conditions. newThe adjusted inkjet pulse width is the new inkjet pulse width that the inkjet control system dynamically calculates based on the risk level after detecting an increased risk of ink mist diffusion.

[0111] This adjustment process does not require manual intervention and can respond quickly at the moment when the diffusion risk increases, achieving flexible suppression of inkjet energy.

[0112] After dynamic pulse width adjustment, the real-time feedback evaluation of the ink droplet initial kinetic energy control effect is conducted to evaluate the change in the ink droplet initial kinetic energy before and after the adjustment to verify the diffusion suppression effect brought about by the shortened pulse width. The formula for calculating the initial kinetic energy change ratio is as follows:

[0113]

[0114] Where, ΔE inkdrop is the ratio of the change in the initial kinetic energy of the ink droplet, which indicates the ratio of the change in the initial kinetic energy of the ink droplet compared to the kinetic energy under the standard pulse width after the dynamic adjustment of the pulse width is implemented. The larger the value, the greater the degree of decrease in the initial kinetic energy of the ink droplet caused by the pulse width adjustment, indicating that the current control measures have a more significant inhibitory effect on ink diffusion. The value range is [0,1). β is the nonlinear response coefficient, which reflects the nonlinear degree of the influence of the pulse width on the initial kinetic energy of the ink droplet. It is the amplification or contraction factor between the pulse width change and the kinetic energy change, and the value range is 1.5-3.

[0115] This method, based on the real-time collection and standardized processing of process data, combines feature engineering with machine learning models to accurately identify ink particle migration trends and inkjet kinetic energy fluctuations. This allows for rapid identification of abnormal ink diffusion conditions at the initial stage. Adaptive pulse width modulation is then used to timely adjust inkjet parameters, suppressing ink droplet deposition in non-target areas and significantly reducing interference and misjudgment risks in variable image areas such as QR codes and barcodes. This solution, which does not rely on additional hardware modifications, offers high real-time performance, strong compatibility, and excellent process stability, significantly improving the overall quality consistency and production yield of printed products, meeting the stringent requirements for image readability and finished product quality in high-precision, high-reliability printing scenarios.

[0116] The present invention provides Figure 2 The full-process management system for product printing shown includes a printing process data acquisition module, a data preprocessing and standardization module, a feature extraction and diffusion quantification module, an ink diffusion intelligent evaluation module, and an adaptive inkjet control and adjustment module.

[0117] The printing process data acquisition module uses the process data acquisition mechanism to continuously collect real-time control parameter data of multiple dimensions generated during the operation of the inkjet printing equipment during the printing process, realizing the full process perception and dynamic monitoring of the printing status;

[0118] The data preprocessing and standardization module preprocesses the acquired raw real-time control parameter data and establishes a structured standard data set to improve data accuracy, consistency, and analysis effectiveness;

[0119] The feature extraction and diffusion quantification module uses feature engineering methods to accurately extract key features that reflect the diffusion state of ink micro-mist from the data set, and conducts comprehensive analysis on the extracted key features to quantify the diffusion degree of ink micro-mist;

[0120] The intelligent ink diffusion assessment module constructs feature vectors from key features that have been comprehensively analyzed and quantified, and inputs them into a pre-trained machine learning model. The machine learning model then quickly and accurately identifies and analyzes the feature vectors, intelligently assessing whether there is ink mist diffusion during the current printing process.

[0121] The adaptive inkjet control and adjustment module, when the machine learning model evaluation results indicate that ink micro-mist diffusion occurs in the current printing process, dynamically shortens the width of the inkjet pulse through the adaptive pulse width modulation (PWM) method to reduce the initial kinetic energy of ink droplets generated in each injection unit, quickly control the ink droplet particle size and speed, and quickly suppress the aerial diffusion of ink particles from a physical level, reducing pollution and grayscale disturbances in non-target areas.

[0122] An embodiment of the present invention provides a full-process management method for product printing, which is implemented through the above-mentioned full-process management system for product printing. The specific methods and processes of a full-process management system for product printing are detailed in the embodiment of the above-mentioned full-process management method for product printing, and will not be repeated here.

[0123] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0124] The above description is merely illustrative of certain exemplary embodiments of the present invention. It goes without saying that those skilled in the art will be able to modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims.

[0125] It should be noted that, in this document, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device that includes the element.

[0126] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0127] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0128] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0129] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0130] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0131] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0132] The above description is merely illustrative of certain exemplary embodiments of the present invention. It goes without saying that those skilled in the art will be able to modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims.

Claims

1. A full-process management method for product printing, characterized in that: The following steps are involved: During the printing process, the process data collection mechanism is used to continuously collect real-time control parameter data of multiple dimensions generated during the operation of the inkjet printing equipment, realizing full-process perception and dynamic monitoring of the printing status; Pre-process the acquired raw real-time control parameter data and establish a structured standard data set to improve data accuracy, consistency and analysis effectiveness; Through feature engineering methods, key features reflecting the ink mist diffusion state are accurately extracted from the data set, and the extracted key features are comprehensively analyzed to quantify the ink mist diffusion degree; The key features that have undergone comprehensive analysis and quantification are constructed into feature vectors and input into a pre-trained machine learning model. The machine learning model then quickly and accurately identifies and analyzes the feature vectors, intelligently assessing whether there is ink mist diffusion in the current printing process. When the evaluation results of the machine learning model indicate that ink mist diffusion occurs in the current printing process, the width of the inkjet pulse is dynamically shortened through the adaptive pulse width modulation method to reduce the initial kinetic energy of ink droplets generated in each injection unit, quickly control the ink droplet particle size and speed, and quickly suppress the aerial diffusion of ink particles from a physical level, reducing pollution and grayscale disturbances in non-target areas.

2. The full-process management method for product printing according to claim 1, characterized in that: During the printing process, the process data collection mechanism is used to continuously collect real-time control parameter data of multiple dimensions generated during the operation of the inkjet printing equipment. The specific steps are as follows: First, define the acquisition parameter range and clarify the key control parameter dimensions that need to be monitored; Secondly, deploy collection interfaces and perception nodes to achieve seamless access to various parameter sources based on existing equipment control system interfaces or embedded sensor modules; Then, the data sampling frequency and synchronization mechanism are set to ensure that the parameters of each dimension are continuously collected with a millisecond sampling period under a unified time base, thus constructing a high-precision, multi-channel time series data stream; Finally, a data cache and upload channel is established to transmit the real-time collected data to the edge controller or central processing platform, forming a continuous and stable data input stream, providing a data basis for subsequent process status analysis, intelligent identification and dynamic control.

3. The full-process management method for product printing according to claim 1, characterized in that: Through feature engineering methods, key features reflecting the diffusion state of ink micro-mist are accurately extracted from the data set. The extracted features include the spatial migration trend of particles in the ink under the action of the thermal flow field and the fluctuation of the initial kinetic energy of the ink droplets driven by the inkjet pulse. The spatial migration trend of particles in the ink under the action of the thermal flow field and the fluctuation of the initial kinetic energy of the ink droplets driven by the inkjet pulse are comprehensively analyzed under the detection window to generate ink particle migration sensitive reference values and pulse kinetic energy fluctuation reference values, respectively. The ink micro-mist diffusion degree is quantified by the ink particle migration sensitive reference values and the pulse kinetic energy fluctuation reference values.

4. The full-process management method for product printing according to claim 3, characterized in that: The specific steps for comprehensively analyzing the spatial migration trend of ink particles under the action of the thermal flow field within the detection window to generate the ink particle migration sensitivity reference value are as follows: Based on the actual deposition trajectory of ink droplets during inkjet imaging, a droplet spatial offset vector field is constructed and superimposed with the corresponding thermal disturbance vector field to characterize the disturbance effect of the thermal flow field on the droplet trajectory. The thermal response offset rate is obtained by calculating the weighted projection relationship between the droplet offset vector and the thermal disturbance direction. The calculation formula is as follows: Where, is the thermal response offset rate, is the ink droplet spatial offset vector field, which represents the spatial offset vector of the ink droplet at the coordinate point (x, y). is the local thermal disturbance vector field, which represents the direction and intensity of the thermal flow disturbance at the coordinate point (x, y), Ω is the detection window range, and λ is the trend enhancement factor, which is used to apply nonlinear weight adjustment to the dot product result and the thermal disturbance modulus. is the modulus of the thermal disturbance vector; On the basis of obtaining the thermal response offset rate, the inkjet direction offset dispersion is further introduced to reflect the overall discrete distribution of the ink drop offset angle relative to the theoretical inkjet direction. The calculation formula of the inkjet direction offset dispersion is as follows: Where, δ θ is the directional discreteness, which is used to quantify the overall discreteness of the inkjet offset direction within the window. θ(x, y) is the offset direction angle, which indicates the spatial offset vector of the ink droplet at the coordinate point (x, y). The angle formed with the ideal inkjet direction, θ ref is the reference inkjet direction, η is the offset enhancement factor, |Ω| is the total number of valid coordinate points contained in the detection window area Ω; Finally, the thermal response offset rate and the directional offset discreteness are combined to generate the ink particle migration sensitivity reference value, which is defined as follows: Where PMSR is the ink particle migration sensitivity reference value.

5. The whole-process management method for product printing according to claim 3, characterized in that: The specific steps for comprehensively analyzing the fluctuation of the initial kinetic energy of the ink droplet driven by the inkjet pulse in the detection window to generate the reference value of the pulse kinetic energy fluctuation are as follows: By performing multi-dimensional fusion processing on the driving parameters collected by the printing equipment during the inkjet operation, a response factor that characterizes the initial kinetic energy state of the ink droplet is formed based on the inkjet head driving voltage, pulse width, jetting frequency, and nozzle flow resistance disturbance factor. The formula for the inkjet pulse kinetic energy response factor is as follows: Where V i is the inkjet drive voltage, which represents the instantaneous voltage applied by the inkjet drive control unit to the inkjet head during the i-th inkjet pulse, τ i is the pulse width, which indicates the length of time the inkjet pulse signal remains at a high level in the i-th inkjet pulse. is the nozzle flow resistance disturbance factor, f i is the injection frequency, tanh is the hyperbolic tangent function, Λ i is the inkjet pulse kinetic energy response factor, which represents the actual kinetic energy response level obtained by the ink droplet under the i-th inkjet pulse; The kinetic energy response factors of multiple inkjet pulses during the continuous inkjet cycle are obtained. i Afterwards, a dynamic analysis of the sequence change trend within the detection window is performed to quantify the continuous stability of the kinetic energy output process. The following trend function is used to define the pulse kinetic energy fluctuation reference value. The formula is as follows: Where, Λ i-1 is the inkjet pulse kinetic energy response factor under the previous inkjet pulse, n is the total number of inkjet pulses, ω is the trend enhancement coefficient, and PKFR is the pulse kinetic energy fluctuation reference value.

6. The full-process management method for product printing according to claim 3, characterized in that: The ink particle migration sensitivity reference value and pulse kinetic energy fluctuation reference value that have been comprehensively analyzed and quantified are constructed into feature vectors and input into a pre-trained machine learning model. The ink micro-mist diffusion risk coefficient is generated by the model, and the ink micro-mist diffusion risk coefficient is used to intelligently evaluate whether there is ink micro-mist diffusion in the current printing process.

7. A full-process management method for product printing according to claim 6, characterized in that: Compare and analyze the ink mist diffusion risk coefficient with the pre-set ink mist diffusion risk coefficient reference threshold to assess whether there is ink mist diffusion in the current printing process. The specific steps are as follows: If the ink mist diffusion risk coefficient is greater than the preset ink mist diffusion risk coefficient reference threshold, it indicates that ink mist diffusion occurs in the current printing process; If the ink micro-mist diffusion risk coefficient is less than or equal to a preset ink micro-mist diffusion risk coefficient reference threshold, it indicates that no ink micro-mist diffusion phenomenon occurs in the current printing process.

8. The full-process management method for product printing according to claim 7, characterized in that: When the machine learning model evaluation results indicate that ink mist diffusion is occurring during the current printing process, the adaptive pulse width modulation method is used to dynamically shorten the inkjet pulse width to reduce the initial kinetic energy of each ink droplet generated within the jetting unit. The specific steps for quickly controlling the ink droplet size and velocity are as follows: When the machine learning model evaluation results indicate that ink mist diffusion occurs during the current printing process, the inkjet pulse width is slightly shortened based on pulse width modulation technology to reduce the initial kinetic energy of the ink droplets and suppress the airborne diffusion trend of the ink particles. The inkjet pulse width adjustment formula is as follows: Where, IMDRC is the ink mist diffusion risk coefficient, IMDRC ref is the reference threshold of the ink mist diffusion risk coefficient, α is the adaptive adjustment sensitivity coefficient, which is used to control the proportional factor of the inkjet pulse width shortening amplitude, and its value range is 0<α<1. std is the standard inkjet pulse width, PW new is the adjusted inkjet pulse width; After dynamic pulse width adjustment, the real-time feedback evaluation of the ink droplet initial kinetic energy control effect is conducted to evaluate the change in the ink droplet initial kinetic energy before and after the adjustment to verify the diffusion suppression effect brought about by the shortened pulse width. The formula for calculating the initial kinetic energy change ratio is as follows: Where, ΔE inkdrop is the change ratio of the initial kinetic energy of the ink droplet, and β is the nonlinear response coefficient.

9. A full-process management system for product printing, used to implement the full-process management method for product printing described in any one of claims 1 to 8, characterized in that: It includes a printing process data acquisition module, a data preprocessing and standardization module, a feature extraction and diffusion quantification module, an ink diffusion intelligent evaluation module, and an adaptive inkjet control and adjustment module. The printing process data acquisition module uses the process data acquisition mechanism to continuously collect real-time control parameter data of multiple dimensions generated during the operation of the inkjet printing equipment during the printing process, realizing the full process perception and dynamic monitoring of the printing status; The data preprocessing and standardization module preprocesses the acquired raw real-time control parameter data and establishes a structured standard data set to improve data accuracy, consistency, and analysis effectiveness; The feature extraction and diffusion quantification module uses feature engineering methods to accurately extract key features that reflect the diffusion state of ink micro-mist from the data set, and conducts comprehensive analysis on the extracted key features to quantify the diffusion degree of ink micro-mist; The intelligent ink diffusion assessment module constructs feature vectors from key features that have been comprehensively analyzed and quantified, and inputs them into a pre-trained machine learning model. The machine learning model then quickly and accurately identifies and analyzes the feature vectors, intelligently assessing whether there is ink mist diffusion during the current printing process. The adaptive inkjet control and adjustment module, when the machine learning model evaluation results indicate that ink micro-mist diffusion occurs in the current printing process, dynamically shortens the width of the inkjet pulse through the adaptive pulse width modulation (PWM) method to reduce the initial kinetic energy of ink droplets generated in each injection unit, quickly control the ink droplet particle size and speed, and quickly suppress the aerial diffusion of ink particles from a physical level, reducing pollution and grayscale disturbances in non-target areas.

Images