Method and apparatus for measuring print pressure distribution in satellite flexographic printing

Abstract

The application discloses a satellite type flexographic printing pressure distribution measurement method and device, collects the contact pressure signals of each printing unit to construct a pressure distribution field and generate a reference pressure line, carries out unit coupling analysis to obtain independent pressure components and identify a pressure transmission path, constructs a pressure measurement key area along the pressure transmission path, carries out boundary response analysis on the measurement key area to generate measurement dead zone parameters, reverses the drum wear characteristics by using the dead zone offset and constructs a wear mapping relationship to realize wear compensation, forms pressure distribution characteristic parameters, decomposes the pressure distribution characteristic parameters into steady-state pressure components and fluctuating pressure components, carries out phase analysis and collaborative correction on the fluctuating components to form a dynamic pressure characteristic library, carries out abnormal detection and correction on the steady-state components to form calibrated pressure data, and fuses to generate a pressure measurement network, constructs a pressure distribution matrix and carries out time lag correction to generate a printing pressure distribution measurement result, and helps to improve printing quality and intelligent control.

Description

Technical Field

[0001] This invention relates to the field of flexographic printing equipment testing technology, and in particular to a method and apparatus for measuring the printing pressure distribution in satellite flexographic printing. Background Technology

[0002] The print quality of satellite flexographic printing presses is directly affected by contact pressure control. Excessive pressure leads to accelerated plate wear and dot gain, while insufficient pressure results in uneven ink coverage and missing images. Because multiple printing color units are arranged around a central cylinder, the pressure states of each unit influence each other. Furthermore, uneven wear occurs on the cylinder surface as the equipment operates for longer periods. These factors present complex technical challenges to pressure control.

[0003] Current technologies primarily rely on manual inspection or fixed-position sensors to obtain pressure information. These methods suffer from problems such as incomplete measurement coverage, inability to distinguish the independent pressure states of each color group, and difficulty in identifying the effects of equipment wear. Particularly during high-speed continuous production, dynamic pressure changes cannot be effectively captured, making it difficult to trace fluctuations in print quality. Therefore, a method is urgently needed to solve at least one of the aforementioned problems. Summary of the Invention

[0004] This invention discloses a method and apparatus for measuring printing pressure distribution in satellite flexographic printing. The aim is to obtain complete pressure distribution information of each printing unit, identify independent pressure states through inter-unit coupling relationship analysis, perform wear characteristic back-inference and adaptive compensation for key measurement areas, and combine steady-state and dynamic pressure synergistic processing to form high-precision pressure distribution measurement results, providing a data foundation for printing quality control and equipment condition assessment.

[0005] The first aspect of this invention provides a method for measuring the printing pressure distribution in satellite flexographic printing, comprising the following steps: The contact pressure signals of each printing unit are collected, a pressure distribution field is constructed based on the contact pressure signals, and a reference pressure line is generated by performing uniformity analysis on the pressure distribution field. Based on the reference pressure line, unit coupling analysis is performed to generate independent pressure components. Correlation identification is performed on the independent pressure components to form pressure transmission paths. Pressure measurement key areas are constructed along the pressure transmission paths. Boundary response analysis is performed on the key pressure measurement area to generate measurement dead zone parameters. The spatial distribution of the measurement dead zone parameters is identified to extract the dead zone offset. The dead zone offset is used to infer the roller wear characteristics to construct a wear mapping relationship. Wear compensation is performed on the key pressure measurement area based on the wear mapping relationship to form pressure distribution characteristic parameters. The pressure distribution characteristic parameters are decomposed into steady-state pressure components and fluctuating pressure components. Phase relationship analysis is performed on the fluctuating pressure components to identify the phase difference between each unit. Based on the phase difference, a unit synchronization index is constructed. The unit synchronization index is used to perform collaborative correction on the fluctuating pressure components to form a dynamic pressure feature library. Anomaly detection is performed on the steady-state pressure component to generate an error compensation amount. The steady-state pressure component is corrected according to the error compensation amount to form calibration pressure data. The calibration pressure data is then fused with the dynamic pressure feature library to generate a pressure measurement network. A pressure distribution matrix is ​​constructed based on the pressure measurement network, and time-delay correction analysis is performed on the pressure distribution matrix to generate printing pressure distribution measurement results.

[0006] A second aspect of the present invention provides a printing pressure distribution measuring device for satellite flexographic printing, comprising: The data acquisition module is used to acquire the contact pressure signals of each printing unit, construct a pressure distribution field based on the contact pressure signals, and perform uniformity analysis on the pressure distribution field to generate a reference pressure line. The coupling analysis module is used to perform unit coupling analysis based on the reference pressure line to generate independent pressure components, identify the correlation of the independent pressure components to form a pressure transmission path, and construct a pressure measurement key area along the pressure transmission path. The wear compensation module is used to perform boundary response analysis on the key pressure measurement area to generate measurement dead zone parameters, identify the spatial distribution of the measurement dead zone parameters to extract the dead zone offset, use the dead zone offset to infer the roller wear characteristics to construct a wear mapping relationship, and perform wear compensation on the key pressure measurement area based on the wear mapping relationship to form pressure distribution feature parameters. The dynamic correction module is used to decompose the pressure distribution characteristic parameters into steady-state pressure components and fluctuating pressure components, perform phase relationship analysis on the fluctuating pressure components to identify the phase difference between each unit, construct a unit synchronization index based on the phase difference, and use the unit synchronization index to perform collaborative correction on the fluctuating pressure components to form a dynamic pressure feature library. An error fusion module is used to detect abnormal features of the steady-state pressure component to generate an error compensation amount, correct the steady-state pressure component according to the error compensation amount to form calibration pressure data, and fuse the calibration pressure data with the dynamic pressure feature library to generate a pressure measurement network. The result output module is used to construct a pressure distribution matrix based on the pressure measurement network, perform time delay correction analysis on the pressure distribution matrix, and generate printing pressure distribution measurement results.

[0007] The beneficial effects of this invention are reflected in the following points: First, by generating independent pressure components through unit coupling analysis, the mixed pressure signal superimposed by multiple printing units is decomposed into independent pressure states for each unit, eliminating measurement deviations caused by inter-unit pressure interference. This ensures that the pressure data of each printing unit accurately reflects its true pressure level, providing accurate data for independent pressure adjustment of each unit and avoiding adjustment errors caused by pressure coupling. Second, by analyzing the spatial offset characteristics of the measurement dead zone, the wear location, wear type, and wear degree of the roller surface are inferred, establishing a quantitative mapping relationship between wear characteristics and pressure measurement deviation. This mapping relationship is used to adaptively compensate and correct the measurement data, effectively eliminating the impact of surface wear caused by long-term roller operation on the accuracy of pressure measurement, enabling the measurement system to maintain stable measurement accuracy throughout the entire life cycle of the equipment. Finally, the pressure data is decomposed into steady-state and fluctuating components and differentiated. By detecting abnormal features and correcting errors in the steady-state components, systematic measurement biases are eliminated. By performing phase relationship analysis and collaborative correction on the fluctuating components, the pressure fluctuations of each unit are synchronized. After fusion, a complete pressure distribution measurement result is formed that simultaneously includes the average pressure level and dynamic fluctuation characteristics, providing comprehensive data support for real-time pressure monitoring, dynamic adjustment, and quality optimization in the printing process.

[0008] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0009] The accompanying drawings illustrate specific examples of the technical solutions described in this invention and, together with the detailed embodiments, form part of the specification, serving to explain the technical solutions, principles, and effects of this invention.

[0010] Unless otherwise specified or defined, the same reference numerals in different figures represent the same or similar technical features, and different reference numerals may be used to represent the same or similar technical features.

[0011] Figure 1 This is a schematic flowchart of a method for measuring printing pressure distribution in satellite flexographic printing according to the present invention.

[0012] Figure 2 This is a structural block diagram of a printing pressure distribution measuring device for satellite-type flexographic printing according to the present invention. Detailed Implementation

[0013] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0014] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0015] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0016] The technical solutions of the embodiments of this application will be described below.

[0017] like Figure 1 As shown, this embodiment of the invention provides a method for measuring the printing pressure distribution in satellite flexographic printing, including the following steps S110-S160: Step S110: Collect the contact pressure signal of each printing unit, construct a pressure distribution field based on the contact pressure signal, and perform uniformity analysis on the pressure distribution field to generate a reference pressure line.

[0018] Specifically, contact pressure signals from each printing unit are collected. Pressure sensor arrays are deployed in the contact area between the printing plate cylinder and the central impression cylinder of each printing unit in the satellite flexographic printing press to collect contact pressure signals in real time. Thin-film pressure sensors or piezoelectric sensors are used, with the sensor thickness controlled between 0.2 and 0.5 mm to ensure no impact on the normal operation of the printing cylinder. Multiple measurement points are evenly distributed along the axial direction of the printing plate cylinder in each printing unit, with a spacing of 50 to 100 mm to cover the effective printing width of the printing plate. Measurement positions are set along the circumferential direction of the cylinder according to the length of the repeating printing unit to ensure the collection of complete pressure change cycles. The measurement range of the pressure sensors is set to 0 to 10 MPa, the measurement accuracy to 0.01 MPa, and the sampling frequency to above 1000 Hz to ensure the capture of dynamic pressure changes during the printing process. The pressure values, timestamps, and spatial location information of each measurement point are recorded. In satellite flexographic printing equipment, different printing units print different colors, and the contact pressure varies between units. The first printing unit has a relatively low contact pressure because the surface of the film being printed is relatively smooth. In subsequent printing units, the surface roughness of the film increases due to the accumulation of ink layers, resulting in a corresponding increase in contact pressure. The pressure values, time information, and spatial location of each measurement point are collected and categorized according to the printing unit number to obtain the contact pressure signal of each printing unit.

[0019] A pressure distribution field is constructed based on contact pressure signals. The pressure values ​​in the contact pressure signals of each printing unit are correlated and matched with the spatial location information of the corresponding measurement points to establish a mapping relationship between pressure and spatial location. A two-dimensional pressure distribution field is constructed using the axial position of the printing cylinder as the horizontal axis coordinate, the circumferential position of the cylinder as the vertical axis coordinate, and the pressure values ​​of the measurement points as the field values. For the area between discrete measurement points, an interpolation algorithm is used to estimate the pressure values ​​at unmeasured locations. The bilinear interpolation method is selected to handle areas with relatively gentle pressure changes, performing weighted calculations based on the pressure values ​​and distances of four adjacent measurement points. The Kriging interpolation method is selected to handle areas with complex pressure changes, considering the spatial correlation between measurement points to accurately reflect the local variation characteristics of the pressure field. Through interpolation calculations, the discrete measurement point data is expanded into a continuous pressure distribution field, forming a pressure value matrix covering the entire contact area of ​​the printing cylinder. In the pressure distribution field of a satellite flexographic printing machine, the pressure distribution is relatively uniform in the central area of ​​the printing cylinder due to its better cylinder rigidity, while the pressure distribution at both ends of the cylinder shows gradual changes or fluctuations due to bearing support and edge effects.

[0020] A baseline pressure line is generated by performing a uniformity analysis of the pressure distribution field. A pressure distribution curve along the axial direction of the printing cylinder is extracted from the pressure distribution field data. The horizontal axis of the pressure distribution curve represents the axial position, and the vertical axis represents the average pressure value at that position. The pressure distribution curve is smoothed to eliminate measurement noise and random fluctuations using either moving average filtering or Gaussian filtering. The moving average filtering window width is set to 3 to 5 measurement points, and the Gaussian filtering standard deviation is set to 0.5 to 1 times the distance between measurement points. The smoothed pressure distribution curve is analyzed to identify the amplitude and frequency characteristics of pressure fluctuations. The amplitude of pressure fluctuations reflects the degree of non-uniformity in pressure distribution; large fluctuation amplitudes indicate non-uniform contact pressure on the printing cylinder. The frequency of pressure fluctuations reflects the periodicity of pressure changes; periodic fluctuations may be related to cylinder eccentricity, gear meshing, or bearing clearance. A baseline pressure value is extracted from the pressure distribution curve, using either the median or a weighted average. The median avoids the influence of extreme values, while the weighted average assigns different weights based on the reliability of the measurement points. A reference pressure line is generated. The reference pressure line is a straight line parallel to the horizontal axis, and its vertical axis value is the reference pressure value. The reference pressure line represents the contact pressure level of the printing unit under ideal conditions.

[0021] Step S120: Based on the reference pressure line, perform unit coupling analysis to generate independent pressure components, identify the correlation of the independent pressure components to form a pressure transmission path, and construct the pressure measurement key area along the pressure transmission path.

[0022] In some embodiments, the step of generating independent pressure components based on the reference pressure line through unit coupling analysis includes: performing unit mapping on the reference pressure line to generate a unit pressure response set; performing interference identification on the unit pressure response set to extract inter-unit pressure interference features; performing interference suppression processing on the pressure interference features to obtain a decoupled pressure signal; and generating independent pressure components through the decoupled pressure signal.

[0023] Cell-level pressure response sets are generated by mapping the reference pressure lines. The pressure response relationships between cells are identified by the numerical differences between the reference pressure lines of each printing cell. The ratio of the reference pressure values ​​of adjacent printing cells is used as the inter-cell response coefficient, R = P_i+1 / P_i, where P_i is the reference pressure value of the i-th printing cell, and P_i+1 is the reference pressure value of the (i+1)-th printing cell. A response coefficient greater than 1 indicates that the pressure of a subsequent cell is higher than that of a preceding cell, and a response coefficient less than 1 indicates that the pressure of a subsequent cell is lower than that of a preceding cell. The deviation between the reference pressure value of each cell and the average reference pressure value of all cells is calculated, and the deviation is D_i = P_i - P_avg, where P_avg is the average value of the reference pressure of all cells. A positive deviation indicates that the pressure of that cell is higher than the average level, and a negative deviation indicates that the pressure of that cell is lower than the average level. The dynamic response pattern of cell pressure is identified by the temporal variation characteristics of the reference pressure values ​​of each cell. The response coefficients and pressure deviations of each cell are integrated into a cell pressure response set. The unit pressure response set records the pressure response data of each printing unit under actual operating conditions, including the pressure relationship between units and the pressure change pattern.

[0024] Interference identification was performed on the unit pressure response set to extract inter-unit pressure interference features. The distribution pattern of the response coefficients of each unit in the unit pressure response set was analyzed to identify inter-unit pressure interference features. These features include the location of the interference source, interference amplitude, interference frequency, and interference propagation direction. The location of the interference source was determined by analyzing the spatial correlation of the pressure deviations of each printing unit. The Pearson correlation coefficient was used to assess the correlation between pressure deviations of each unit pair, with the correlation coefficient being ρ = Cov(X,Y) / [σ(X)·σ(Y)], where ρ is the correlation coefficient, Cov(X,Y) is the covariance of the deviations of the two units, and σ(X) and σ(Y) are the standard deviations of the deviations of the two units, respectively. Unit pairs with an absolute correlation coefficient greater than 0.7 exhibited strong pressure coupling; a positive correlation coefficient indicated co-directional coupling, while a negative correlation coefficient indicated anti-directional coupling. The interference amplitude was obtained from the root mean square value of the pressure deviations of each unit, and the interference frequency was extracted by performing a fast Fourier transform on the pressure deviation signal. In satellite flexographic printing, the pressure interference frequency between adjacent printing units is typically in the range of 5 to 20 Hz, corresponding to the rotation frequency of the central impression cylinder and its harmonics. The direction of interference propagation is determined by analyzing the phase relationship of the pressure deviation signals of each unit; units with leading phase are the interference sources, and units with lagging phase are the interference receivers. The extracted pressure interference features form an interference feature matrix, whose elements include the interference frequency, interference amplitude, and interference propagation direction between unit pairs.

[0025] The pressure interference characteristics are processed to suppress interference and obtain a decoupled pressure signal. Filter parameters are designed based on the interference frequency characteristics. For periodic interference with stable frequencies, a notch filter is used. The center frequency of the notch filter is set to the main interference frequency identified in the interference characteristics, and the notch bandwidth is set to 5% to 10% of the center frequency. After filtering, a decoupled pressure signal suppressing periodic interference is obtained. For time-varying non-stationary interference, an adaptive filtering method is used. The adaptive filter selects a reference signal based on the location characteristics of the interference source and dynamically adjusts the filter coefficients using a minimum mean square error algorithm to gradually eliminate coupled pressure components and obtain a decoupled pressure signal. The filter strength is adjusted according to the interference amplitude characteristics. Units with large interference amplitudes use strong suppression parameters to fully filter out coupled components, while units with small interference amplitudes use weak suppression parameters to retain useful signal components. The phase compensation parameters of the filter are determined based on the interference propagation direction characteristics. After phase correction of the phase-leading interference signal, the decoupled pressure signal is extracted. Filtering is performed in the time domain or frequency domain. Time domain processing uses a recursive filter to output the decoupled pressure signal in real time. Frequency domain processing performs a Fourier transform on the pressure signal to suppress interference components and then performs an inverse transform to obtain the decoupled pressure signal. The obtained decoupled pressure signal is closer to the real independent pressure state of each printing unit, and the pressure coupling effect between units is effectively suppressed.

[0026] Independent pressure components are generated by decoupling the pressure signals. The decoupling pressure signals of each printing unit are calibrated and adjusted to eliminate systematic biases and random errors in the measurement system. The calibration process uses the average value of the baseline pressure line as a reference, calculating the deviation between the mean of the decoupling pressure signal and the mean of the baseline pressure line. Based on this deviation, the decoupling pressure signals are adjusted by overall translation, aligning the mean of the decoupling signals to the mean position of the baseline pressure line for the corresponding unit. The calibration process identifies the difference in fluctuation amplitude between the decoupling pressure signal and the baseline pressure line by comparing their standard deviations. The ratio of the standard deviations is used as the calibration coefficient, K_cal = σ_baseline / σ_decoupled, where σ_baseline is the standard deviation of the baseline pressure line and σ_decoupled is the standard deviation of the decoupling pressure signal. The fluctuation amplitude of the decoupling pressure signal is scaled and adjusted according to the calibration coefficient to ensure its fluctuation characteristics are consistent with the baseline pressure line. The calibrated and adjusted decoupling pressure signals are the independent pressure components. These independent pressure components retain the pressure variation characteristics of each printing unit while eliminating the influence of inter-unit coupling interference.

[0027] In some embodiments, the step of identifying the correlation of the independent pressure components to form a pressure transmission path includes: sorting the independent pressure components by amplitude to generate a pressure gradient sequence; identifying energy dissipation locations from the pressure gradient sequence and extracting dissipation feature points; connecting the dissipation feature points to form an energy transmission trajectory; and determining the pressure transmission path based on the energy transmission trajectory.

[0028] A pressure gradient sequence is generated by sorting the independent pressure components by amplitude. The values ​​of the independent pressure components for each printing unit are extracted and sorted from highest to lowest pressure amplitude. The sorted sequence reflects the distribution of pressure levels and pressure energy in each printing unit. Units with high pressure amplitudes are typically sources or convergence points of pressure energy; these units have higher pressure adjustment settings or bear a larger printing load. Units with low pressure amplitudes are typically the ends or dissipation points of pressure energy; these units have lower pressure adjustment settings or bear a smaller printing load. The pressure difference between adjacent sorted units is obtained; the pressure difference represents the magnitude of the pressure gradient, which is... Where ΔP is the pressure difference between adjacent elements, and Δx is the equivalent distance between elements. The pressure gradient reflects the driving force for pressure energy transfer between elements; the larger the gradient, the stronger the driving force. A pressure gradient sequence is generated, recording the magnitude and direction of the pressure gradient between each pair of elements. The direction of the pressure gradient points from high-pressure elements to low-pressure elements, representing the natural direction of pressure energy transfer. Element pairs with large pressure gradients have a strong driving force for pressure transfer and high energy transfer efficiency. Element pairs with small pressure gradients have a weak driving force for pressure transfer and low or almost no energy transfer efficiency.

[0029] Energy dissipation locations are identified and dissipation feature points are extracted from pressure gradient sequences. During pressure energy transmission, some is dissipated due to friction, material deformation, and damping, converting into heat or other forms of energy. Energy dissipation locations correspond to unit locations where the pressure gradient changes abruptly or the pressure amplitude drops sharply. The criteria for identifying energy dissipation locations include two aspects: first, the rate of change of the pressure gradient exceeds a set threshold, defined as... The first indicator is the rate of change of the pressure gradient along the location, with a threshold typically set to 1.5 to 2 times the average gradient change rate. The second indicator is when the pressure amplitude decays beyond a set proportion, defined as the ratio of the pressure in that unit to the pressure in the previous unit; a proportion below 0.8 indicates significant energy dissipation. Dissipation feature points are extracted, recording the unit number, pressure amplitude, and degree of energy dissipation at the location. The degree of dissipation is quantified by the pressure gradient change rate or the pressure amplitude decay rate; a higher value indicates more severe dissipation. In satellite flexographic printing presses, energy dissipation typically occurs at the exit end of the contact area between the printing plate cylinder and the central impression cylinder, where pressure energy is rapidly released due to the printing substrate detaching from contact.

[0030] Connecting the dissipation feature points forms an energy transfer trajectory. Based on the pressure gradient sequence and the spatial location of the dissipation feature points, connect each dissipation feature point according to the timing and direction of pressure transfer. Determine the starting point of the connection, selecting the printing unit with the highest amplitude of the independent pressure component as the source of energy transfer. Starting from the starting point, connect adjacent dissipation feature points sequentially according to the pressure gradient direction, with the connection following the order of decreasing pressure values. The connection path strictly follows the pressure gradient direction, ensuring that energy is spontaneously transferred along the direction of decreasing pressure. This forms an energy transfer trajectory, which is a path curve connecting high-pressure units to low-pressure units. Each point on the trajectory curve corresponds to a different dissipation feature point location, marking the intermediate nodes of energy transfer. In a six-color satellite flexographic printing press, because the pressure setting on the surface of the smooth film in the first printing unit is relatively high, energy is transferred from the first unit through the central impression cylinder to the second and third units. Significant energy dissipation occurs at the exit end of the contact area in the third unit, and then continues to the fourth and fifth units, finally terminating in the sixth unit, forming a complete energy transfer trajectory. Record the unit number, pressure amplitude, and dissipation degree of each node on the trajectory curve to form trajectory node information. Record the pressure gradient and transmission direction of each connecting segment to form trajectory connection information.

[0031] The pressure transmission path is determined based on the energy transmission trajectory. The main path and key nodes of pressure transmission are extracted based on the energy transmission trajectory. The main path is the segment with the maximum energy flow or minimum dissipation in the energy transmission trajectory; these segments undertake the main pressure energy transmission task. Key nodes are the printing unit positions where energy converges or diverges in the energy transmission trajectory; these nodes play a controlling role in the pressure transmission process. In a satellite flexographic printing press, the pressure energy of the first printing unit is transmitted to the adjacent second and third units through the central impression cylinder. The third unit receives pressure from the first and second units and continues to transmit energy to the fourth unit; this unit position becomes a key node for energy convergence and divergence. The pressure transmission path is determined, extending along the main path of the energy transmission trajectory, connecting the starting unit, intermediate units, and terminating unit. The starting unit is usually the first printing unit or the unit with the highest pressure setpoint; intermediate units are transitional units in the energy transmission process; and the terminating unit is the unit where energy is finally dissipated. The pressure transmission path reflects the propagation law and influence range of pressure changes between printing units. The determined pressure transmission path is recorded in the order of the starting unit, intermediate unit, and ending unit number to form pressure transmission path data.

[0032] A critical pressure measurement zone is constructed along the pressure transmission path. This critical zone is the area on the pressure transmission path that is most sensitive to pressure changes and has the most significant impact on print quality. The pressure gradient distribution at each location along the pressure transmission path is analyzed to identify the peak pressure gradient locations and sections with drastic pressure changes. These locations correspond to key points for pressure measurement; deploying measurement sensors at these locations can most effectively capture the changing characteristics of the pressure distribution. The importance of each printing unit in the pressure transmission path is evaluated, with importance comprehensively assessed through the unit's pressure transmission intensity, energy dissipation, and the range of its influence on downstream units. Highly important printing units are marked as critical pressure measurement units, requiring increased measurement point density or higher measurement frequency. The branching and converging locations of the pressure transmission path are analyzed. Branching locations correspond to the starting points of pressure energy diversion, and converging locations correspond to areas where multiple pressure sources act together. These locations are challenging and crucial for pressure measurement due to the complex pressure interactions. A critical pressure measurement zone is constructed, encompassing key units on the pressure transmission path, peak pressure gradient locations, energy dissipation locations, and pressure branch convergence locations. Planning the arrangement of measurement sensors within critical areas involves increasing the number of sensors, reducing the spacing between measurement points, or using high-precision sensors to ensure accurate measurement of pressure distribution in these areas. Critical pressure measurement areas typically include the inlet area of ​​the first printing unit, the center of the contact area of ​​the intermediate printing units, and the outlet area of ​​the last printing unit. The pressure distribution in these areas directly affects registration accuracy and print quality.

[0033] Step S130: Perform boundary response analysis on the key area of ​​pressure measurement to generate measurement dead zone parameters; identify the spatial distribution of the measurement dead zone parameters to extract the dead zone offset; use the dead zone offset to infer the roller wear characteristics and construct a wear mapping relationship; and perform wear compensation on the key area of ​​pressure measurement based on the wear mapping relationship to form pressure distribution characteristic parameters.

[0034] Specifically, boundary response analysis is performed on the critical pressure measurement area to generate measurement dead zone parameters. Spatial boundary locations and pressure distribution data of the critical pressure measurement area are extracted. Boundary lines of each critical area are identified, separating them from non-critical areas. Pressure values ​​are extracted along the boundary lines, and the pressure gradient variation characteristics at the boundaries are analyzed. The pressure difference between the inner and outer sides of the boundary is calculated; this pressure difference reflects the degree of abrupt change in pressure response at the boundary. Boundary locations with larger pressure differences indicate sensitive pressure response, while those with smaller or near-zero pressure differences indicate insensitive pressure response. Boundary areas with insensitive pressure responses are marked as candidate locations for measurement dead zones. The temporal variation characteristics of the pressure signal at these candidate locations are analyzed to identify the amplitude and frequency characteristics of pressure signal fluctuations. Locations where the pressure signal fluctuation amplitude is less than 50% of the average fluctuation amplitude of the critical pressure measurement area are identified as measurement dead zones. In satellite flexographic printing presses, the edge of the contact area between the printing plate cylinder and the central impression cylinder is prone to forming measurement dead zones due to rapid changes in contact pressure and sensor response lag. Extract the spatial location parameters of the measurement dead zone, including the axial and circumferential coordinates of the dead zone. Generate measurement dead zone parameters, which record the spatial location information of each measurement dead zone.

[0035] The spatial distribution of measurement dead zone parameters is identified, and dead zone offset is extracted. The spatial distribution pattern of each measurement dead zone in the axial and circumferential directions of the printing cylinder is analyzed. The spatial coordinates of each measurement dead zone, including axial and circumferential coordinates, are extracted. The distribution density of measurement dead zones in the axial direction is analyzed to identify concentrated and sparse dead zone regions. The periodicity of the distribution of measurement dead zones in the circumferential direction is analyzed to identify the periodic patterns of dead zone occurrence. The circumferential periodicity pattern is obtained by calculating the circumferential spacing of adjacent measurement dead zones; dead zones with similar spacing indicate a periodic distribution pattern. The centroid coordinates of the spatial positions of all measurement dead zones are calculated, and the centroid coordinates serve as the geometric center reference point for the dead zone distribution. The offset distance between the actual position of each measurement dead zone and the centroid position is calculated; this offset distance is the dead zone offset. The dead zone offset is represented in vector form, including the offset magnitude and offset direction. The offset is the Euclidean distance between the actual position and the centroid position, calculated as d = √[(x_i - x_c)² + (y_i - y_c)²], where x_i and y_i are the coordinates of the i-th dead zone, and x_c and y_c are the coordinates of the centroid. The offset direction is the vector direction from the centroid position to the actual position, calculated using the arctangent function θ = arctan[(y_i - y_c) / (x_i - x_c)]. In actual printing, wear on the printing cylinder surface can cause the measurement dead zone position to shift, with severely worn areas exhibiting larger shifts.

[0036] In some embodiments, the step of using the dead zone offset to infer the roller wear characteristics and construct a wear mapping relationship includes: performing spatial direction analysis on the dead zone offset to extract the offset direction vector; performing vector decomposition on the offset direction vector to obtain radial offset components and tangential offset components; inferring the roller wear characteristics based on the ratio of the radial offset components and the tangential offset components; and constructing a wear mapping relationship based on the roller wear characteristics.

[0037] Spatial orientation analysis is performed on the dead zone offset to extract the offset direction vector. Offset direction information is extracted from the dead zone offset dataset, and the distribution characteristics of the offset direction for each measurement dead zone are analyzed. The offset direction is represented in angle form, ranging from 0 to 360 degrees. 0 degrees corresponds to the positive axial direction of the printing cylinder, and 90 degrees corresponds to the positive circumferential direction. The angular distribution of the offset directions for all measurement dead zones is statistically analyzed to identify the primary and secondary offset directions. The primary offset direction corresponds to the direction interval where the offset angles are concentrated, while the secondary offset direction corresponds to the direction interval where the offset angles are dispersed. Through cluster analysis of the angle distribution, the offset directions of the dead zone offset are divided into different direction groups, with the offset directions within the same direction group being similar. The offset direction angle of each measurement dead zone is converted into a unit offset direction vector, with the x-component being cosθ and the y-component being sinθ, where θ is the offset angle. The unit offset direction vector represents the pure directional information of the dead zone offset, and the vector length is normalized to 1. A complete offset direction vector is generated by combining the magnitude of the dead zone offset for each measurement dead zone. The x-component of the offset direction vector is d×cosθ, and the y-component is d×sinθ, where d is the magnitude of the dead zone offset. The offset direction vector contains both the magnitude and direction information of the dead zone offset; the length of the vector represents the magnitude of the dead zone offset, and the direction of the vector represents the offset direction.

[0038] The offset direction vector is decomposed to obtain radial and tangential offset components. Orthogonal decomposition is performed on each offset direction vector. A radial-tangential coordinate system is established for the printing cylinder, with the radial direction perpendicular to the cylinder axis and pointing towards the cylinder center, and the tangential direction tangential to the cylinder surface along the cylinder circumference. The offset direction vector is decomposed in the radial-tangential coordinate system to obtain radial and tangential offset components. The radial offset component V_r = V·e_r, where V is the offset direction vector, e_r is the radial unit vector, and · represents the vector dot product operation. The tangential offset component V_t = V·e_t, where e_t is the tangential unit vector. The radial offset component reflects the offset distance of the measurement dead zone in the radial direction of the cylinder, which is mainly caused by cylinder diameter wear or cylinder eccentricity. The tangential offset component reflects the offset distance of the measurement dead zone in the circumferential direction of the cylinder, which is mainly caused by circumferential wear of the cylinder surface or cylinder rotation phase deviation. The radial and tangential offset component values ​​of each measurement dead zone are extracted to form radial and tangential offset component datasets.

[0039] For example, the step of inferring the roller wear characteristics based on the ratio of the radial offset component and the tangential offset component includes: determining the ratio of the radial offset component to the tangential offset component as the wear azimuth ratio; performing spatial gradient analysis on the wear azimuth ratio to extract the azimuth gradient vector; identifying the dominant expansion direction from the azimuth gradient vector to form an expansion direction feature; and inferring the roller wear characteristics based on the expansion direction feature.

[0040] The ratio of the radial offset component to the tangential offset component is defined as the wear orientation ratio. The wear orientation ratio is defined as K = V_r / V_t, and its magnitude reflects the relative intensity of roller wear in the radial and tangential directions. When the tangential offset component is close to zero, a correction formula K = V_r / (V_t + ε) is used, where ε is a small variable to prevent the denominator from being zero. Wear orientation ratio values ​​from all measurement dead zones are extracted to form a wear orientation ratio data sequence. The statistical distribution characteristics of the wear orientation ratio data sequence are analyzed to obtain the mean, median, maximum, and minimum values ​​of the wear orientation ratio. The mean wear orientation ratio reflects the dominant trend of overall wear; a mean greater than 1 indicates radial wear is dominant, and a mean less than 1 indicates tangential wear is dominant. The standard deviation of the wear orientation ratio reflects the degree of dispersion of wear direction; a large standard deviation indicates significant differences in wear direction at different locations, while a small standard deviation indicates relatively consistent wear direction. Based on the numerical range of the wear orientation ratio, the roller surface is divided into three regions: a radially dominant region, a tangentially dominant region, and a mixed wear region.

[0041] Spatial gradient analysis is performed on the wear azimuth ratio to extract the azimuth gradient vector. The spatial variation gradient of the wear azimuth ratio is analyzed from the spatial distribution map. The difference in wear azimuth ratio between adjacent measurement dead zones is calculated; this difference reflects the spatial rate of change of the wear azimuth ratio. The axial component of the spatial gradient is... Where ΔK is the difference in wear orientation ratio between adjacent dead zones in the axial direction, and Δx is the axial spacing. The component of the spatial gradient in the circumferential direction is... Where ΔK is the difference in wear azimuth ratio between adjacent dead zones in the circumferential direction, and Δy is the circumferential spacing. The synthesized azimuth gradient vector has the following x-component: The y-component is The magnitude of the azimuth gradient vector represents the spatial rate of change of the wear azimuth ratio; a larger vector magnitude indicates a more drastic change in the wear azimuth ratio. The direction of the azimuth gradient vector indicates the direction of the fastest increase in the wear azimuth ratio; the vector points in the direction of increasing wear azimuth ratio. Analyzing the spatial distribution characteristics of the azimuth gradient vector identifies regions with large and small gradients. Regions with large gradients correspond to locations where wear characteristics change drastically; these locations are the boundaries between different wear types. Regions with small gradients correspond to locations where wear characteristics are relatively uniform.

[0042] The dominant expansion direction is identified from the azimuth gradient vectors to form expansion direction features. The directional angles of all azimuth gradient vectors are statistically analyzed to identify the angular intervals where the directions are concentrated. These angular intervals correspond to the dominant direction of wear expansion, representing the direction with the most significant changes in wear characteristics. The dominant expansion direction is determined by statistically analyzing the frequency distribution of the azimuth gradient vector directional angles. The dominant expansion direction corresponds to the directional angle interval with the highest frequency, where the number of azimuth gradient vectors accounts for the largest proportion of the total. The average directional angle of the dominant expansion direction is extracted; this average directional angle represents the main trend direction of wear expansion. Secondary expansion directions are analyzed; these correspond to the directional angle intervals with the second highest frequency. Secondary expansion directions reflect auxiliary trends in wear expansion. The concentration of the dominant expansion direction is evaluated; this concentration is characterized by the standard deviation or variance of the azimuth gradient vector directional distribution, with smaller values ​​indicating greater concentration. Expansion direction features are formed, including the angles of the dominant and secondary expansion directions, the frequency proportion of each direction, and the directional concentration. These expansion direction features describe the spatial expansion pattern and development trend of roller wear. The angular relationship between the expansion direction features and the axial and circumferential directions of the roller is analyzed to identify the spatial orientation of wear expansion.

[0043] The wear characteristics of the roller are inferred from the characteristics of the expansion direction. Based on the angular characteristics of the dominant expansion direction, the direction of the main force causing wear is determined. A dominant expansion direction close to radial (perpendicular to the roller axis) indicates that wear is mainly caused by radial pressure; a dominant expansion direction close to tangential (along the roller circumference) indicates that wear is mainly caused by friction. The uniformity of wear is determined based on the concentration characteristics of the expansion direction. High concentration indicates consistent wear direction and regular wear pattern; low concentration indicates dispersed wear direction and irregular wear pattern. Combined with the angular characteristics of the secondary expansion direction, complex wear patterns are identified. The presence of a significant secondary expansion direction indicates that wear is influenced by multiple forces, resulting in a complex wear mechanism. The specific types of roller wear characteristics are inferred, including uniform radial wear, localized pitting wear, circumferential scratch wear, and complex wear. Uniform radial wear corresponds to a dominant radial expansion direction with high concentration; localized pitting wear corresponds to a dispersed expansion direction with low concentration; and circumferential scratch wear corresponds to a dominant tangential expansion direction with high concentration. The degree of roller wear is assessed by combining the frequency proportion of the extended direction features and the magnitude of the azimuth gradient vector. A high frequency proportion and a large gradient indicate severe wear. The spatial distribution range of roller wear features is determined, and the boundaries of the wear area are identified by the spatial distribution of the extended direction features.

[0044] A wear mapping relationship is constructed based on the characteristics of roller wear. A mapping relationship between roller wear characteristics and pressure measurement compensation is established. The influence patterns of different wear types on pressure measurement are analyzed: uniform radial wear leads to a lower overall pressure measurement value, local pitting wear leads to increased local pressure measurement fluctuations, and circumferential scratch wear leads to uneven circumferential pressure distribution. Based on the quantified values ​​of wear degree in the roller wear characteristics, the range of pressure measurement compensation is determined. Wear degree is characterized by the magnitude of dead zone offset; a larger offset indicates more severe wear, and the corresponding compensation is larger. A quantitative mapping function of wear degree and compensation is established, with the form E=f(D), where E is the compensation and D is the dead zone offset. Based on the spatial distribution characteristics of roller wear, the spatial distribution pattern of compensation is determined. Areas of concentrated wear correspond to high-value areas of compensation, and areas of dispersed wear correspond to low-value areas of compensation. A wear mapping relationship database is constructed, containing a complete mapping relationship between wear type, wear degree, wear location, and corresponding compensation.

[0045] In some embodiments, the step of performing wear compensation on the key pressure measurement area based on the wear mapping relationship to form pressure distribution characteristic parameters includes: evaluating the compensation amount of the wear mapping relationship to generate a partition compensation coefficient; mapping the partition compensation coefficient to the key pressure measurement area to form a compensation distribution map; extracting the compensation correction value of each measurement point from the compensation distribution map; and using the compensation correction value to perform wear compensation on the key pressure measurement area to form pressure distribution characteristic parameters.

[0046] The wear mapping relationship is evaluated to generate zonal compensation coefficients. The required pressure compensation for different regions in the wear mapping relationship database is assessed. The roller surface is divided into multiple compensation zones along the axial and circumferential directions, each corresponding to a specific wear type and degree. For each compensation zone, the wear type and degree data are extracted from the wear mapping relationship. Based on the wear degree value, the base compensation amount for that zone is determined. The base compensation amount is proportional to the wear degree; zones with severe wear have larger base compensation amounts. Correction coefficients for the compensation amount are adjusted according to the wear type differences. The correction coefficient for uniform radial wear is set to 1.0 to 1.2, for local pitting wear to 0.8 to 1.0, and for circumferential scratch wear to 1.1 to 1.3. The zonal compensation coefficient for each compensation zone is extracted; the zonal compensation coefficient is equal to the base compensation amount multiplied by the correction coefficient. The zonal compensation coefficient reflects the magnitude and direction of the pressure measurement value adjustment required for that region. A smooth transition is applied to the compensation coefficients of adjacent compensation zones to avoid abrupt changes in compensation amount at zone boundaries. The smoothing process uses a weighted average method, where the compensation coefficient at the boundary is the weighted average of the compensation coefficients of the two sub-regions.

[0047] The zonal compensation coefficients are mapped to the key pressure measurement area to form a compensation distribution map. The axial and circumferential coordinates of each measurement point are extracted from the key pressure measurement area. Based on the spatial coordinates of the measurement points, the corresponding compensation zones are located in the zonal compensation coefficient distribution map. The zonal compensation coefficient of each compensation zone is assigned to the corresponding measurement point, establishing a mapping relationship between the measurement point and the compensation coefficient. For measurement points located at the boundaries of compensation zones, interpolation methods are used to determine their compensation coefficients. Bilinear interpolation or nearest neighbor interpolation is used, with weighted calculations based on the distance from the measurement point to the center of the surrounding compensation zones. After mapping the compensation coefficients of all measurement points, the compensation coefficient distribution data for the key pressure measurement area is formed. A compensation distribution map is drawn, using the spatial location of the measurement points as coordinates and the compensation coefficient as the field value. A color gradient visually displays the spatial distribution of the compensation coefficients, with areas of large compensation coefficients displayed in dark colors and areas of small compensation coefficients displayed in light colors. Contour intervals are set on the compensation distribution map to clearly show the gradient of compensation coefficient changes. The spatial characteristics of the compensation distribution map are analyzed to identify high-compensation and low-compensation areas. High-compensation areas correspond to locations with severe wear, while low-compensation areas correspond to locations with slight or no wear.

[0048] Extract the compensation correction value for each measurement point from the compensation distribution map. Read the compensation coefficient value corresponding to the location of each measurement point in the compensation distribution map. Convert the compensation coefficient to a pressure compensation correction value using the formula ΔP=K×P_base, where ΔP is the compensation correction value, K is the compensation coefficient, and P_base is the reference pressure value for that measurement point. The reference pressure value is extracted from the pressure distribution data of the key pressure measurement area, using the theoretical pressure value of the measurement point under no-wear conditions or the average pressure value of the adjacent area. For measurement points with a positive compensation coefficient in the compensation distribution map, the compensation correction value is positive, indicating that the pressure measurement value at that point needs to be increased to compensate for the underestimation caused by wear. For measurement points with a negative compensation coefficient, the compensation correction value is negative, indicating that the pressure measurement value at that point needs to be decreased to compensate for the overestimation caused by abnormal protrusions. For measurement points with a compensation coefficient close to zero, the compensation correction value is close to zero, indicating that the pressure measurement value at that point is basically accurate and does not require adjustment. Extract the compensation correction value for each measurement point to form a compensation correction value dataset. The compensation correction value dataset records the spatial location, original pressure value, compensation coefficient, and compensation correction value for each measurement point. Analyze the statistical distribution characteristics of the compensation correction values ​​to obtain their mean, standard deviation, and distribution range. The mean of the compensation correction values ​​reflects the systematic deviation of the overall pressure measurement, while the standard deviation reflects the degree of difference in compensation requirements at different locations. Identify the extreme points of the compensation correction values: the maximum value corresponds to the measurement location with the most severe wear, and the minimum value corresponds to the measurement location with abnormal protrusions.

[0049] Wear compensation is applied to key pressure measurement areas using compensation correction values ​​to form pressure distribution characteristic parameters. Point-by-point pressure compensation is performed on each measurement point within the key pressure measurement area. The compensation formula is P_compensated = P_measured + ΔP, where P_compensated is the compensated pressure value, P_measured is the original measured pressure value, and ΔP is the compensation correction value for that point. After completing wear compensation for all measurement points, compensated pressure distribution data is obtained. The compensated pressure distribution data eliminates the influence of roller wear on pressure measurement, more accurately reflecting the true contact pressure distribution during the printing process. Pressure distribution characteristic parameters are extracted from the compensated pressure distribution data. The average pressure value is extracted; it is the arithmetic mean of the compensated pressure values ​​at all measurement points. The maximum and minimum pressure values ​​are extracted to identify extreme values ​​in the pressure distribution. The standard deviation of the pressure is extracted; it reflects the dispersion and fluctuation range of the pressure distribution. The pressure gradient is extracted; it is the rate of pressure change between adjacent measurement points, reflecting the spatial variation characteristics of the pressure distribution. The pressure uniformity index is extracted; it is expressed as the ratio of the standard deviation of the pressure to the average pressure value; the smaller the value, the more uniform the pressure distribution.

[0050] Step S140 decomposes the pressure distribution characteristic parameters into steady-state pressure components and fluctuating pressure components. Phase relationship analysis is performed on the fluctuating pressure components to identify the phase difference between each unit. Based on the phase difference, a unit synchronization index is constructed. The unit synchronization index is used to perform collaborative correction on the fluctuating pressure components to form a dynamic pressure feature library.

[0051] Specifically, the pressure distribution characteristic parameters are decomposed into steady-state pressure components and fluctuating pressure components. The average pressure, standard deviation of pressure, and pressure gradient of the key pressure measurement areas for each printing unit are extracted from the pressure distribution characteristic parameters. Time-domain analysis is performed on the pressure time-series data of each unit to separate the steady-state and fluctuating components. The steady-state pressure component corresponds to the DC or low-frequency component of the pressure signal, reflecting the average pressure level and slow change trend of the printing unit. The fluctuating pressure component corresponds to the AC or high-frequency component of the pressure signal, reflecting the dynamic pressure fluctuations during the printing process. The moving average method is used to extract the steady-state pressure component, with the sliding window width set to the length of a complete printing cycle. The average pressure value within the window is taken as the steady-state pressure component at that moment. The fluctuating pressure component is obtained by subtracting the steady-state pressure component from the original pressure signal. The fluctuating pressure component is represented as P_fluct(t) = P_total(t) - P_steady(t), where P_total is the original pressure signal, P_steady is the steady-state pressure component, and P_fluct is the fluctuating pressure component. The amplitude and frequency characteristics of the fluctuating pressure component are analyzed to identify the main fluctuating frequency components. The main frequency of the fluctuating pressure component usually corresponds to the rotation frequency of the central impression cylinder and its multiples. Cylinder eccentricity or uneven printing plate thickness will produce pressure fluctuations at specific frequencies.

[0052] Phase relationship analysis is performed on the fluctuating pressure components to identify the phase difference between units. Fourier transform is applied to the fluctuating pressure components of each unit to convert them to the frequency domain for phase analysis. In the frequency domain, the dominant frequency component and its corresponding phase angle of the fluctuating pressure component of each unit are extracted. The dominant frequency component usually corresponds to the frequency component with the largest amplitude, which reflects the main vibration characteristics of the printing system. The phase angle of the dominant frequency component between adjacent printing units is extracted to determine the phase difference. The phase difference is defined as Δφ = φ_i+1 - φ_i, where φ_i is the phase angle of the i-th unit and φ_i+1 is the phase angle of the (i+1)-th unit. The sign of the phase difference indicates whether the phase is leading or lagging; a positive phase difference indicates that the subsequent unit is leading, and a negative phase difference indicates that the subsequent unit is lagging. The distribution pattern of the phase difference is analyzed to identify phase-synchronized and phase-out unit groups. Unit pairs with an absolute phase difference of less than 30 degrees are marked as phase-synchronized, and unit pairs with an absolute phase difference of more than 30 degrees are marked as phase-out. When the central impression cylinder rotates at a constant speed, the pressure fluctuation peaks of each printing unit should appear sequentially at the same time intervals. However, if the pressure adjustment mechanism of a certain unit responds slowly, the peak value of its fluctuating pressure component will appear delayed, resulting in phase asynchrony. The phase difference values ​​between each pair of units are statistically analyzed to form a phase difference matrix. The phase difference matrix records the phase relationship between any two printing units.

[0053] A unit synchronization index is constructed based on phase difference. The unit synchronization index S_i is defined as the average phase difference between the i-th unit and its adjacent units, expressed as S_i = (|Δφ_i,i-1| + |Δφ_i,i+1|) / 2, where Δφ_i,i-1 is the phase difference between the i-th unit and the preceding unit, and Δφ_i,i+1 is the phase difference between the i-th unit and the following unit. A smaller unit synchronization index value indicates that the phase of the unit is closer to that of its adjacent units, and the operation is more synchronized. A larger unit synchronization index value indicates that the phase of the unit deviates further from that of its adjacent units, and the operation is less synchronized. For the first printing unit, the unit synchronization index only considers the phase difference with the following unit, S_1 = |Δφ_1,2|. For the last printing unit, the unit synchronization index only considers the phase difference with the preceding unit. The unit synchronization index values ​​of all printing units are extracted to form a synchronization index sequence. Analyze the statistical distribution characteristics of the synchronization index sequence to obtain the mean, standard deviation, and extreme values ​​of the unit synchronization index. The mean of the unit synchronization index reflects the overall synchronization level of the entire printing system. The standard deviation of the unit synchronization index reflects the degree of difference in synchronization status among units. Identify the locations of the maximum values ​​of the unit synchronization index; the maximum values ​​correspond to the printing units with the worst synchronization.

[0054] In some embodiments, the step of using the unit synchronization index to collaboratively correct the fluctuating pressure components to form a dynamic pressure feature library includes: determining a threshold for the unit synchronization index to identify out-of-step units; extracting out-of-step mode features from the out-of-step units to generate out-of-step feature vectors; using the out-of-step feature vectors to perform phase reconstruction on the fluctuating pressure components to obtain corrected fluctuating components; and forming a dynamic pressure feature library based on the corrected fluctuating components.

[0055] A threshold-based method is used to identify out-of-sync units based on the unit synchronization index. A synchronization threshold is set to determine out-of-sync units; this threshold is determined based on the statistical distribution characteristics of the unit synchronization index, and is set as the average value of the unit synchronization index plus one standard deviation. Units with synchronization index values ​​exceeding the threshold are marked as out-of-sync units, while those below the threshold are marked as synchronized units. The threshold is applied to each printing unit individually, recording the out-of-sync unit number and its synchronization index value. The number and distribution of out-of-sync units are statistically analyzed to determine their spatial distribution patterns within the printing unit sequence. A concentrated distribution of out-of-sync units indicates a systemic synchronization problem in a localized area, while a dispersed distribution indicates independent synchronization deviations in each unit. The degree of out-of-sync for each unit is extracted, defined as the difference between the unit's synchronization index and the threshold. A higher degree of out-of-sync indicates a more severe synchronization deviation. Out-of-sync units are graded by severity: units with a degree of out-of-sync exceeding twice the threshold are marked as severely out of sync, while units with a degree of out-of-sync between one and two times the threshold are marked as slightly out of sync.

[0056] Out-of-step mode features are extracted from out-of-step units to generate out-of-step feature vectors. Time-series data of the fluctuating pressure components of each out-of-step unit are extracted, and the amplitude, frequency, and phase characteristics of the fluctuating pressure are analyzed. Out-of-step mode features include the fluctuation amplitude deviation, fluctuation frequency deviation, and phase offset of the out-of-step unit. The fluctuation amplitude deviation is defined as the difference between the fluctuation pressure amplitude of the out-of-step unit and the average amplitude of the synchronizing unit. The fluctuation frequency deviation is defined as the difference between the main fluctuation frequency of the out-of-step unit and the system's main frequency. The phase offset is obtained through the phase difference between the out-of-step unit and its adjacent synchronizing unit. The spatial distribution pattern of the out-of-step mode features is analyzed to identify the type of out-of-step mode. Out-of-step mode types include phase-leading, phase-lag, excessively large amplitude, and excessively small amplitude. The phase-leading type corresponds to out-of-step units with a positive and large absolute phase difference, while the phase-lag type corresponds to out-of-step units with a negative and large absolute phase difference. If the pressure regulation response of a certain printing unit is too fast, its fluctuating pressure peak will arrive prematurely, exhibiting a phase-leading and large fluctuation amplitude out-of-step characteristic. Conversely, if the pressure regulation response of another unit is lagging, its fluctuating pressure peak will arrive late, exhibiting a phase-lag out-of-step characteristic. An out-of-step feature vector is generated, which includes the out-of-step unit number, out-of-step mode type, fluctuation amplitude deviation, fluctuation frequency deviation, and phase offset.

[0057] For example, the step of using the out-of-step feature vector to perform phase reconstruction on the fluctuating pressure component to obtain the corrected fluctuating component includes: performing amplitude analysis on the out-of-step feature vector to extract the out-of-step amplitude distribution; identifying the minimum amplitude position from the out-of-step amplitude distribution to obtain a stable reference point; extracting a phase reference value from the stable reference point; and using the phase reference value to perform phase reconstruction on the fluctuating pressure component to obtain the corrected fluctuating component.

[0058] Amplitude analysis is performed on the out-of-step feature vector to extract the out-of-step amplitude distribution. The fluctuation amplitude deviation data of each out-of-step unit is extracted from the feature vector. The fluctuation amplitude deviation reflects the amplitude difference between the out-of-step unit and the synchronization unit. The fluctuation amplitude deviations of each out-of-step unit are arranged according to the unit position to form a out-of-step amplitude distribution sequence. The spatial variation pattern of the out-of-step amplitude distribution sequence is analyzed to identify the peak and trough positions. The peak of the out-of-step amplitude distribution corresponds to the out-of-step unit position with the largest fluctuation amplitude deviation, indicating the most severe out-of-step condition. The trough of the out-of-step amplitude distribution corresponds to the out-of-step unit position with the smallest fluctuation amplitude deviation, indicating a relatively minor out-of-step condition. Statistical features of the out-of-step amplitude distribution are extracted, including the average amplitude deviation, maximum amplitude deviation, minimum amplitude deviation, and standard deviation of the amplitude deviation. The average amplitude deviation reflects the overall amplitude deviation of the out-of-step unit, and the standard deviation reflects the degree of difference in amplitude deviation between different out-of-step units. A out-of-step amplitude distribution curve is plotted, with the horizontal axis representing the out-of-step unit position and the vertical axis representing the fluctuation amplitude deviation. The curve shape visually illustrates the spatial distribution characteristics of the step loss amplitude.

[0059] A stable reference point is obtained by identifying the minimum amplitude location from the out-of-step amplitude distribution. The minimum amplitude deviation is found in the out-of-step amplitude distribution sequence. The out-of-step cell location corresponding to the minimum amplitude deviation in the out-of-step amplitude distribution is marked as a candidate stable reference point. The candidate stable reference points are verified by extracting the fluctuating pressure time-series data at that location and analyzing its stability characteristics. Stability characteristics are evaluated using the standard deviation and coefficient of variation of the fluctuating pressure; smaller values ​​indicate greater stability. The candidate location with the best stability characteristics is selected as the final stable reference point. The stable reference point is the cell location with the smallest fluctuation and closest to the ideal synchronization state among the out-of-step cells. Cells located in the middle of the printed cell sequence are usually relatively stable in pressure fluctuation due to the bidirectional constraint from the cells before and after them, with smaller fluctuation amplitude deviations, making them more likely to become stable reference points. The cell number, spatial location, and fluctuation amplitude deviation value of the stable reference point are recorded. The complete fluctuating pressure component time-series data of the stable reference point is extracted as the data source for subsequent phase reference value extraction.

[0060] Phase reference values ​​are extracted from stable reference points. The fluctuating pressure components at the stable reference points are Fourier transformed to obtain spectral data in the frequency domain. The frequency component with the largest amplitude is identified in the frequency domain spectrum; this frequency component corresponds to the dominant frequency of the fluctuating pressure. The complex representation of the dominant frequency component is extracted, and the phase angle is determined from the real and imaginary parts of the complex number. The phase angle is obtained using the arctangent function, with the formula φ = arctan(Im / Re), where Im is the imaginary part and Re is the real part. This phase angle serves as the phase reference value, representing the most stable phase reference closest to the ideal synchronization state in the out-of-synchronization unit. The dominant frequency phase angle of each synchronization unit is extracted, and the deviation range between the stable reference point phase and the synchronization unit phase is analyzed. The phase deviation range is determined by the maximum and minimum values ​​of the phase angle. The rationality of the phase reference value is verified; the phase reference value should be within a reasonable range of the synchronization unit phase, with a deviation not exceeding 50% of the synchronization unit phase distribution range. The phase reference value is normalized, adjusting the phase angle to the standard range of -180 to 180 degrees. When the phase angle exceeds 180 degrees, subtract 360 degrees for folding. When the phase angle is less than -180 degrees, add 360 degrees for folding. Record the phase reference value, the corresponding frequency component, and the stable reference point number.

[0061] Phase reconstruction of the fluctuating pressure component using a phase reference value yields the corrected fluctuating component. The dominant frequency phase angle of the fluctuating pressure component for each out-of-synchronization unit is extracted to determine the phase deviation between the out-of-synchronization unit's phase and the phase reference value. The phase deviation equals the out-of-synchronization unit's phase minus the phase reference value. Based on the phase deviation, the required phase adjustment for each out-of-synchronization unit is determined; the phase adjustment is equal to the negative of the phase deviation. The fluctuating pressure component of each out-of-synchronization unit undergoes phase translation, with the translation direction opposite to the phase deviation direction. Phase translation is achieved through signal cyclic shift in the time domain or phase angle rotation in the frequency domain. The time delay of the time-domain cyclic shift is obtained using the formula Δt = Δφ / (2πf), where Δφ is the phase adjustment and f is the dominant frequency. The frequency-domain phase rotation is achieved by adding or subtracting the phase adjustment from the phase angle of the spectrum. After phase adjustment, the phase of the fluctuating pressure component of each out-of-synchronization unit is aligned with the phase reference value, and the peak occurrence time of the pressure fluctuation in the out-of-synchronization unit remains consistent with the stable reference point. The fluctuating pressure component of the synchronization unit retains its original phase. The phase-reconstructed wave pressure components of all cells are integrated to form a corrected wave component dataset. The corrected wave components achieve phase consistency of the wave pressure of each printing cell.

[0062] A dynamic pressure feature library is formed based on the corrected fluctuation components. The steady-state pressure components and corrected fluctuation components of each printing unit are superimposed to reconstruct the complete pressure signal of each unit. The complete pressure signal is represented as P_total(t) = P_steady(t) + P_fluct_corrected(t), where P_steady is the steady-state pressure component and P_fluct_corrected is the corrected fluctuation component. Feature parameters of the reconstructed pressure signal are extracted, including the average pressure, peak pressure, pressure fluctuation amplitude, main fluctuation frequency, and phase distribution. The average pressure reflects the steady-state pressure level of the unit, the peak pressure reflects the maximum instantaneous pressure, and the pressure fluctuation amplitude reflects the dynamic range of pressure change. A dynamic pressure feature library data structure is constructed, which includes the steady-state pressure components, corrected fluctuation components, unit synchronization index, out-of-synchronization feature vector, and reconstructed pressure feature parameters of each printing unit.

[0063] Step S150: Detect abnormal features of steady-state pressure components to generate error compensation amount, correct the steady-state pressure components according to the error compensation amount to form calibration pressure data, and fuse the calibration pressure data with the dynamic pressure feature library to generate a pressure measurement network.

[0064] Specifically, anomaly detection is performed on steady-state pressure components to generate error compensation. Statistical analysis is conducted on the steady-state pressure components to obtain the mean and standard deviation. Anomaly detection thresholds are set, with the upper and lower limits being the mean plus or minus three times the standard deviation. Measurement points whose steady-state pressure values ​​exceed the upper and lower limits are marked as anomalies. The spatial distribution characteristics of anomalies are analyzed to identify clustered regions and isolated points. Clustered regions correspond to multiple adjacent measurement points showing anomalies simultaneously, indicating a systematic pressure deviation in that region. Isolated points correspond to a single measurement point showing anomalies, possibly caused by measurement noise or sensor malfunction. Anomaly feature analysis is performed on the anomalies in clustered regions to extract the anomaly amplitude and direction. The anomaly amplitude is defined as the absolute value of the deviation between the anomaly point's pressure value and the mean. The anomaly direction is divided into positive and negative anomalies; positive anomalies correspond to pressure values ​​above the upper limit, and negative anomalies correspond to pressure values ​​below the lower limit. When the installation position of the pressure sensor in a printing unit is slightly offset, the steady-state pressure measurement value at that position will continuously deviate from the normal range, forming positive or negative anomaly characteristics. The error compensation amount is determined based on the abnormal amplitude of the steady-state pressure component. The error compensation amount is equal to the negative of the abnormal amplitude. Positive anomalies correspond to negative error compensation amounts, and negative anomalies correspond to positive error compensation amounts. For isolated anomaly points, the error compensation amount is the difference between the average value of surrounding normal measurement points and the anomaly point.

[0065] The steady-state pressure component is corrected using error compensation to generate calibrated pressure data. For measurement points marked as abnormal in the steady-state pressure component, the corresponding error compensation is added to the steady-state pressure value to obtain the corrected pressure value. The correction formula is P_corrected = P_measured + E_compensation, where P_corrected is the corrected pressure value, P_measured is the original steady-state pressure value, and E_compensation is the error compensation. For normal measurement points not marked as abnormal in the steady-state pressure component, the original steady-state pressure value remains unchanged. After correcting all measurement points, the corrected steady-state pressure data is obtained. The corrected steady-state pressure data undergoes a continuity check to identify abrupt changes in pressure distribution before and after correction. Locations with abrupt changes exceeding a threshold are smoothed using a weighted average method of neighboring measurement points to eliminate discontinuities introduced by the correction. This generates calibrated pressure data, which eliminates abnormal deviations in the steady-state pressure component and more accurately reflects the true average pressure level of each printing unit.

[0066] In some embodiments, fusing the calibration pressure data with the dynamic pressure feature library to generate a pressure measurement network includes: extracting the difference between the calibration pressure data and the dynamic pressure feature library to generate a static-dynamic difference distribution; identifying difference amplitude features from the static-dynamic difference distribution to form a sensitivity distribution map; dividing the sensitivity distribution map into high-sensitivity measurement regions and low-sensitivity measurement regions; and constructing a pressure measurement network by hierarchical fusion based on the high-sensitivity measurement regions and low-sensitivity measurement regions.

[0067] A static-dynamic differential distribution is generated by extracting the difference between calibration pressure data and a dynamic pressure feature library. The corrected fluctuating pressure component is extracted from the dynamic pressure feature library. The calibration pressure data values ​​for each measurement point in each printing unit are extracted; these values ​​represent the steady-state pressure level. The amplitude of the corrected fluctuating pressure component for the corresponding measurement point is extracted; this amplitude represents the degree of dynamic pressure fluctuation. For each measurement point, the difference between the calibration pressure data and the fluctuating pressure amplitude is extracted, defined as D = P_steady - A_fluct, where P_steady is the calibration pressure data and A_fluct is the fluctuating pressure amplitude. The difference reflects the relative strength of the relationship between steady-state pressure and dynamic pressure. A larger difference indicates that steady-state pressure dominates, while a smaller difference indicates that dynamic fluctuation dominates. The differences for each measurement point are arranged according to spatial location to form a static-dynamic differential distribution sequence. The spatial variation pattern of the static-dynamic differential distribution sequence is analyzed to identify high-value and low-value regions. High-value regions correspond to locations where the steady-state pressure is much greater than the fluctuating pressure, indicating relatively stable pressure at these locations. The low-value region corresponds to a location where the steady-state pressure is close to or less than the fluctuating pressure, and the pressure fluctuates drastically at this location.

[0068] Sensitivity distribution maps are generated by identifying the amplitude characteristics of static and dynamic difference distributions. Spatial gradient analysis is performed on the static and dynamic difference distribution to extract the rate of change of static and dynamic difference values ​​between adjacent measurement points. The rate of change is defined as gradient G = ΔD / Δx, where ΔD is the change in the difference between adjacent points in the static and dynamic difference distribution, and Δx is the distance between measurement points. The gradient reflects the drastic spatial change of the static and dynamic difference distribution; a large gradient indicates drastic difference changes, while a small gradient indicates gradual difference changes. The amplitude characteristics of the static and dynamic difference distribution are identified, including the magnitude of the difference gradient, the direction of difference change, and the frequency of difference fluctuations. Locations with large difference gradients are sensitive to pressure changes; small pressure disturbances will cause a significant measurement response, and these are marked as high-sensitivity regions. Locations with small difference gradients are not sensitive to pressure changes; larger pressure disturbances are required to cause a measurement response, and these are marked as low-sensitivity regions. At the center of the contact area of ​​the printing unit, the pressure distribution is relatively uniform and stable, the static and dynamic difference distribution gradient is small, and the measurement sensitivity is low. At the edge of the contact area, rapid pressure changes result in a larger static and dynamic difference distribution gradient, and the measurement sensitivity is high. A sensitivity distribution map is generated based on the difference amplitude characteristics identified from the static and dynamic difference distributions. The sensitivity distribution map uses the spatial location of the measurement point as coordinates and the magnitude of the difference gradient as the field value.

[0069] Based on the sensitivity distribution map, high-sensitivity and low-sensitivity measurement areas were divided. The differential gradient values ​​of each measurement point were extracted from the sensitivity distribution map, and a sensitivity threshold was set for region division. The sensitivity threshold was determined using the median or average of the differential gradients in the sensitivity distribution map. Measurement points with differential gradients greater than the threshold were marked as high-sensitivity measurement areas, and those with differential gradients less than the threshold were marked as low-sensitivity measurement areas. Spatial clustering was performed on the high-sensitivity and low-sensitivity measurement areas, grouping adjacent measurement points of the same type into continuous regions. Boundary lines were identified to separate high-sensitivity and low-sensitivity areas. Geometric features of each region were extracted, including area, shape, and location distribution. High-sensitivity measurement areas are typically located at the edges of the printing unit contact area, pressure gradient peak locations, and key nodes in the pressure transmission path; these locations respond rapidly to pressure changes. Low-sensitivity measurement areas are typically located in the center of the printing unit contact area and in flat areas with uniform pressure distribution; these locations have relatively stable pressure. The number and spatial proportion of high-sensitivity and low-sensitivity measurement areas were statistically analyzed.

[0070] A pressure measurement network was constructed by hierarchically fusing data from high-sensitivity and low-sensitivity measurement zones. Corrected fluctuating pressure components were extracted from a dynamic pressure feature library. For measurement points in the high-sensitivity zone, a high-weight fusion strategy was adopted, assigning a larger fusion weight to the fluctuating pressure components. Pressure data in the high-sensitivity zone is sensitive to dynamic changes, and fluctuating pressure components contain important dynamic feature information. The fusion formula is P_fused = w_s × P_steady + w_d × P_fluct, where w_s is the steady-state weight and w_d is the dynamic weight; for the high-sensitivity zone, w_s = 0.3 and w_d = 0.7. For measurement points in the low-sensitivity zone, a low-weight fusion strategy was adopted, assigning a larger fusion weight to the calibrated pressure data. Pressure in the low-sensitivity zone is relatively stable, and steady-state pressure components are the main features. For the low-sensitivity zone, w_s = 0.7 and w_d = 0.3 were set. After completing the hierarchical fusion of each measurement point, the fused pressure data was obtained. The pressure measurement network is constructed using nodes and edges. Each node corresponds to a measurement point, and its attributes include the fused pressure value, sensitivity level, and spatial location. Network edges connect adjacent measurement points, with edge weights represented by the pressure gradient or correlation coefficient between the measurement points. The pressure measurement network integrates calibration pressure data and dynamic fluctuation characteristics, adaptively fusing data based on the differences between high-sensitivity and low-sensitivity measurement areas.

[0071] Step S160: Construct a pressure distribution matrix based on the pressure measurement network, and perform time delay correction analysis on the pressure distribution matrix to generate printing pressure distribution measurement results.

[0072] Specifically, a pressure distribution matrix is ​​constructed based on a pressure measurement network. The fused pressure values ​​of each measurement point in the pressure measurement network are arranged in two dimensions according to the axial and circumferential positions of the printing plate cylinder to construct the pressure distribution matrix. Rows in the pressure distribution matrix correspond to the axial measurement point positions, columns correspond to the circumferential measurement point positions, and matrix elements are the fused pressure values ​​at the corresponding positions. For blank areas between measurement points in the pressure measurement network, interpolation methods are used to fill the matrix elements. Bilinear interpolation or cubic spline interpolation is selected as the interpolation method, and a weighted estimate is performed based on the pressure values ​​and distances of surrounding measurement points. A complete pressure distribution matrix is ​​formed, covering the entire contact area of ​​the printing plate cylinder and providing continuous pressure distribution data. The numerical distribution characteristics of the pressure distribution matrix are analyzed to identify pressure concentration areas and pressure deficiency areas. Pressure concentration areas correspond to the clustering locations of elements with higher pressure values ​​in the pressure distribution matrix. Pressure deficiency areas correspond to the clustering locations of elements with lower pressure values ​​in the pressure distribution matrix. Statistical characteristic parameters of the pressure distribution matrix are extracted, including the average pressure, pressure variance, pressure gradient distribution, and pressure uniformity.

[0073] Time-delay correction analysis is performed on the pressure distribution matrix to generate printing pressure distribution measurement results. The time delay characteristics of the pressure measurement data in the pressure distribution matrix are analyzed. Pressure sensors have inherent response delays; it takes time for the sensor to convert the pressure signal into an electrical signal. Data acquisition system processing and transmission also introduce time delays. The time delay characteristics of each measurement point in the pressure distribution matrix are identified; the time delay is obtained by summing the sensor response time and data transmission delay. The time delay may differ between different measurement points, varying depending on the sensor type, installation location, and signal transmission distance. The time delay values ​​of each measurement point in the pressure distribution matrix are extracted to form time-delay distribution data. Time-delay correction is performed on the pressure distribution matrix using either time-reverse shift or phase compensation. Time-reverse shift adjusts the measured pressure values ​​in the pressure distribution matrix forward according to the time delay, compensating for the measurement delay. Time-shifting is applied to the pressure time-series data of each measurement point in the pressure distribution matrix, with the shift amount equal to the corresponding time delay value. After time-delay correction, the pressure data of each measurement point in the pressure distribution matrix is ​​aligned to the actual occurrence time, eliminating the time deviation introduced by measurement delay. The printing pressure distribution measurement results are extracted from the corrected pressure distribution matrix. The printing pressure distribution measurement results include the pressure distribution matrix of each printing unit, pressure statistical characteristic parameters, location of pressure concentration areas, and pressure uniformity evaluation. The pressure distribution matrix provides detailed spatial pressure distribution information, the pressure statistical characteristic parameters provide a quantitative assessment of pressure levels, the location of pressure concentration areas indicates key locations requiring adjustment, and the pressure uniformity evaluation reflects the stability of printing quality.

[0074] To implement the above-described method embodiment, a method for measuring printing pressure distribution in satellite flexographic printing is provided to achieve the corresponding functions and technical effects. See also... Figure 2 , Figure 2 This diagram illustrates a structural block diagram of a printing pressure distribution measuring device 200 for satellite flexographic printing according to an embodiment of this application. For ease of explanation, only the parts relevant to this embodiment are shown. The printing pressure distribution measuring device 200 for satellite flexographic printing according to an embodiment of this application includes: The data acquisition module 201 is used to acquire the contact pressure signals of each printing unit, construct a pressure distribution field based on the contact pressure signals, and perform uniformity analysis on the pressure distribution field to generate a reference pressure line. The coupling analysis module 202 is used to perform unit coupling analysis based on the reference pressure line to generate independent pressure components, identify the correlation of the independent pressure components to form a pressure transmission path, and construct a pressure measurement key area along the pressure transmission path. Wear compensation module 203 is used to perform boundary response analysis on the key pressure measurement area to generate measurement dead zone parameters, perform spatial distribution identification on the measurement dead zone parameters to extract dead zone offset, use the dead zone offset to back-infer the roller wear characteristics to construct a wear mapping relationship, and perform wear compensation on the key pressure measurement area based on the wear mapping relationship to form pressure distribution characteristic parameters; The dynamic correction module 204 is used to decompose the pressure distribution characteristic parameters into steady-state pressure components and fluctuating pressure components, perform phase relationship analysis on the fluctuating pressure components to identify the phase difference between each unit, construct a unit synchronization index based on the phase difference, and use the unit synchronization index to perform collaborative correction on the fluctuating pressure components to form a dynamic pressure feature library. Error fusion module 205 is used to detect abnormal features of the steady-state pressure component to generate error compensation amount, correct the steady-state pressure component according to the error compensation amount to form calibration pressure data, and fuse the calibration pressure data with the dynamic pressure feature library to generate a pressure measurement network. The result output module is used to construct a pressure distribution matrix based on the pressure measurement network, perform time delay correction analysis on the pressure distribution matrix, and generate printing pressure distribution measurement results.

[0075] The aforementioned satellite flexographic printing pressure distribution measuring device 200 can implement a satellite flexographic printing pressure distribution measuring method according to the above-described method embodiments. The options in the above method embodiments are also applicable to this embodiment, and will not be detailed here. The remaining contents of this application embodiment can be referred to the contents of the above method embodiments, and will not be repeated in this embodiment.

[0076] The purpose of the above embodiments is to reproduce and derive the technical solution of the present invention by way of example, and to fully describe the technical solution, purpose and effect of the present invention. The purpose is to enable the public to have a more thorough and comprehensive understanding of the disclosure of the present invention, and not to limit the scope of protection of the present invention.

Claims

1. A method for measuring printing pressure distribution in satellite flexographic printing, characterized in that, include: The contact pressure signals of each printing unit are collected, a pressure distribution field is constructed based on the contact pressure signals, and a reference pressure line is generated by performing uniformity analysis on the pressure distribution field. This includes: extracting the axial pressure distribution curve from the pressure distribution field and smoothing and denoising it, analyzing its fluctuation amplitude and frequency characteristics, and extracting the reference pressure value to generate the reference pressure line. The process of generating independent pressure components based on the reference pressure line includes: mapping the reference pressure line to generate a set of unit pressure responses; identifying inter-unit pressure interference features from the set of unit pressure responses; performing interference suppression processing on the pressure interference features to obtain a decoupled pressure signal; generating independent pressure components from the decoupled pressure signal; and identifying the correlation of the independent pressure components to form a pressure transmission path, including: sorting the independent pressure components by amplitude to generate a pressure gradient sequence; identifying energy dissipation locations and extracting dissipation feature points from the pressure gradient sequence; connecting the dissipation feature points to form an energy transmission trajectory; determining the pressure transmission path based on the energy transmission trajectory; and constructing a key pressure measurement area along the pressure transmission path. The process of generating measurement dead zone parameters by performing boundary response analysis on the key pressure measurement area includes: identifying the boundary line from the key pressure measurement area; calculating the pressure difference between the inner and outer sides along the boundary line to screen for insensitive areas; determining the measurement dead zone by combining fluctuation characteristics; extracting its spatial location parameters to generate measurement dead zone parameters; identifying the spatial distribution of the measurement dead zone parameters to extract the dead zone offset; using the dead zone offset to infer the roller wear characteristics to construct a wear mapping relationship; and performing wear compensation on the key pressure measurement area based on the wear mapping relationship to form pressure distribution characteristic parameters. The pressure distribution characteristic parameters are decomposed into steady-state pressure components and fluctuating pressure components. Phase relationship analysis is performed on the fluctuating pressure components to identify the phase difference between each printing unit. Based on the phase difference, a unit synchronization index is constructed using the average phase difference between each printing unit and its adjacent units. The fluctuating pressure components are then collaboratively corrected using the unit synchronization index to form a dynamic pressure feature library. This process includes: threshold determination of the unit synchronization index to identify out-of-synchronization units; extraction of out-of-synchronization mode features from the out-of-synchronization units to generate out-of-synchronization feature vectors; phase reconstruction of the fluctuating pressure components using the out-of-synchronization feature vectors to obtain corrected fluctuating components; and the formation of a dynamic pressure feature library based on the corrected fluctuating components. Anomaly detection is performed on the steady-state pressure component to generate an error compensation amount. The steady-state pressure component is corrected according to the error compensation amount to form calibration pressure data. The calibration pressure data is then fused with the dynamic pressure feature library to generate a pressure measurement network. A pressure distribution matrix is ​​constructed based on the pressure measurement network, and time-delay correction analysis is performed on the pressure distribution matrix to generate printing pressure distribution measurement results.

2. The method according to claim 1, characterized in that, The method of constructing a wear mapping relationship by using the dead zone offset to infer the roller wear characteristics includes: Spatial direction analysis is performed on the dead zone offset to extract the offset direction vector; The offset direction vector is decomposed to obtain the radial offset component and the tangential offset component; The wear characteristics of the roller are inferred from the ratio of the radial offset component to the tangential offset component. A wear mapping relationship is constructed based on the roller wear characteristics.

3. The method according to claim 1, characterized in that, The process of compensating for wear in the key pressure measurement area based on the wear mapping relationship to form pressure distribution characteristic parameters includes: The wear mapping relationship is evaluated to generate partition compensation coefficients. The zoning compensation coefficients are mapped to the key pressure measurement areas to form a compensation distribution map; Extract the compensation correction value for each measurement point from the compensation distribution map; The wear compensation value is used to compensate for the pressure measurement critical area to form pressure distribution characteristic parameters.

4. The method according to claim 1, characterized in that, The step of fusing the calibration pressure data with the dynamic pressure feature library to generate a pressure measurement network includes: The static and dynamic difference distributions are generated by extracting the difference between the calibration pressure data and the dynamic pressure feature library. A sensitivity distribution map is formed by identifying the difference amplitude characteristics from the static and dynamic difference distributions. Based on the sensitivity distribution map, high-sensitivity measurement areas and low-sensitivity measurement areas are divided; A pressure measurement network is constructed by hierarchically fusing the high-sensitivity measurement zone and the low-sensitivity measurement zone.

5. The method according to claim 2, characterized in that, The method of inferring roller wear characteristics based on the ratio of the radial offset component to the tangential offset component includes: The ratio of the radial offset component to the tangential offset component is determined as the wear orientation ratio; Spatial gradient analysis is performed on the wear azimuth ratio to extract the azimuth gradient vector; The dominant expansion direction is identified from the azimuth gradient vector to form an expansion direction feature; The wear characteristics of the roller are inferred from the aforementioned expansion direction characteristics.

6. The method according to claim 1, characterized in that, The step of using the out-of-step feature vector to perform phase reconstruction on the fluctuating pressure component to obtain the corrected fluctuating component includes: Amplitude analysis is performed on the out-of-step feature vector to extract the out-of-step amplitude distribution; A stable reference point is obtained by identifying the position of minimum amplitude from the out-of-step amplitude distribution; Extract the phase reference value at the stable reference point; The phase of the fluctuating pressure component is reconstructed using the phase reference value to obtain the corrected fluctuating component.

7. A printing pressure distribution measuring device for satellite-type flexographic printing, characterized in that, include: The data acquisition module is used to acquire contact pressure signals from each printing unit, construct a pressure distribution field based on the contact pressure signals, and perform uniformity analysis on the pressure distribution field to generate a reference pressure line. This includes: extracting the axial pressure distribution curve from the pressure distribution field and smoothing and denoising it, analyzing its fluctuation amplitude and frequency characteristics, and extracting the reference pressure value to generate the reference pressure line. The coupling analysis module is used to perform unit coupling analysis based on the reference pressure line to generate independent pressure components, including: mapping the reference pressure line to generate a unit pressure response set; identifying interference in the unit pressure response set to extract inter-unit pressure interference features; performing interference suppression processing on the pressure interference features to obtain a decoupled pressure signal; generating independent pressure components through the decoupled pressure signal; and identifying the correlation of the independent pressure components to form a pressure transmission path, including: sorting the independent pressure components by amplitude to generate a pressure gradient sequence; identifying energy dissipation locations and extracting dissipation feature points from the pressure gradient sequence; connecting the dissipation feature points to form an energy transmission trajectory; determining the pressure transmission path based on the energy transmission trajectory; and constructing a pressure measurement key area along the pressure transmission path. The wear compensation module is used to perform boundary response analysis on the key pressure measurement area to generate measurement dead zone parameters. This includes: identifying the boundary line from the key pressure measurement area; calculating the pressure difference between the inner and outer sides along the boundary line to filter insensitive areas; determining the measurement dead zone based on fluctuation characteristics; extracting its spatial location parameters to generate measurement dead zone parameters; identifying the spatial distribution of the measurement dead zone parameters to extract the dead zone offset; using the dead zone offset to infer the roller wear characteristics and construct a wear mapping relationship; and performing wear compensation on the key pressure measurement area based on the wear mapping relationship to form pressure distribution characteristic parameters. A dynamic correction module is used to decompose the pressure distribution characteristic parameters into steady-state pressure components and fluctuating pressure components, perform phase relationship analysis on the fluctuating pressure components to identify the phase difference between each printing unit, construct a unit synchronization index based on the average phase difference between each printing unit and its adjacent units, and use the unit synchronization index to collaboratively correct the fluctuating pressure components to form a dynamic pressure feature library. This includes: determining a threshold for the unit synchronization index to identify out-of-synchronization units; extracting out-of-synchronization mode features from the out-of-synchronization units to generate an out-of-synchronization feature vector; using the out-of-synchronization feature vector to perform phase reconstruction on the fluctuating pressure components to obtain corrected fluctuating components; and forming a dynamic pressure feature library based on the corrected fluctuating components. An error fusion module is used to detect abnormal features of the steady-state pressure component to generate an error compensation amount, correct the steady-state pressure component according to the error compensation amount to form calibration pressure data, and fuse the calibration pressure data with the dynamic pressure feature library to generate a pressure measurement network. The result output module is used to construct a pressure distribution matrix based on the pressure measurement network, perform time delay correction analysis on the pressure distribution matrix, and generate printing pressure distribution measurement results.

Images