A plate roller printing online design optimization method and system

Abstract

The present application relates to the technical field of printing equipment control, and specifically discloses a method and system for online design optimization of printing based on a plate roller, which comprises: arranging a multi-type sensor array on a printing machine cylinder assembly to collect multi-dimensional data such as dynamic pressure, pressure distribution, cylinder radial displacement, and nip environment temperature in real time; performing abnormal rejection, noise reduction, and standardization processing on the collected data; calculating a pressure distribution uniformity index and comparing it with a preset threshold value, and locating a target area with the largest pressure deviation when the requirement is not met; using a closed-loop control strategy, combining cylinder speed and printing material thickness to dynamically correct the adjustment amount, driving an actuator to adjust the pressure until the deviation rate is below the allowable threshold value. Through multi-sensor fusion and closed-loop adjustment, the present application realizes online monitoring and intelligent optimization of printing pressure, significantly improves the pressure distribution uniformity and printing quality consistency, and is suitable for high-speed gravure printing, flexographic printing, and other scenarios.

Description

Technical Field

[0001] This invention relates to the field of printing equipment control technology, specifically to an online design optimization method and system for printing rollers. Background Technology

[0002] In roll-to-roll printing processes such as gravure and flexographic printing, the uniformity of the contact pressure distribution between the printing roller and the pressure roller directly affects the ink transfer efficiency and the quality of the printed product. Excessive pressure can lead to dot gain, image distortion, and even damage to the printing substrate, while insufficient pressure can cause insufficient ink transfer, pale colors, or broken lines. Uneven pressure distribution can also cause defects such as localized color differences and misregistration in the printed product.

[0003] Currently, traditional printing equipment mainly uses mechanical pressure regulation, relying on the operator's experience to manually adjust the contact pressure between the printing roller and the pressure roller by adjusting the pressure of the clamping bolts or cylinders on both sides. This method has the following prominent problems: First, the adjustment depends on manual experience, which is highly subjective and lacks quantitative basis, making it difficult to ensure the accuracy and consistency of pressure distribution; second, it cannot monitor pressure changes at different positions along the roller axis in real time. When factors such as printing speed, substrate thickness, and ambient temperature fluctuate, the pressure distribution changes accordingly, and manual adjustment cannot respond in time; third, it lacks comprehensive consideration of multiple factors such as roller radial displacement and temperature interference, and single pressure detection is insufficient to reflect the true contact state.

[0004] Therefore, this invention proposes an online design optimization method and system based on printing rollers. Summary of the Invention

[0005] The purpose of this invention is to provide an online design optimization method and system for roller printing, thereby solving the above-mentioned technical problems: The objective of this invention can be achieved through the following technical solutions: A method for online design optimization based on printing rollers, the method comprising: S1. Sensor layout and calibration: Multiple types of sensor arrays are arranged on the printing press roller assembly, including piezoelectric pressure sensors, displacement sensors and temperature sensors. S2. Multi-source pressure data acquisition: During the printing process, multi-dimensional data is acquired in real time through a sensor array, including: dynamic pressure data of different positions of the roller acquired by the piezoelectric pressure sensor, radial displacement data of the roller acquired by the displacement sensor, and ambient temperature data of the pressure zone acquired by the temperature sensor. The sampling frequency is set to 100-500Hz to ensure the real-time nature of data acquisition, and the sampling timestamp is recorded to achieve time synchronization of multi-source data. S3. Multi-sensor data fusion processing: Processing the collected multi-source data; S4. Pressure Distribution Uniformity Assessment: Set a threshold for printing pressure distribution uniformity, analyze the data processed in S3, and assess whether the current pressure distribution meets the printing quality requirements based on the analysis results. S5. Pressure distribution optimization and adjustment: Based on the evaluation results of step S4, a closed-loop control strategy is adopted to dynamically adjust the printing pressure.

[0006] As a further description of the technical solution of the present invention, S1 includes uniformly calibrating all sensors to determine the measurement range, accuracy, and error compensation parameters of the sensors, ensuring the consistency of the detection data; wherein, piezoelectric pressure sensors are arranged at both ends and the middle area of ​​the roller to detect dynamic pressure values, with a measurement range covering 3-999 N / cm. 2 It is adapted to the needs of different printing scenarios; the pressure-sensitive thin-film sensor is laid in the contact pressure area of ​​the roller to detect local pressure peaks and pressure distribution profiles; the displacement sensor is used to detect the radial displacement of the roller, indirectly reflecting pressure changes; the temperature sensor is used to detect the ambient temperature of the pressure area to eliminate the interference of temperature on pressure detection.

[0007] As a further description of the technical solution of the present invention, the multi-source data sampling method in S2 includes the following steps: S21. Spatial Layout of Sampling Points: m sampling sections are evenly spaced along the axial direction of the roller. k sampling points are set at equal angles along the circumference of the roller on each section, forming... Each discrete sampling point ensures coverage of the two ends, the middle, and key locations in the pressure zone of the roller. S22. Sampling frequency and time synchronization: Set the sampling frequency Satisfying 100Hz≤ ≤500Hz, all sensors use the same clock source to trigger sampling and record a unified timestamp. j=1,2,…,N, to achieve time alignment of multi-source data; S23, Parallel acquisition of multi-dimensional data: at each sampling point i and at each sampling time... The following four types of data are collected in parallel: dynamic pressure value, pressure peak value and pressure distribution profile width, roller radial displacement and pressure zone ambient temperature; S24. Data sliding window sampling: A sliding window mechanism is adopted, with a window length L = 0.5-1.0 seconds and a sliding step Δt = 0.1 seconds; The average pressure at each sampling point is calculated within each window, which is used for real-time calculation of the pressure distribution uniformity index. S25. Abnormal sampling data removal and completion: Real-time detection of the validity of data at each sampling point. If data at a certain sampling point is missing or exceeds the sensor's range, spatial interpolation of nearby sampling points or temporal interpolation of the previous moment is used to complete the data, ensuring the integrity of data at n sampling points.

[0008] As a further description of the technical solution of the present invention, the specific process of S3 includes: The collected raw data were subjected to outlier removal, noise reduction, and standardization. The 3σ criterion was used to remove outlier data, and wavelet transform was used to reduce noise and eliminate environmental interference and sensor errors. Data of different dimensions were standardized to map pressure data, displacement data, and temperature data to the [0,1] interval for easy subsequent fusion calculation.

[0009] As a further description of the technical solution of the present invention, the specific process of S4 includes: S41. Based on the multi-source data processed in S3, calculate the pressure distribution uniformity index at the current moment. ; S42. Set the pressure distribution uniformity threshold. , calculate and Comparison: like ≥ The current pressure distribution is determined to meet the printing quality requirements. like If the current pressure distribution is determined to be unsatisfactory for printing quality, pressure optimization adjustment is triggered. S43, when At that time, calculate the pressure deviation rate of each sampling point: ,Will Sort the samples from largest to smallest, locate the r sampling points with the largest pressure deviations, and output them as the target area for pressure adjustment to S5. This represents the average pressure at the i-th sampling point within the current sliding window. This is the arithmetic mean of the pressure values ​​within the current sliding window for all sampling points.

[0010] As a further description of the technical solution of the present invention, S41 includes: Calculate the mathematical average of the actual pressure values ​​of all sampling points within the current sliding window; Calculate the ratio of the absolute value of the pressure deviation at each sampling point to the mathematical average value, average the ratio over all sampling points to obtain the average pressure deviation rate, and subtract the average pressure deviation rate from 1 to obtain the pressure consistency coefficient. Obtain the radial displacement of the roller, calculate the ratio of this displacement to the preset maximum allowable displacement, multiply it by the preset displacement weighting coefficient, and subtract the product from 1 to obtain the displacement influence coefficient. Obtain the ambient temperature of the pressure zone, calculate the ratio of the absolute deviation of this temperature from the standard operating temperature to the standard operating temperature, multiply it by the preset temperature weighting coefficient, and subtract the product from 1 to obtain the temperature influence coefficient. Multiplying the pressure uniformity coefficient, displacement influence coefficient, and temperature influence coefficient yields the pressure distribution uniformity index. The pressure distribution uniformity index is between 0 and 1. The closer the value is to 1, the more uniform the pressure distribution is, and the closer the value is to 0, the less uniform the pressure distribution is.

[0011] As a further description of the technical solution of the present invention, S5 includes: S51. Receive the pressure regulation target region from the non-uniform positioning analysis output of S44. The target region contains the location information of the top r sampling points with the largest pressure deviation rate, where r≥1. S52. For each target adjustment point, based on its pressure deviation rate... and the average pressure within the current sliding window Calculate the required target pressure value and adjustment amount: like > The pressure at this point is determined to be too high, and the pressure needs to be reduced. like < The pressure at this point is determined to be too low, and the pressure needs to be increased. Adjustment amount ,in, The coefficient representing the influence of drum speed. The coefficient representing the influence of the thickness of the printing substrate; S53. The calculated adjustment amount Converted into corresponding actuator control signals; S54. During the adjustment process, pressure change data at each adjustment point are collected in real time. The deviation between the actual pressure value and the target pressure value is used as feedback input, and the adjustment is continuously iterated until the pressure deviation rate at that adjustment point is reached. Below the preset allowable deviation threshold; S55. When there are multiple adjustment points, adjust them in order of pressure deviation rate from large to small, or use parallel adjustment method and prioritize the adjustment point with the largest deviation.

[0012] An online design optimization system for roller printing, the system comprising a data acquisition layer, a processing layer, and an execution layer; The acquisition layer includes: The sensor array module, deployed on the printing press roller assembly, includes a piezoelectric pressure sensor, a displacement sensor, and a temperature sensor; the sensor array module is used to collect dynamic pressure data, radial displacement data, and ambient temperature data of the pressure zone at different positions of the roller in real time. The processing layer includes: A multi-source data acquisition module, connected to the sensor array module, is used to acquire multi-dimensional data in real time at a set sampling frequency and achieve time synchronization; The data fusion processing module is connected to the multi-source data acquisition module and is used to perform outlier removal, noise reduction and standardization processing on the acquired multi-source data. The pressure distribution uniformity assessment module is connected to the data fusion processing module. It is used to set the pressure distribution uniformity threshold, analyze and evaluate the processed data, predict future pressure distribution trends, and identify potential risks in advance. The execution layer includes: The pressure distribution optimization and adjustment module is connected to the pressure distribution uniformity evaluation module and is used to dynamically adjust the printing pressure using a closed-loop control strategy based on the evaluation results and predicted trends. The actuator module, connected to the pressure distribution optimization and adjustment module, is used to receive control signals and perform pressure adjustment actions.

[0013] The beneficial effects of this invention are: First, by deploying multi-type sensor arrays and combining them with a sliding window sampling mechanism, multi-dimensional, high-density, and synchronized real-time monitoring of printing pressure distribution, roller radial displacement, and pressure zone temperature is achieved, significantly improving the completeness and timeliness of data acquisition. Second, a pressure distribution uniformity index integrating pressure consistency coefficient, displacement influence coefficient, and temperature influence coefficient is constructed, enabling a comprehensive and quantitative assessment of the uniformity of pressure distribution and precise location of abnormal areas with the largest pressure deviations, overcoming the limitations of traditional single-point detection or manual experience judgment. Third, a closed-loop control strategy is adopted, incorporating dynamic correction coefficients for roller speed and substrate thickness into the adjustment calculation, allowing the adjustment process to adapt to changes in printing conditions, improving the robustness and accuracy of control. Finally, the entire process achieves an automated closed loop from sensing and evaluation to adjustment, requiring no manual intervention during machine downtime. It can quickly eliminate pressure unevenness, effectively improve the color consistency and image clarity of printed materials, reduce scrap rates, and is suitable for high-speed, high-precision printing production lines. Attached Figure Description

[0014] The invention will now be further described with reference to the accompanying drawings.

[0015] Figure 1 This is a schematic diagram of part of the process of the online design optimization method for printing rollers according to the present invention. Detailed Implementation

[0016] 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.

[0017] Please see Figure 1 As shown, this invention provides an online design optimization method for printing rollers, the method comprising: S1. Sensor layout and calibration: Multiple types of sensor arrays are arranged on the printing press roller assembly, including piezoelectric pressure sensors, displacement sensors and temperature sensors. S1 includes uniformly calibrating all sensors to determine their measurement range, accuracy, and error compensation parameters, ensuring the consistency of the detection data. Specifically, piezoelectric pressure sensors are installed at both ends and in the middle of the roller to detect dynamic pressure values, with a measurement range covering 3-999 N / cm². 2 It is adapted to the needs of different printing scenarios; the pressure-sensitive thin-film sensor is laid in the contact pressure area of ​​the roller to detect local pressure peaks and pressure distribution profiles; the displacement sensor is used to detect the radial displacement of the roller, indirectly reflecting pressure changes; the temperature sensor is used to detect the ambient temperature of the pressure area to eliminate the interference of temperature on pressure detection.

[0018] The above technical solution involves scientifically deploying piezoelectric pressure sensors, displacement sensors, and temperature sensors on the printing press cylinder assembly to form a multi-type sensor array. All sensors undergo unified calibration to determine their measurement range, accuracy, and error compensation parameters, ensuring the consistency and reliability of the detection data. Specifically, piezoelectric pressure sensors are deployed at both ends and the middle area of ​​the cylinder to detect dynamic pressure, displacement sensors indirectly reflect pressure changes, and temperature sensors are used for temperature compensation. This layout and calibration lays the hardware foundation for subsequent high-precision, multi-dimensional, and synchronized pressure data acquisition, ensuring the effectiveness of online monitoring and optimization design of printing pressure.

[0019] S2. Multi-source pressure data acquisition: During the printing process, multi-dimensional data is acquired in real time through a sensor array, including: dynamic pressure data of different positions of the roller acquired by the piezoelectric pressure sensor, radial displacement data of the roller acquired by the displacement sensor, and ambient temperature data of the pressure zone acquired by the temperature sensor. The sampling frequency is set to 100-500Hz to ensure the real-time nature of data acquisition, and the sampling timestamp is recorded to achieve time synchronization of multi-source data. The multi-source data sampling method in S2 includes the following steps: S21. Spatial Layout of Sampling Points: m sampling sections are evenly spaced along the axial direction of the roller. k sampling points are set at equal angles along the circumference of the roller on each section, forming... Each discrete sampling point ensures coverage of the two ends, the middle, and key locations in the pressure zone of the roller. The sampling point density at both ends of the roller is higher than that in the middle area, and the distance between adjacent sampling points at both ends is 0.5 times the distance in the middle area. 0.8 times.

[0020] S22. Sampling frequency and time synchronization: Set the sampling frequency Satisfying 100Hz≤ ≤500Hz, all sensors use the same clock source to trigger sampling and record a unified timestamp. j=1,2,…,N, to achieve time alignment of multi-source data; S23, Parallel acquisition of multi-dimensional data: at each sampling point i and at each sampling time... The following four types of data are collected in parallel: dynamic pressure value, pressure peak value and pressure distribution profile width, roller radial displacement and pressure zone ambient temperature; S24. Data sliding window sampling: A sliding window mechanism is adopted, with a window length L = 0.5-1.0 seconds and a sliding step Δt = 0.1 seconds; The average pressure at each sampling point is calculated within each window, which is used for real-time calculation of the pressure distribution uniformity index. The formula for calculating the average pressure within the sliding window is: ,in, For window length, Sampling frequency, Let be the pressure value at the j-th sampling time of the i-th sampling point. This represents the average pressure at the i-th sampling point within the sliding window.

[0021] S25. Abnormal sampling data removal and completion: Real-time detection of the validity of data at each sampling point. If data at a certain sampling point is missing or exceeds the sensor's range, spatial interpolation of nearby sampling points or temporal interpolation of the previous moment is used to complete the data, ensuring the integrity of data at n sampling points.

[0022] The above technical solution involves several steps. First, multiple sampling sections are evenly spaced along the roller axis. Each section has sampling points set at equal angles along its circumference, with additional sampling points at both ends of the roller to improve data resolution in critical areas. Second, a sampling frequency of 100-500Hz is set, and all sensors use the same clock source to trigger sampling and record a unified timestamp, ensuring time alignment of multi-source data. Then, four types of data are collected in parallel at each sampling point: dynamic pressure value, pressure peak and distribution profile, roller radial displacement, and ambient temperature in the pressure zone. To improve real-time calculation efficiency, a sliding window mechanism is introduced, with a window length of 0.5-1.0 seconds and a sliding step of 0.1 seconds. The average values ​​of pressure, displacement, and temperature are calculated within each window for real-time evaluation of pressure distribution uniformity. Furthermore, the system monitors data validity in real-time, using spatial or temporal interpolation to complete missing or out-of-range data, ensuring data integrity. Through these steps—spatial layout, time synchronization, parallel acquisition, sliding window calculation, and abnormal data completion—S2 achieves stable, reliable, and real-time acquisition of multi-dimensional pressure-related data during the printing process, providing a high-quality data foundation for subsequent data fusion and pressure optimization.

[0023] S3. Multi-sensor data fusion processing: Processing the collected multi-source data; The specific process of S3 includes: The collected raw data were subjected to outlier removal, noise reduction, and standardization. The 3σ criterion was used to remove outlier data, and wavelet transform was used to reduce noise and eliminate environmental interference and sensor errors. Data of different dimensions were standardized to map pressure data, displacement data, and temperature data to the [0,1] interval for easy subsequent fusion calculation.

[0024] S4. Pressure Distribution Uniformity Assessment: Set a threshold for printing pressure distribution uniformity, analyze the data processed in S3, and assess whether the current pressure distribution meets the printing quality requirements based on the analysis results. The specific process of S4 includes: S41. Based on the multi-source data processed in S3, calculate the pressure distribution uniformity index at the current moment. ; S42. Set the pressure distribution uniformity threshold. , calculate and Comparison: like ≥ The current pressure distribution is determined to meet the printing quality requirements. like If the current pressure distribution is determined to be unsatisfactory for printing quality, pressure optimization adjustment is triggered. S43, when At that time, calculate the pressure deviation rate of each sampling point: ,Will Sort the samples from largest to smallest, locate the r sampling points with the largest pressure deviations, and output them as the target area for pressure adjustment to S5. This represents the average pressure at the i-th sampling point within the current sliding window. This is the arithmetic mean of the pressure values ​​within the current sliding window for all sampling points.

[0025] S41 includes: Calculate the mathematical average of the actual pressure values ​​of all sampling points within the current sliding window; Calculate the ratio of the absolute value of the pressure deviation at each sampling point to the mathematical average value, average the ratio over all sampling points to obtain the average pressure deviation rate, and subtract the average pressure deviation rate from 1 to obtain the pressure consistency coefficient. Obtain the radial displacement of the roller, calculate the ratio of this displacement to the preset maximum allowable displacement, multiply it by the preset displacement weighting coefficient, and subtract the product from 1 to obtain the displacement influence coefficient. Obtain the ambient temperature of the pressure zone, calculate the ratio of the absolute deviation of this temperature from the standard operating temperature to the standard operating temperature, multiply it by the preset temperature weighting coefficient, and subtract the product from 1 to obtain the temperature influence coefficient. Multiplying the pressure uniformity coefficient, displacement influence coefficient, and temperature influence coefficient yields the pressure distribution uniformity index. The pressure distribution uniformity index is between 0 and 1. The closer the value is to 1, the more uniform the pressure distribution is, and the closer the value is to 0, the less uniform the pressure distribution is.

[0026] The calculation formula is as follows: ; In the formula, This represents the average pressure at the i-th sampling point within the current sliding window. This is the arithmetic mean of the pressure values ​​within the current sliding window for all sampling points; n is the total number of sampling points, and i belongs to n. This represents the radial displacement of the roller. This is the preset maximum allowable radial displacement threshold; The ambient temperature of the pressure zone within the current sliding window. For standard operating temperature, and These are preset empirical weighting coefficients.

[0027] Through the above technical solution, after the multi-sensor data fusion processing is completed, the uniformity of the current printing pressure distribution is quantitatively evaluated, thresholded, and abnormal areas are located, thus providing a precise basis for pressure optimization and adjustment. First, based on the processed multi-source data, the system calculates the pressure distribution uniformity index within the current sliding window. This indicator is not a single pressure parameter, but a comprehensive evaluation coefficient, obtained by multiplying three sub-coefficients: first, the pressure consistency coefficient, which is the average of the deviation rates between the average pressure of all sampling points and the overall average pressure, and then subtracting this average from 1 to reflect the dispersion of the pressure itself; second, the displacement influence coefficient, which is obtained by obtaining the ratio of the radial displacement of the roller to the preset maximum allowable displacement, and multiplying it by a weighting coefficient. Then, subtract from 1 to reflect the negative impact of drum deformation on pressure uniformity; thirdly, the temperature influence coefficient, calculated by multiplying the relative deviation between the current temperature and the standard working temperature by a weighting factor. Then subtract from 1 to compensate for pressure measurement errors caused by temperature changes. The result is obtained by multiplying the three coefficients. The value ranges from 0 to 1, with values ​​closer to 1 indicating a more uniform pressure distribution and values ​​closer to 0 indicating a less uniform distribution. Secondly, the system sets a preset uniformity threshold. , calculate Compare with this threshold: If ≥ If the current pressure distribution meets the printing quality requirements, no adjustment is needed; if If the requirements are not met, the system will trigger a pressure optimization and adjustment process. Finally, when adjustment is deemed necessary, the system further calculates the pressure deviation rate for each sampling point. ,in This is the arithmetic mean of the pressure values ​​within the current sliding window for all sampling points. (This refers to all sampling points.) Sort the samples from largest to smallest, locate the r sampling points with the largest pressure deviations, and output them as the target area for pressure regulation to S5. This mechanism not only qualitatively determines whether the overall pressure distribution is acceptable, but also quantitatively identifies the specific location and severity of pressure anomalies, achieving a complete analysis from macroscopic assessment to microscopic location, providing clear targets and basis for subsequent closed-loop regulation.

[0028] S5. Pressure distribution optimization and adjustment: Based on the evaluation results of step S4, a closed-loop control strategy is adopted to dynamically adjust the printing pressure.

[0029] S5 includes: S51. Receive the pressure regulation target region from the non-uniform positioning analysis output of S44. The target region contains the location information of the top r sampling points with the largest pressure deviation rate, where r≥1. S52. For each target adjustment point, based on its pressure deviation rate... and the average pressure within the current sliding window Calculate the required target pressure value and adjustment amount: like > The pressure at this point is determined to be too high, and the pressure needs to be reduced. like < The pressure at this point is determined to be too low, and the pressure needs to be increased. Adjustment amount ,in, The coefficient representing the influence of drum speed. The coefficient representing the influence of the thickness of the printing substrate; , , This is the current drum speed. As the reference speed, The weighting coefficient for the influence of rotational speed, Given the current thickness of the printing substrate, As the reference thickness, The weighting coefficient is used to determine the impact of the thickness of the printing material.

[0030] S53. The calculated adjustment amount Converted into corresponding actuator control signals; S54. During the adjustment process, pressure change data at each adjustment point are collected in real time. The deviation between the actual pressure value and the target pressure value is used as feedback input, and the adjustment is continuously iterated until the pressure deviation rate at that adjustment point is reached. Below the preset allowable deviation threshold; S55. When there are multiple adjustment points, adjust them in order of pressure deviation rate from large to small, or use parallel adjustment method and prioritize the adjustment point with the largest deviation.

[0031] Using the above technical solution, firstly, the system receives the pressure regulation target area from S4. This area contains the specific location information of the top r sampling points with the largest pressure deviation rates, where r ≥ 1. For each target regulation point, the system calculates the pressure average within its current sliding window. Overall pressure mean across all sampling points Comparison and judgment: If > If the pressure at that point is too high, it is determined that the pressure needs to be reduced; if < If the pressure is too low, it is determined that the pressure needs to be increased. The system then calculates the required adjustment amount. Two dynamic correction coefficients were introduced: the drum speed influence coefficient. Influence coefficient of printing material thickness , and These are the current speed and the reference speed, respectively. and These are the current thickness and the reference thickness, respectively. and These are the corresponding weighting coefficients. The introduction of these two coefficients allows the calculation of the adjustment amount to adapt to changes in printing speed and batch variations in materials, improving the robustness and adaptability of the adjustment. After calculating the adjustment amount, the system converts it into the corresponding actuator control signal, driving the actuator to adjust the pressure. During the adjustment process, the system collects pressure change data at each adjustment point in real time, using the deviation between the actual pressure value and the target pressure value as feedback input, and continuously iterates the adjustment until the pressure deviation rate at that adjustment point is reached. The pressure deviation is below a preset allowable threshold, thus forming a closed-loop control. When multiple adjustment points exist, the system can adjust them sequentially in descending order of pressure deviation rate, prioritizing the abnormal point with the largest deviation. Alternatively, it can use parallel adjustment to adjust multiple points simultaneously. Through this mechanism, S5 achieves a complete closed loop from anomaly location, adjustment calculation, control signal output to feedback iterative correction, enabling it to quickly and accurately eliminate uneven pressure distribution and significantly improve the stability and consistency of printing quality.

[0032] An online design optimization system for roller printing, the system comprising a data acquisition layer, a processing layer, and an execution layer; The acquisition layer includes: The sensor array module, deployed on the printing press roller assembly, includes a piezoelectric pressure sensor, a displacement sensor, and a temperature sensor; the sensor array module is used to collect dynamic pressure data, radial displacement data, and ambient temperature data of the pressure zone at different positions of the roller in real time. The processing layer includes: A multi-source data acquisition module, connected to the sensor array module, is used to acquire multi-dimensional data in real time at a set sampling frequency and achieve time synchronization; The data fusion processing module is connected to the multi-source data acquisition module and is used to perform outlier removal, noise reduction and standardization processing on the acquired multi-source data. The pressure distribution uniformity assessment module is connected to the data fusion processing module. It is used to set the pressure distribution uniformity threshold, analyze and evaluate the processed data, predict future pressure distribution trends, and identify potential risks in advance. The execution layer includes: The pressure distribution optimization and adjustment module is connected to the pressure distribution uniformity evaluation module and is used to dynamically adjust the printing pressure using a closed-loop control strategy based on the evaluation results and predicted trends. The actuator module, connected to the pressure distribution optimization and adjustment module, is used to receive control signals and perform pressure adjustment actions.

[0033] In summary, the working principle of this invention revolves around a closed-loop control logic of sensing, fusion, evaluation, and adjustment, constructing an online printing pressure optimization system based on multi-sensor data fusion. First, piezoelectric pressure sensors, pressure-sensitive thin-film sensors, displacement sensors, and temperature sensors are deployed on the printing press roller assembly, forming a high-precision, multi-dimensional sensing array after unified calibration. During printing, the system uses a sampling frequency of 100-500Hz, combined with a sliding window mechanism (window length 0.5-1.0 seconds, step size 0.1 seconds), to collect data on roller dynamic pressure, pressure distribution in the pressure zone, radial displacement, and ambient temperature in parallel, ensuring data integrity through spatial and temporal interpolation. Next, the raw data undergoes anomaly removal (3σ criterion), wavelet denoising, and standardization to eliminate dimensional and environmental interference. Based on this, the core evaluation module calculates the pressure distribution uniformity index. ≥ This index integrates the pressure uniformity coefficient, displacement influence coefficient, and temperature influence coefficient, and can quantitatively reflect the uniformity of pressure distribution. With preset threshold If the pressure deviation falls below a threshold, an optimization process is triggered, and the pressure deviation rate at each sampling point is further calculated. The top few locations with the largest deviations are identified as the adjustment target areas. Subsequently, the adjustment module dynamically corrects the adjustment amount based on the direction and magnitude of the pressure deviation at each target point, combined with the current roller speed and substrate thickness, generating control signals to drive the actuator to increase or decrease pressure. During the adjustment process, the system continuously collects feedback data, forming a closed-loop iterative control until the pressure deviation rate at each adjustment point drops to within the allowable range. The entire process achieves a complete closed loop from data acquisition and fusion evaluation to anomaly location and adaptive adjustment, enabling real-time response to changes in printing conditions and effectively improving the uniformity of pressure distribution and the stability of printing quality.

[0034] For ease of understanding, the following is a specific working example of the present invention: Optimization and adjustment of pressure distribution in a gravure printing press. Printing scenario: A packaging and printing company uses a gravure printing press to print on thin film materials with a thickness of 50 mm. The PET film was used, with a roller reference speed of 150 m / min and a standard operating temperature of 25℃. After the equipment ran continuously for 2 hours, the operator found quality defects such as uneven color depth and blurry images in some areas of the printed material, suspecting that uneven roller pressure distribution was the cause.

[0035] S1. Sensor Layout and Calibration: The equipment has been equipped with piezoelectric pressure sensors (measuring range 3~999N / cm) at both ends and in the middle area of ​​the drum, as required by this invention. 2 Pressure-sensitive thin-film sensors are installed in the pressing zone, displacement sensors are installed at the roller shaft end, and temperature sensors are arranged near the pressing zone. All sensors are uniformly calibrated before startup, and error compensation parameters are stored in the control system.

[0036] S2. Multi-source pressure data acquisition: The system is set to a sampling frequency of 200Hz, a sliding window length of L=0.8 seconds, and a sliding step size of Δt=0.1 seconds. M=12 sampling sections are set along the roller axis, and k=8 sampling points are set circumferentially for each section, forming a total of 96 discrete sampling points. The system triggers all sensors with the same clock source, acquiring dynamic pressure, pressure distribution profile, radial displacement, and temperature data in parallel. Within a sliding window, the system calculates the average pressure at each sampling point. .

[0037] S3. Multi-sensor data fusion processing: The system sequentially performs 3σ criterion anomaly removal, wavelet transform noise reduction, and [0,1] interval standardization on the collected raw data. For example, the original pressure value at a certain sampling point is 850 N / cm². 2 After processing, the values ​​were mapped to 0.85, the displacement data of 0.02mm was mapped to 0.10, and the temperature data of 28℃ was mapped to 0.60, in preparation for subsequent fusion calculations.

[0038] S4. Pressure distribution uniformity assessment: The system calculates the average pressure of all 96 sampling points within the current window and calculates the pressure deviation rate of each point.

[0039] Preset experience weighting coefficient =0.3、 =0.2, maximum permissible radial displacement =0.05mm, current measured displacement D=0.03mm; standard operating temperature 25℃, current temperature 28℃.

[0040] The pressure distribution uniformity index is obtained as follows: =0.704 System preset threshold =0.85. Since 0.704 < 0.85, the pressure distribution is determined to be unsatisfactory, triggering regulation.

[0041] The system calculated the pressure deviation rate at each sampling point and found that the average pressure at the third sampling point on the right end of the drum was the highest in the entire field. Several other points had lower pressure. The system output the locations of the top five sampling points with the highest deviation rates as the target adjustment areas.

[0042] S5, Pressure Distribution Optimization Adjustment: The system receives the positions of 5 target adjustment points. For the high-pressure point on the right, it determines that the pressure is too high and needs to be reduced.

[0043] The current drum speed is 155 m / min, and the reference speed is 150 m / min. =0.02; Material thickness 52 The base thickness is 50. , =0.03.

[0044] Calculate the adjustment amount: =151.6 N / cm 2 The system converts the adjustment amount into a control signal for the actuator (such as an electric cylinder or a pneumatic regulating valve), which drives the corresponding pressure roller on the right end to perform a pressure reduction action.

[0045] During the adjustment process, the system collects pressure changes at this point in real time at a frequency of 200Hz, updating the average pressure every 0.1 seconds via a sliding window, forming a closed-loop feedback. After three iterative adjustments (with each adjustment amount gradually decreasing), the pressure at this point stabilizes at 715 N / cm². 2 Around 0.7%, the deviation rate dropped to below 0.7%, which is lower than the preset allowable deviation threshold (2%).

[0046] Subsequently, the system adjusted the remaining four low-pressure points sequentially according to their pressure deviation rates, from largest to smallest, increasing the pressure using a similar method. After all adjustments were completed, the system recalculated. It increased to 0.91, exceeding the threshold of 0.85.

[0047] Results after adjustment: After the system completed the adjustment, the color uniformity of the printed materials was significantly improved, and the clarity of the images and text returned to the acceptable standard. The entire adjustment process took approximately 12 seconds, requiring no manual intervention, significantly improving production efficiency and product quality consistency. The system recorded the pressure distribution data, adjustment parameters, and results during this adjustment process in a log for future process optimization reference.

[0048] The system preset thresholds and coefficients involved in this application are all empirical values, selected by those skilled in the art based on actual conditions. The formula is obtained by software simulation based on a large amount of collected data, and is the closest to the real situation. All parameters involved in this application have been uniformly normalized to remove dimensions and calculate their values.

[0049] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for online design optimization based on printing rollers, characterized in that, The method includes: S1. Sensor layout and calibration: Multiple types of sensor arrays are arranged on the printing press roller assembly, including piezoelectric pressure sensors, displacement sensors and temperature sensors. S2. Multi-source pressure data acquisition: During the printing process, multi-dimensional data is acquired in real time through a sensor array, including: dynamic pressure data of different positions of the roller acquired by the piezoelectric pressure sensor, radial displacement data of the roller acquired by the displacement sensor, and ambient temperature data of the pressure zone acquired by the temperature sensor. The sampling frequency is set to 100-500Hz to ensure the real-time nature of data acquisition, and the sampling timestamp is recorded to achieve time synchronization of multi-source data. S3. Multi-sensor data fusion processing: Processing the collected multi-source data; S4. Pressure Distribution Uniformity Assessment: Set a threshold for printing pressure distribution uniformity, analyze the data processed in S3, and assess whether the current pressure distribution meets the printing quality requirements based on the analysis results. S5. Pressure distribution optimization and adjustment: Based on the evaluation results of step S4, a closed-loop control strategy is adopted to dynamically adjust the printing pressure.

2. The online design optimization method for printing based on printing rollers according to claim 1, characterized in that, S1 includes uniformly calibrating all sensors to determine their measurement range, accuracy, and error compensation parameters, ensuring the consistency of the detection data. Specifically, piezoelectric pressure sensors are installed at both ends and in the middle of the roller to detect dynamic pressure values, with a measurement range covering 3-999 N / cm². 2 It is adapted to the needs of different printing scenarios; the pressure-sensitive thin-film sensor is laid in the contact pressure area of ​​the roller to detect local pressure peaks and pressure distribution profiles; the displacement sensor is used to detect the radial displacement of the roller, indirectly reflecting pressure changes; the temperature sensor is used to detect the ambient temperature of the pressure area to eliminate the interference of temperature on pressure detection.

3. The online design optimization method for printing based on printing rollers according to claim 1, characterized in that, The multi-source data sampling method in S2 includes the following steps: S21. Spatial Layout of Sampling Points: m sampling sections are evenly spaced along the axial direction of the roller. k sampling points are set at equal angles along the circumference of the roller on each section, forming... Each discrete sampling point ensures coverage of the two ends, the middle, and key locations in the pressure zone of the roller. S22. Sampling frequency and time synchronization: Set the sampling frequency Satisfying 100Hz≤ ≤500Hz, all sensors use the same clock source to trigger sampling and record a unified timestamp. j=1,2,…,N, to achieve time alignment of multi-source data; S23, Parallel acquisition of multi-dimensional data: at each sampling point i and at each sampling time... The following four types of data are collected in parallel: dynamic pressure value, pressure peak value and pressure distribution profile width, roller radial displacement and pressure zone ambient temperature; S24. Data sliding window sampling: A sliding window mechanism is adopted, with a window length L = 0.5-1.0 seconds and a sliding step Δt = 0.1 seconds; The average pressure at each sampling point is calculated within each window, which is used for real-time calculation of the pressure distribution uniformity index. S25. Abnormal sampling data removal and completion: Real-time detection of the validity of data at each sampling point. If data at a certain sampling point is missing or exceeds the sensor's range, spatial interpolation of nearby sampling points or temporal interpolation of the previous moment is used to complete the data, ensuring the integrity of data at n sampling points.

4. The online design optimization method for printing based on printing rollers according to claim 1, characterized in that, The specific process of S3 includes: The collected raw data were subjected to outlier removal, noise reduction, and standardization. The 3σ criterion was used to remove outlier data, and wavelet transform was used to reduce noise and eliminate environmental interference and sensor errors. Data of different dimensions were standardized to map pressure data, displacement data, and temperature data to the [0,1] interval for easy subsequent fusion calculation.

5. The online design optimization method for printing based on printing rollers according to claim 1, characterized in that, The specific process of S4 includes: S41. Based on the multi-source data processed in S3, calculate the pressure distribution uniformity index at the current moment. ; S42. Set the pressure distribution uniformity threshold. , calculate and Comparison: like ≥ The current pressure distribution is determined to meet the printing quality requirements. like If the current pressure distribution is determined to be unsatisfactory for printing quality, pressure optimization adjustment is triggered. S43, when At that time, calculate the pressure deviation rate of each sampling point: ,Will Sort the samples from largest to smallest, locate the r sampling points with the largest pressure deviations, and output them as the target area for pressure adjustment to S5. This represents the average pressure at the i-th sampling point within the current sliding window. This is the arithmetic mean of the pressure values ​​within the current sliding window for all sampling points.

6. The online design optimization method for printing based on printing rollers according to claim 6, characterized in that, S41 includes: Calculate the mathematical average of the actual pressure values ​​of all sampling points within the current sliding window; Calculate the ratio of the absolute value of the pressure deviation at each sampling point to the mathematical average value, average the ratio over all sampling points to obtain the average pressure deviation rate, and subtract the average pressure deviation rate from 1 to obtain the pressure consistency coefficient. Obtain the radial displacement of the roller, calculate the ratio of this displacement to the preset maximum allowable displacement, multiply it by the preset displacement weighting coefficient, and subtract the product from 1 to obtain the displacement influence coefficient. Obtain the ambient temperature of the pressure zone, calculate the ratio of the absolute deviation of this temperature from the standard operating temperature to the standard operating temperature, multiply it by the preset temperature weighting coefficient, and subtract the product from 1 to obtain the temperature influence coefficient. Multiplying the pressure uniformity coefficient, displacement influence coefficient, and temperature influence coefficient yields the pressure distribution uniformity index. The pressure distribution uniformity index is between 0 and 1. The closer the value is to 1, the more uniform the pressure distribution is, and the closer the value is to 0, the less uniform the pressure distribution is.

7. The online design optimization method for printing based on printing rollers according to claim 6, characterized in that, S5 includes: S51. Receive the pressure regulation target region from the non-uniform positioning analysis output of S44. The target region contains the location information of the top r sampling points with the largest pressure deviation rate, where r≥1. S52. For each target adjustment point, based on its pressure deviation rate... and the average pressure within the current sliding window Calculate the required target pressure value and adjustment amount: like > The pressure at this point is determined to be too high, and the pressure needs to be reduced. like < The pressure at this point is determined to be too low, and the pressure needs to be increased. Adjustment amount ,in, The coefficient representing the influence of drum speed. The coefficient representing the influence of the thickness of the printing substrate; S53. Calculate the adjustment amount. Converted into corresponding actuator control signals; S54. During the adjustment process, pressure change data at each adjustment point are collected in real time. The deviation between the actual pressure value and the target pressure value is used as feedback input, and the adjustment is continuously iterated until the pressure deviation rate at that adjustment point is reached. Below the preset allowable deviation threshold; S55. When there are multiple adjustment points, adjust them in order of pressure deviation rate from large to small, or use parallel adjustment method and prioritize the adjustment point with the largest deviation.

8. An online design optimization system based on printing rollers, characterized in that, For implementing the online design optimization method based on printing rollers as described in any one of claims 1-8, the system includes a data acquisition layer, a processing layer, and an execution layer; The acquisition layer includes: The sensor array module, deployed on the printing press roller assembly, includes a piezoelectric pressure sensor, a displacement sensor, and a temperature sensor; the sensor array module is used to collect dynamic pressure data, radial displacement data, and ambient temperature data of the pressure zone at different positions of the roller in real time. The processing layer includes: A multi-source data acquisition module, connected to the sensor array module, is used to acquire multi-dimensional data in real time at a set sampling frequency and achieve time synchronization; The data fusion processing module is connected to the multi-source data acquisition module and is used to perform outlier removal, noise reduction and standardization processing on the acquired multi-source data. The pressure distribution uniformity assessment module is connected to the data fusion processing module. It is used to set the pressure distribution uniformity threshold, analyze and evaluate the processed data, predict future pressure distribution trends, and identify potential risks in advance. The execution layer includes: The pressure distribution optimization and adjustment module is connected to the pressure distribution uniformity evaluation module and is used to dynamically adjust the printing pressure using a closed-loop control strategy based on the evaluation results and predicted trends. The actuator module, connected to the pressure distribution optimization and adjustment module, is used to receive control signals and perform pressure adjustment actions.

Images