Four-color offset press registration error dynamic compensation method and system based on multi-sensor fusion

Abstract

The invention provides a four-color offset press registration error dynamic compensation method and system based on multi-sensor fusion. According to the method, the rotating phase of the printing cylinder, the tension fluctuation value of the printing stock and the periodic offset of the color register mark are collected, the rotating phase of the printing cylinder is matched with the reference phase in real time, the first dynamic compensation signal is generated, and then the color register mark is obtained according to the tension fluctuation of the printing stock and the deviation degree of the tension interval. Generating a second dynamic compensation signal, and performing overlay analysis on the periodic offset, the first dynamic compensation signal and the second dynamic compensation signal to generate a composite compensation instruction when the periodic offset of the chromatography mark is detected to exceed a preset tolerance range; according to the technical scheme provided by the invention, the technical problems of single monitoring dimension, serious compensation hysteresis and poor multi-link collaboration in a traditional method are effectively solved, and the registration precision and the system robustness in a high-speed multi-color printing scene are remarkably improved.

Description

Technical Field

[0001] The present application relates to the field of printing machinery automation control technology, and in particular to a method and system for dynamic compensation of registration error of a four-color offset printing press based on multi-sensor fusion. Background Art

[0002] With the printing industry's increasing demand for high-precision color prints, register accuracy control of four-color offset printing presses has become a key technical link in ensuring printing quality. During high-speed multi-color overprinting, the registration error between color plates will change dynamically due to the combined influence of factors such as mechanical vibration, material deformation, and changes in ambient temperature and humidity. This requires the compensation system to monitor and quickly respond to error changes in real time, while taking into account the high-speed operating characteristics of the printing unit and the physical properties of the printing material.

[0003] Existing technical solutions primarily employ closed-loop feedback control methods based on a single sensor. These methods acquire register mark position information by installing displacement sensors or visual inspection devices at key locations within the printing unit. A PID controller then generates compensation signals to drive the actuators. Some improved solutions attempt to incorporate tension sensors or angle encoders to assist with compensation, but each sensor data is often processed independently, lacking a mechanism for in-depth fusion of multi-source information. Furthermore, existing methods often employ fixed-parameter compensation algorithms, making them difficult to adapt to the dynamic demands of varying printing speeds and material properties. Summary of the Invention

[0004] The present application provides a dynamic compensation method and system for registration error of a four-color offset printing machine based on multi-sensor fusion, which is used to solve the problems in the prior art of insufficient monitoring accuracy of a single sensor, poor coordination of multiple compensation links, and weak adaptability of fixed parameter algorithms.

[0005] In a first aspect, the present application provides a dynamic compensation method for registration error of a four-color offset printing press based on multi-sensor fusion, comprising:

[0006] Collect the rotation phase of the printing cylinder, the tension fluctuation value of the substrate, and the periodic offset of the color mark generated during the operation of the multi-color offset printing unit;

[0007] Matching the rotation phase of the printing cylinder with a preset reference phase in real time to generate a first dynamic compensation signal;

[0008] generating a second dynamic compensation signal according to a degree of deviation between the tension fluctuation of the substrate and a preset tension range;

[0009] When it is detected that the periodic offset of the color mark exceeds a preset tolerance range, the periodic offset is superimposed and analyzed with the first dynamic compensation signal and the second dynamic compensation signal, and a composite compensation instruction is generated according to the analysis result.

[0010] Optionally, matching the rotation phase of the printing cylinder with a preset reference phase in real time to generate a first dynamic compensation signal includes:

[0011] Calculating a dynamic phase difference between the rotation phase of the printing cylinder and a preset reference phase;

[0012] Using a preset sliding window to extract the gradient characteristics of the dynamic phase difference in the current printing cycle;

[0013] Determining an increasing weight of the compensation amplitude as the phase difference increases based on a nonlinear relationship model between the rate of change of the gradient feature and a preset phase compensation curve;

[0014] Performing a product operation on the dynamic phase difference and the incremental weight to generate an initial compensation amount;

[0015] The initial compensation amount is input into a pre-trained phase smoothing filter, and the high-frequency oscillation component is suppressed according to the cumulative trend of the cylinder angle deviation in adjacent printing cycles, and a first dynamic compensation signal is output.

[0016] Optionally, calculating the dynamic phase difference between the rotation phase of the printing cylinder and a preset reference phase includes:

[0017] monitoring a rotational position signal of the printing cylinder and decomposing the rotational position signal into an instantaneous deviation between a current phase angle and a reference phase angle;

[0018] Performing multi-cycle rolling sampling on the instantaneous deviation, and quantifying the instantaneous fluctuation value of the phase jump by a discrete integrator with a preset window length being an integer multiple of the circumference of the printing cylinder;

[0019] Calculating the cumulative offset of the printing cylinder angle deviation according to the change direction and amplitude of the instantaneous fluctuation value during the continuous printing cycle;

[0020] The instantaneous fluctuation value and the accumulated offset are vector-superimposed to generate a dynamic phase difference.

[0021] Optionally, generating a second dynamic compensation signal according to a degree of deviation between the tension fluctuation of the substrate and a preset tension range includes:

[0022] Real-time monitoring of tension fluctuations in the substrate's running path, and performing mean filtering on the tension fluctuations within a preset time window to generate dynamic tension fluctuation values;

[0023] Performing a difference operation on the dynamic tension fluctuation value and the upper and lower limit thresholds of the preset tension range to obtain an instantaneous deviation coefficient;

[0024] Extracting a characteristic vector of the tension change direction according to the historical fluctuation trend of the dynamic tension fluctuation value, and generating a trend factor by rolling and accumulating the characteristic vector over multiple periods;

[0025] Performing nonlinear mapping between the trend factor and a preset tension compensation sensitivity parameter to generate an elastic weight that varies with printing speed, wherein the elastic weight is fixed to a constant value in a low-speed registration stage;

[0026] The instantaneous deviation coefficient and the elastic weight are weighted and superimposed to generate an initial tension compensation amount, and the initial tension compensation amount is input into a preconfigured tension compensation filter. The tension compensation filter is used to suppress low-frequency interference components according to the tension coupling relationship between adjacent printing units, and a second dynamic compensation signal is output.

[0027] Optionally, nonlinear mapping is performed between the trend factor and a preset tension compensation sensitivity parameter to generate an elastic weight that varies with printing speed, including:

[0028] Normalizing the trend factor to generate a standardized trend value, wherein a value range of the standardized trend value matches the physical dimension of the tension compensation sensitivity parameter;

[0029] Determine the speed adjustment coefficient of tension compensation based on the proportional relationship between the current printing speed of the multi-color offset printing unit and the preset speed threshold;

[0030] performing a product operation on the normalized trend value and the tension compensation sensitivity parameter to generate an initial mapping factor, and inputting the initial mapping factor into a preconfigured dynamic scaling module;

[0031] The dynamic scaling module performs nonlinear scaling on the initial mapping factor according to the speed adjustment coefficient to generate an intermediate elastic coefficient;

[0032] The intermediate elastic coefficient is superimposed on a preset elastic reference value to generate an elastic weight, wherein the output direction of the elastic weight is determined by the sign of the characteristic vector of the tension change direction.

[0033] Optionally, when it is detected that the periodic offset of the color mark exceeds a preset tolerance range, the periodic offset is superimposed and analyzed with the first motion compensation signal and the second motion compensation signal, and a composite compensation instruction is generated according to the analysis result, including:

[0034] Detecting the periodic offset of the color mark, and when the periodic offset exceeds the preset tolerance range, extracting the change trajectory of the periodic offset in the continuous printing cycle, and generating a deviation trend prediction curve by least squares fitting, wherein the deviation trend prediction curve includes a transverse deviation component and a longitudinal deviation component;

[0035] decomposing the first dynamic compensation signal into a roller angle compensation component and decomposing the second dynamic compensation signal into a tension fluctuation compensation component;

[0036] Establishing a three-dimensional compensation space coordinate system, wherein a first coordinate axis of the three-dimensional compensation space coordinate system represents the roller angle compensation component, a second coordinate axis represents the tension fluctuation compensation component, and a third coordinate axis represents the predicted offset of the offset trend prediction curve;

[0037] In the three-dimensional compensation space coordinate system, calculating the vector angle between the roller angle compensation component and the tension fluctuation compensation component, and starting the collaborative compensation mode when the vector angle is less than a preset threshold;

[0038] In the collaborative compensation mode, the roller angle compensation component, the tension fluctuation compensation component and the predicted offset are weightedly fused to generate an initial composite compensation vector;

[0039] The initial composite compensation vectors are dynamically normalized to balance the contribution of the initial composite compensation vectors in the high-speed printing stage and the low-speed registration stage, and composite compensation instructions are output.

[0040] Optionally, in the collaborative compensation mode, weighted fusion of the roller angle compensation component, the tension fluctuation compensation component, and the predicted offset is performed to generate an initial composite compensation vector, including:

[0041] Obtaining a speed weight coefficient corresponding to the current printing speed from the operation control system of the multi-color offset printing unit;

[0042] Calculating a synergy factor based on the magnitude of the vector angle;

[0043] Decomposing the predicted offset into a transverse prediction component and a longitudinal prediction component, and establishing mapping relationships with the tension fluctuation compensation component and the roller angle compensation component respectively;

[0044] A dynamic weighting algorithm is used to calculate the real-time fusion weights of the roller angle compensation component, the tension fluctuation compensation component, the transverse prediction component, and the longitudinal prediction component by taking the speed weight coefficient and the synergy factor as input parameters;

[0045] performing a weighted sum operation on the roller angle compensation component, the tension fluctuation compensation component, the transverse prediction component, and the longitudinal prediction component according to the real-time fusion weight to generate a three-dimensional compensation vector;

[0046] Amplitude constraint processing is performed on the three-dimensional compensation vector, and when the component values of the three-dimensional compensation vector after constraint do not exceed a preset maximum compensation limit, an initial composite compensation vector is generated.

[0047] In a second aspect, the present application provides a dynamic compensation system for registration error of a four-color offset printing press based on multi-sensor fusion, comprising:

[0048] The acquisition module is used to collect the rotation phase of the printing cylinder, the tension fluctuation value of the printing material, and the periodic offset of the color mark generated during the operation of the multi-color offset printing unit;

[0049] a matching module, configured to match the rotation phase of the printing cylinder with a preset reference phase in real time to generate a first dynamic compensation signal;

[0050] a generating module, configured to generate a second dynamic compensation signal according to a degree of deviation between the tension fluctuation of the substrate and a preset tension range;

[0051] The analysis module is used to, when detecting that the periodic offset of the color mark exceeds a preset tolerance range, superimpose and analyze the periodic offset with the first dynamic compensation signal and the second dynamic compensation signal, and generate a composite compensation instruction according to the analysis result.

[0052] In a third aspect, an embodiment of the present application provides a computing device comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the dynamic compensation method for registration error of a four-color offset printing machine based on multi-sensor fusion as described in the first aspect above.

[0053] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program. When the computer program is executed by a computer, the method for dynamic compensation of registration error of a four-color offset printing machine based on multi-sensor fusion as described in the first aspect is implemented.

[0054] In the embodiment of the present application, a multi-sensor collaborative monitoring mechanism is used to achieve the synchronous collection and joint analysis of the printing roller phase, substrate tension, and color offset, breaking through the limitations of traditional single sensor monitoring. The first dynamic compensation signal is generated based on the real-time matching of the roller phase deviation, and the second dynamic compensation signal is obtained through the dynamic deviation analysis of the tension fluctuation and the preset interval. The superposition analysis of the two and the color offset forms a composite compensation instruction, which significantly improves the comprehensiveness and coordination of error compensation. This method effectively distinguishes the causes of errors and avoids compensation conflicts through the deep integration of multi-source data, solving the lag problems caused by the single monitoring dimension, isolated compensation links, and solidified parameters in traditional solutions, and significantly improves registration accuracy and system response speed.

[0055] Furthermore, by extracting the gradient features of the dynamic phase difference and adjusting its nonlinear weights, combined with the cumulative trend suppression mechanism of the phase smoothing filter, roller angle compensation can accurately track instantaneous phase jumps during high-speed printing, while effectively filtering out high-frequency noise interference caused by mechanical vibration. Furthermore, by employing a tension fluctuation trend factor and speed-adaptive elastic weight mapping technology, combined with the tension compensation filter's low-frequency interference suppression capability, differentiated compensation for tension fluctuations in the substrate at different printing speeds is achieved. When these two components work together, the phase compensation component prioritizes correcting longitudinal errors, while the tension compensation component focuses on suppressing lateral offsets. Through the phased priority adjustment mechanism of the composite compensation instruction, tension fluctuations are prioritized for stabilization during high-speed printing, while the roller angle is precisely calibrated during low-speed registration. This ultimately significantly improves the convergence speed and compensation stability of registration errors.

[0056] These and other aspects of the present application will become more readily apparent from the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0058] Figure 1 A flow chart of a dynamic compensation method for registration error of a four-color offset printing press based on multi-sensor fusion provided by the present application is shown;

[0059] Figure 2 The present invention provides a schematic structural diagram of a dynamic compensation system for register error of a four-color offset printing press based on multi-sensor fusion;

[0060] Figure 3 A schematic structural diagram of a computing device provided by the present application is shown. DETAILED DESCRIPTION

[0061] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0062] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to being different types.

[0063] Researchers have discovered that existing methods for compensating for register errors in four-color offset printing machines suffer from problems such as a single monitoring dimension, isolated compensation steps, and rigid parameters. These problems make them difficult to adapt to the dynamic error characteristics of high-speed printing, which are caused by the coupling of mechanical vibration and material deformation. Based on this, a dynamic compensation method based on multi-sensor fusion is proposed. This method, through collaborative analysis of multi-source data including print cylinder phase, substrate tension, and color offset, enables precise tracing of error causes and dynamic adaptation of compensation strategies.

[0064] The technical solution of the present application can be applied to the complex registration error compensation needs caused by mechanical transmission deviation, material ductility and environmental disturbance in high-speed multi-color overprinting scenarios, especially in complex working conditions where the printing speed is frequently switched or the physical properties of the substrate change significantly.

[0065] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0066] Figure 1 The present invention provides a flowchart of a method for dynamic compensation of registration error of a four-color offset printing press based on multi-sensor fusion, as shown in FIG. Figure 1 As shown, the method includes:

[0067] Step 101, collecting the rotation phase of the printing cylinder, the tension fluctuation value of the substrate, and the periodic offset of the color mark generated during the operation of the multi-color offset printing unit;

[0068] In this step, the rotational phase of the printing cylinder refers to the real-time rotational angular position of each color group cylinder in the printing unit, which is used to represent the relative phase relationship between the cylinders. The tension fluctuation value of the printed material refers to the real-time change in the tension of the printed material (such as paper, film, etc.) along the conveyor path during the printing process. The periodic offset of the color mark refers to the deviation of the registration mark pre-printed on the substrate from the standard position, which occurs periodically during the printing process. These three parameters respectively reflect the mechanical transmission status of the printing press, the stability of material conveyance, and the color registration accuracy.

[0069] In this embodiment, a rotary encoder mounted on the end of the printing cylinder shaft first collects the rotation angle signals of each color group's rollers in real time (e.g., the current phase of the C color group roller is 152.3°). Simultaneously, a tension sensor array distributed along the substrate conveyor path (e.g., a tension detection roller located after the third color group) monitors material tension changes (e.g., the current tension value is 25.3N±1.2N). Next, a high-speed CCD camera mounted on the paper delivery captures images of the color mark on the substrate (e.g., a 0.15mm longitudinal offset of the M color mark is detected). The periodic offset of each color group is calculated using an image processing algorithm. The rotary encoder phase signal, tension sensor data, and color offset are then timestamped to ensure data synchronization. Finally, all collected data is transmitted to a central control unit via industrial Ethernet, forming a real-time dataset containing roller phase, tension fluctuations, and color offset, which provides input for subsequent compensation signal generation.

[0070] For example, when a Heidelberg XL-106 four-color offset printing press was printing at a high speed of 20,000 sheets per hour, the system detected that the phase of the roller of color group 1 (black) lagged behind the reference phase by 3.2°, the tension fluctuation of the substrate between color groups 2 and 3 reached ±2.5N (beyond the normal range), and through image analysis, it was found that the cyan overprint mark had a periodic fluctuation of ±0.12mm in the longitudinal direction.

[0071] Step 102: Match the rotation phase of the printing cylinder with a preset reference phase in real time to generate a first dynamic compensation signal;

[0072] In this step, the reference phase refers to the preset ideal relative rotational angle between the rollers of each color group, serving as a reference for phase matching. The first dynamic compensation signal is the control command signal generated through real-time phase difference analysis to correct roller angle deviation. These two parameters constitute the reference and output control variable for phase compensation, respectively.

[0073] In this embodiment, first, the preset reference phase data (e.g., the black reference phase is 155.5°) is retrieved from the central control unit and the difference calculation is performed with the real-time collected rotation phase of the printing cylinder (152.3° in step 101) to generate a dynamic phase difference (-3.2°). Secondly, the gradient characteristics of the phase difference within the last five printing cycles are extracted using a sliding window algorithm (e.g., the phase change rate is -0.8° / cycle). The incremental weight (currently 38.4%) is calculated in combination with the preset nonlinear compensation curve (e.g., the compensation weight increases by 12% for every 1° increase in phase difference). Next, the dynamic phase difference is multiplied by the incremental weight to obtain the initial compensation amount (-3.2°×38.4%=-1.23°). After inputting into the phase smoothing filter, the high-frequency component is suppressed based on the cumulative deviation trend of the previous three cycles (e.g., the cumulative lag is 4.1°), and finally the first dynamic compensation signal (axial adjustment +1.15°) is output. All intermediate parameters (dynamic phase difference, incremental weight) are written to the shared data area for subsequent steps.

[0074] For example, in the Heidelberg XL-106 printing scenario mentioned above, the system calculates a 38.4% incremental weight through a nonlinear model based on the difference (-3.2°) between the real-time phase of the black drum (152.3°) and the reference phase (155.5°), combined with the detected phase change rate (-0.8° / cycle). After processing by the phase smoothing filter, a +1.15° compensation instruction is generated to drive the axial adjustment mechanism. Post-execution detection shows that the black drum phase deviation is reduced to 0.8°, providing a corrected phase reference for subsequent tension compensation and color offset analysis.

[0075] Step 103: generating a second dynamic compensation signal according to the degree of deviation between the tension fluctuation of the substrate and a preset tension range;

[0076] In this step, the preset tension range refers to the safe operating range of tension (e.g., 22-28N) set based on the characteristics of the printing material. This range is used to determine whether tension fluctuations are controllable. The second dynamic compensation signal is the control command signal for adjusting the tension roller group, generated through tension deviation analysis and dynamic weight calculation. These two parameters define the boundary conditions for tension control and the compensation output mechanism, respectively.

[0077] In this embodiment, first, the dynamic tension fluctuation value (26.1N±0.8N after mean filtering) is extracted from the collected tension fluctuation value (25.3N±1.2N). Secondly, the dynamic tension fluctuation value is compared with the upper and lower limits of the preset tension range (22-28N), and the instantaneous deviation coefficient is calculated (such as the deviation between the current value 26.1N and the upper limit 28N is -1.9N). Then, the tension history data of the last 10 printing cycles are analyzed, and the characteristic vector of the tension change direction (such as a continuous downward trend) is extracted, and a trend factor (such as a cumulative decrease of -4.2N) is generated by multi-cycle accumulation. Subsequently, based on the proportional relationship between the printing speed (20,000 sheets / hour) and the preset speed threshold (18,000 sheets / hour), an elastic weight is generated (such as an elastic weight of 0.68 corresponding to a speed adjustment coefficient of 1.11). Finally, the instantaneous deviation coefficient (-1.9N) and the elasticity weight (0.68) are weighted and superimposed to generate the initial tension compensation value (-1.29N). After the tension compensation filter suppresses low-frequency interference, the second dynamic compensation signal (driving the tension roller to increase by 1.2N) is output. All intermediate parameters (dynamic tension fluctuation value, trend factor, elasticity weight) are stored in the shared data area.

[0078] For example, after completing roller phase compensation, the system detected that the tension fluctuation between color groups 2 and 3 still exceeded the preset range (26.1N ± 0.8N, with a preset upper limit of 28N). Analysis revealed a continuous downward trend in tension (trend factor -4.2N). Combined with the current high-speed printing state (speed adjustment coefficient 1.11), an elasticity weight of 0.68 was generated. After compensation calculations, the tension roller was driven to increase the tension by 1.2N to 27.3N ± 0.5N.

[0079] Step 104: When it is detected that the periodic offset of the color mark exceeds a preset tolerance range, the periodic offset is superimposed and analyzed with the first motion compensation signal and the second motion compensation signal, and a composite compensation instruction is generated according to the analysis result.

[0080] In this step, the preset tolerance range refers to the maximum offset threshold allowed for the color mark (such as ±0.1mm in the longitudinal direction and ±0.08mm in the transverse direction), which is used to determine whether to trigger composite compensation; the composite compensation instruction refers to a comprehensive control instruction that integrates roller angle compensation, tension compensation and color offset prediction, and is used to coordinate the adjustment of multiple actuators.

[0081] In this embodiment, first, the detected longitudinal offset of the cyan color mark (±0.12mm) is determined to be outside the preset tolerance range (±0.1mm), triggering compensation analysis. Next, the generated first dynamic compensation signal (+1.15° axial adjustment) and the generated second dynamic compensation signal (1.2N tension increase) are extracted and decomposed into a longitudinal compensation component (+1.15°) and a lateral compensation component (+1.2N), respectively. Next, a least squares fit is used to generate an offset trend prediction curve (e.g., the longitudinal offset expands at a rate of 0.02mm / cycle), and a three-dimensional compensation spatial coordinate system is established (X-axis: roller angle compensation component, Y-axis: tension compensation component, Z-axis: predicted offset). The XY plane vector angle (e.g., 38°) is then calculated. Since it is less than a preset threshold (45°), the coordinated compensation mode is activated. The roller angle compensation component (+1.15°), the tension compensation component (+1.2N), and the predicted offset (+0.14mm) are fused according to the speed weight (70% tension weight in the high-speed stage) to generate an initial composite compensation vector (longitudinal +1.03°, lateral +0.86N). Finally, dynamic normalization is used to balance the high-speed / low-speed contributions and output a composite compensation instruction (drive axial adjustment +1.0°, tension roller fine-tuning +0.8N).

[0082] For example, after completing the tension compensation in step 103, the system detected that the longitudinal offset of the cyan overprint mark was still +0.12mm (out of tolerance +0.1mm). Analysis shows that the offset trend prediction curve shows that it will increase to +0.14mm if no intervention is made, and the vector angle (38°) between the roller angle and the tension compensation component shows that the two have good synergy. The system generates a composite instruction according to the high-speed printing weight (tension compensation priority 70%), drives the axial mechanism to adjust by another +1.0° (cumulative compensation +2.15°), and fine-tunes the tension roller by +0.8N (cumulative increase 2.0N). Post-execution detection shows that the cyan overprint offset converges to +0.05mm, and the tension stabilizes at 27.8N±0.3N, verifying the effectiveness of the composite compensation.

[0083] Because conventional printing roller phase compensation methods only use static phase difference calculations, they cannot effectively capture dynamic phase jumps and accumulated deviations during high-speed printing. Therefore, in some embodiments, according to step 102, the rotational phase of the printing roller is matched with a preset reference phase in real time to generate a first dynamic compensation signal, including:

[0084] Step 201, calculating a dynamic phase difference between the rotation phase of the printing cylinder and a preset reference phase;

[0085] In this step, the dynamic phase difference refers to the real-time angular deviation between the actual rotation phase of the printing roller and the preset reference phase. The difference contains an instantaneous jump component and a cumulative offset component; the instantaneous fluctuation value refers to the amplitude characteristic of the phase jump within a single printing cycle quantified by a discrete integrator; the cumulative offset refers to the directional accumulation result of the phase deviation within multiple consecutive cycles.

[0086] In this embodiment, first, the real-time rotational position signal of the black printing cylinder (e.g., the current phase is 152.3°) is acquired via a rotary encoder, and the instantaneous deviation (-3.2°) is calculated relative to the preset reference phase (155.5°). Secondly, a discrete integrator with a window length three times the cylinder circumference (corresponding to three printing cycles) is used to perform rolling sampling of the instantaneous deviation, quantifying the instantaneous fluctuation value (e.g., the fluctuation values for the most recent three cycles are -3.2°, -2.9°, and -3.5°, respectively). Next, the phase change direction for 10 consecutive printing cycles is analyzed (continuous lag trend), and the cumulative offset is calculated (e.g., the cumulative lag over 10 cycles is -32.1°). Finally, the instantaneous fluctuation value (-3.5° for the current cycle) and the cumulative offset (-32.1°) are vector-superimposed (a vector synthesis angle deviation of -35.6°) to generate a dynamic phase difference. All intermediate parameters (instantaneous fluctuation value, cumulative offset) are written to the shared data area for reference in step 202.

[0087] For example, the system detected the instantaneous phase fluctuations of the black drum over three consecutive cycles (-3.2°, -2.9°, and -3.5°). Combined with the accumulated hysteresis over 10 cycles (-32.1°), the system calculated a dynamic phase difference of -35.6° through vector superposition. This parameter serves as input for gradient feature extraction and subsequent compensation weight calculation. The data after execution shows that the dynamic phase difference more accurately reflects the cumulative error characteristics of the mechanical transmission system than the single-cycle measurement.

[0088] Step 202: extracting the gradient feature of the dynamic phase difference in the current printing cycle using a preset sliding window;

[0089] In this step, the sliding window refers to a data processing unit that intercepts continuous printing cycle data in chronological order (such as a window length of 3 cycles) and is used to extract the phase change characteristics of the local time period; the gradient feature refers to the phase difference change rate obtained by differential calculation (such as the phase difference change per cycle).

[0090] In this embodiment, the current printing cycle and the previous two cycle data (-35.6°, -32.1°, -28.5°) are first intercepted from the dynamic phase difference data stream (-35.6°) generated in step 201. Next, the phase difference variation between adjacent cycles is calculated by first-order difference (e.g., the variation of the current cycle: -35.6° - (-32.1°) = -3.5°), and then the average gradient within the window is calculated (the average gradient of 3 cycles is -3.0° / cycle). Finally, the gradient feature (-3.0° / cycle) is written into the feature database.

[0091] Step 203, determining an increasing weight of the compensation amplitude as the phase difference increases based on a nonlinear relationship model between the rate of change of the gradient feature and a preset phase compensation curve;

[0092] In this step, the preset phase compensation curve refers to a nonlinear function model (such as an exponential growth curve) that reflects the relationship between the gradient change rate and the compensation weight; the incremental weight refers to the compensation coefficient that is dynamically adjusted according to the degree of increase in the phase difference (such as the weight increases by 10% for every 1° increase in the phase difference).

[0093] In this example, the gradient feature (-3.0° / cycle) is first retrieved and the compensation curve corresponding to the current printing speed (20,000 sheets / hour) is matched in the preset curve database (including a speed-gradient-weight mapping table). Next, the incremental weight coefficient (42%) is obtained based on the gradient change rate (-3.0° / cycle). This weight coefficient is then associated with the dynamic phase difference (-35.6°) and stored in the compensation parameter library.

[0094] Step 204: multiplying the dynamic phase difference by the incremental weight to generate an initial compensation amount;

[0095] In this step, the initial compensation amount refers to the original compensation value that has not been filtered and contains high-frequency noise components; the product operation refers to the process of performing scalar multiplication calculation on the dynamic phase difference and the incremental weight.

[0096] In this embodiment, the dynamic phase difference (-35.6°) and the incremental weight (42%) are first read from the shared data area. Next, a scalar multiplication operation (-35.6 × 0.42 = -14.95°) is performed to generate the initial compensation value. This value, along with the sign bit (the negative sign indicates the lag compensation direction), is then written to the compensation instruction buffer.

[0097] Step 205: Input the initial compensation amount into a pre-trained phase smoothing filter, suppress the high-frequency oscillation component according to the cumulative trend of the cylinder angle deviation in adjacent printing cycles, and output a first dynamic compensation signal;

[0098] In this step, the phase smoothing filter refers to a digital filter (such as a second-order Butterworth filter) that suppresses high-frequency noise based on historical data; the high-frequency oscillation component refers to a phase rapid fluctuation signal (frequency > 10 Hz) caused by mechanical vibration.

[0099] In this embodiment, the pre-trained filter parameters (cutoff frequency 8Hz, order 2) are first loaded. Next, the initial compensation value (-14.95°) and the compensation values for three adjacent cycles (-12.3°, -13.1°, and -14.2°) are input into the filter. Recursive calculations are used to eliminate high-frequency components (filtering out ±0.8° jitter). Next, the low-frequency compensation value (outputting -12.8°) is enhanced based on the cumulative offset trend (-32.1° accumulated over 10 cycles). Finally, the first dynamic compensation signal (axial adjustment +12.8°) is generated.

[0100] Since simple instantaneous phase difference measurement in the prior art is difficult to reflect the dynamic change characteristics of the printing roller angle deviation, especially unable to distinguish between instantaneous jumps and cumulative offsets, as another embodiment, according to step 201, calculating the dynamic phase difference between the rotation phase of the printing roller and a preset reference phase includes:

[0101] Step 301: monitor the rotation position signal of the printing cylinder and decompose the rotation position signal into an instantaneous deviation between a current phase angle and a reference phase angle;

[0102] In this step, the rotational position signal refers to the electrical signal (such as a pulse sequence) of the printing cylinder angular position obtained in real time by the rotary encoder; the instantaneous deviation refers to the instantaneous angular difference (such as -3.2°) between the current phase angle and the reference phase angle.

[0103] In this example, the real-time rotational position signal (152.3°) is read from the black print cylinder's rotary encoder, and the preset reference phase (155.5°) is retrieved. Next, the angle difference calculation module generates an instantaneous deviation (-3.2°). Finally, the deviation value is associated with a timestamp and stored in the shared data area.

[0104] Step 302: Perform multi-cycle rolling sampling on the instantaneous deviation, and quantify the instantaneous fluctuation value of the phase jump by using a discrete integrator with a preset window length that is an integer multiple of the circumference of the printing cylinder;

[0105] In this step, multi-cycle rolling sampling refers to a data processing method that continuously intercepts the instantaneous deviation sequence according to a fixed window length (such as 3 printing cycles); discrete integrator refers to a phase fluctuation quantization device designed based on a window of integer multiples of the drum circumference.

[0106] In this embodiment, the window length is first set to three times the drum circumference (corresponding to three printing cycles). Next, the instantaneous deviation data (-3.2°, -2.9°, and -3.5°) from step 301 are integrated to calculate the average value of the phase jump within the window (-3.2°). Finally, the instantaneous fluctuation value is output and written to the feature database.

[0107] Step 303, calculating the cumulative offset of the printing roller angle deviation according to the change direction and amplitude of the instantaneous fluctuation value in the continuous printing cycle;

[0108] In this step, the cumulative offset refers to the algebraic sum of phase deviations over multiple consecutive cycles, reflecting the continuous development trend of the angle deviation (e.g., the cumulative lag is -32.1° over 10 cycles).

[0109] In this example, the instantaneous deviation data for the last ten printing cycles (e.g., -3.2°, -2.9°, and -3.5°) is retrieved. Next, the direction of the deviation change (all negative) and the amplitude fluctuation range (-2.5° to -3.8°) are analyzed. Next, the algebraic sum is calculated to obtain the cumulative offset (-32.1°). Finally, the result is associated with the print speed tag and stored in the historical database.

[0110] Step 304: performing vector superposition on the instantaneous fluctuation value and the accumulated offset to generate a dynamic phase difference;

[0111] In this step, vector superposition refers to the process of synthesizing and calculating the instantaneous fluctuation value (amplitude characteristic) and the cumulative offset (trend characteristic) according to the space vector.

[0112] In this embodiment, first, the instantaneous fluctuation value (-3.2°) and the cumulative offset (-32.1°) of step 303 are read. Secondly, the two are converted into vector form in the polar coordinate system (amplitude -3.2° direction 180°, amplitude -32.1° direction 180°). Then, a vector addition operation is performed (synthetic amplitude -35.3°, direction 180°). Finally, the dynamic phase difference (-35.3°) is output to the compensation calculation module. In order to solve the problem that the existing tension compensation method does not take into account the historical trend and speed adaptability of tension fluctuations, resulting in compensation lag, in some embodiments, according to step 103, a second dynamic compensation signal is generated according to the degree of deviation of the tension fluctuation of the substrate from the preset tension range, including:

[0113] Step 401: monitoring the tension fluctuation value in the substrate running path in real time, and performing mean filtering on the tension fluctuation value through a preset time window to generate a dynamic tension fluctuation value;

[0114] In this step, the dynamic tension fluctuation value refers to the tension fluctuation data after filtering, which eliminates the stable fluctuation characteristics after short-term random interference; the preset time window refers to the fixed-length data segment used for mean filtering (such as a 5-second window).

[0115] In this example, the original tension fluctuation value (25.3N ± 1.2N) is first collected in real time from the tension sensor group (the third color group detection roller). Next, a 5-second time window (corresponding to 50 sampling points) is set, and a moving average calculation is performed on the data within the window (for example, the average of 50 data points is 26.1N). Finally, the dynamic tension fluctuation value (26.1N ± 0.8N) is output to the compensation analysis module.

[0116] Step 402: performing a difference operation on the dynamic tension fluctuation value and the upper and lower thresholds of the preset tension range to obtain an instantaneous deviation coefficient;

[0117] In this step, the instantaneous deviation coefficient refers to the real-time difference between the dynamic tension fluctuation value and the preset interval boundary, which quantifies the risk level of tension out of control (such as exceeding the upper limit of -1.9N).

[0118] In this example, the upper and lower limits of the preset tension range (28N) are first read. Next, the difference between the dynamic tension fluctuation value (26.1N) and the upper limit (26.1-28 = -1.9N) is calculated. If the value is positive, the difference is compared to the lower limit. Next, the difference (-1.9N) is mapped to a deviation coefficient (-0.38, normalized to the range [-1, 1]). Finally, the deviation coefficient and direction flag are stored (negative values indicate proximity to the lower limit).

[0119] Step 403: extracting a characteristic vector of the tension change direction based on the historical fluctuation trend of the dynamic tension fluctuation value, and generating a trend factor by rolling and accumulating the characteristic vector over multiple periods;

[0120] In this step, the eigenvector refers to a multidimensional data sequence that reflects the direction of tension change (e.g., [-0.3, -0.5, -0.4] indicates a continuous decrease); the trend factor refers to the total amount of tension change accumulated over multiple periods.

[0121] In this example, dynamic tension data for the most recent 10 printing cycles is first extracted (e.g., 26.1N, 25.8N, 25.3N). Next, the difference between adjacent cycles is calculated to generate a direction vector (e.g., [-0.3N, -0.5N, -0.4N]). Next, a rolling sum is applied to the vector (-0.3 - 0.5 - 0.4 = -1.2N) to generate a trend factor (-4.2N). Finally, the associated timestamp is stored in a historical database.

[0122] Step 404: nonlinearly mapping the trend factor with a preset tension compensation sensitivity parameter to generate an elastic weight that varies with printing speed, wherein the elastic weight is fixed to a constant value in a low-speed registration stage;

[0123] In this step, the tension compensation sensitivity parameter refers to a preset compensation response intensity reference value (eg, 0.5 N / unit deviation); the elastic weight refers to an adaptive compensation coefficient combining a trend factor and a speed change.

[0124] In this example, the trend factor (-4.2N) is first normalized (normalized trend value -0.84). Next, a speed adjustment coefficient (1.11) is calculated based on the current printing speed (20,000 sheets / hour) and the threshold (18,000 sheets / hour). Next, the normalized trend value (-0.84) is multiplied by the sensitivity parameter (0.5) to yield -0.42, which is then passed through the dynamic scaling module (coefficient 1.11) to produce an intermediate elasticity coefficient (-0.47). Finally, the elasticity baseline value (0.2) is superimposed to generate an elasticity weight (-0.27), with the direction determined by the sign of the eigenvector (negative).

[0125] Step 405: Perform weighted superposition of the instantaneous deviation coefficient and the elastic weight to generate an initial tension compensation value, and input the initial tension compensation value into a pre-configured tension compensation filter. The tension compensation filter suppresses low-frequency interference components based on the tension coupling relationship between adjacent printing units, and outputs a second dynamic compensation signal.

[0126] In this step, the low-frequency interference component refers to the slow-changing tension fluctuation (frequency <1 Hz) caused by mechanical wear or material creep; the tension coupling relationship refers to the interaction mechanism of tension transmission between adjacent printing units.

[0127] In this embodiment, the instantaneous deviation coefficient (-0.38) and the elastic weight (-0.27) are first weighted and superimposed (weight ratio 6:4) to generate an initial tension compensation value (-0.38 × 0.6 + (-0.27) × 0.4 = -0.33). Next, the value is input into a low-pass filter with a cutoff frequency of 0.5Hz to filter out high-frequency noise (such as ±0.1N jitter). Next, the compensation value is adjusted based on the tension coupling coefficient (0.9) for color group 2-3 (-0.33 × 0.9 = -0.297N). Finally, a second dynamic compensation signal is output (driving the tension roller to raise by 0.3N).

[0128] To further address the problem that conventional tension compensation weight setting lacks standardization and speed-dependency considerations, which affects compensation accuracy, as another embodiment, according to step 404, the trend factor is nonlinearly mapped with a preset tension compensation sensitivity parameter to generate an elastic weight that varies with printing speed, including:

[0129] Step 501: normalize the trend factor to generate a standardized trend value, wherein the range of the standardized trend value matches the physical dimension of the tension compensation sensitivity parameter;

[0130] In this step, normalization refers to the data standardization process of linearly converting the numerical range of the trend factor to the target interval (such as [-1, 1]); the standardized trend value refers to the dimensionless trend parameter after dimension matching, which can be directly calculated with the tension compensation sensitivity parameter.

[0131] In this example, the generated trend factor (-4.2 N) was first read and its original range (historical maximum ±15 N) was determined. Next, a linear transformation formula (normalized value = trend factor / 5) was used to convert -4.2 N to -0.84 (interval [-1, 1]). Finally, the physical dimensionality matching of the normalized trend value (-0.84) and the tension compensation sensitivity parameter (0.5 N / unit) was verified (0.84 × 0.5 = 0.42 N).

[0132] Step 502 , determining a speed adjustment coefficient for tension compensation based on a proportional relationship between the current printing speed of the multi-color offset printing unit and a preset speed threshold;

[0133] In this step, the speed adjustment coefficient refers to a parameter that is dynamically adjusted according to the ratio of the printing speed to a preset threshold value, and is used to reflect the degree of influence of the speed on the tension compensation.

[0134] In this example, the current printing speed (20,000 sheets / hour) and the preset speed threshold (18,000 sheets / hour) are first obtained. Next, the speed ratio is calculated (20,000 / 18,000 ≈ 1.11). Next, a preset speed-adjustment coefficient mapping table is consulted (a ratio of 1.0-1.2 corresponds to a coefficient of 1.0-1.3), and the speed adjustment coefficient is determined to be 1.15. Finally, this coefficient is stored and associated with a timestamp.

[0135] Step 503 , performing a product operation on the normalized trend value and the tension compensation sensitivity parameter to generate an initial mapping factor, and inputting the initial mapping factor into a preconfigured dynamic scaling module;

[0136] In this step, the initial mapping factor refers to an intermediate parameter generated by multiplying the normalized trend value and the sensitivity parameter; the dynamic scaling module refers to an operation unit that dynamically adjusts the compensation amount according to the working condition parameters.

[0137] In this embodiment, first, the normalized trend value (-0.84) and the sensitivity parameter (0.5N / unit) are retrieved. Next, a product operation is performed (-0.84×0.5=-0.42). Next, the initial mapping factor (-0.42) is input into the buffer of the dynamic scaling module and waits for step 504 to process.

[0138] Step 504: the dynamic scaling module performs nonlinear scaling on the initial mapping factor according to the speed adjustment coefficient to generate an intermediate elastic coefficient;

[0139] In this step, nonlinear scaling refers to the operation process of adjusting the initial mapping factor by exponential or piecewise function according to the speed adjustment coefficient; the intermediate elastic coefficient refers to the compensation strength reference value after scaling.

[0140] In this example, the speed adjustment coefficient (1.15) and the initial mapping factor (-0.42) are first loaded. Next, a segmented scaling strategy is employed: when the coefficient is greater than 1.1, the formula (intermediate value = initial value × coefficient ^ 1.2) is used to calculate (-0.42 × 1.15 ^ 1.2 ≈ -0.47). Finally, the intermediate elasticity coefficient (-0.47) is output to the weight synthesis module.

[0141] Step 505 , superimposing the intermediate elastic coefficient with a preset elastic reference value to generate an elastic weight, wherein the output direction of the elastic weight is determined by the sign of the characteristic vector of the tension change direction;

[0142] In this step, the elastic reference value refers to the static base value of the compensation weight (such as 0.2), which is used to balance the compensation strength of the dynamic adjustment; the output direction refers to the direction of the compensation action (increasing or reducing tension).

[0143] In this embodiment, the intermediate spring coefficient (-0.47) and the preset spring reference value (0.2) are first read. Next, algebraic superposition is performed (-0.47 + 0.2 = -0.27). Next, the output direction is determined as reverse compensation (increasing tension) based on the negative sign of the eigenvector in step 403. Finally, a spring weight (-0.27) is generated and associated with a direction flag.

[0144] Researchers have found that existing registration error compensation methods do not adequately address the synergistic relationship between the various compensation components, which can easily lead to compensation conflicts. To address this issue, in some embodiments, according to step 104, when it is detected that the periodic offset of the color mark exceeds a preset tolerance range, the periodic offset is superimposed and analyzed with the first dynamic compensation signal and the second dynamic compensation signal, and a composite compensation instruction is generated based on the analysis results, including:

[0145] Step 601: Detecting the periodic offset of the color mark. When the periodic offset exceeds the preset tolerance range, extracting the change trajectory of the periodic offset within the continuous printing cycle, and generating a deviation trend prediction curve by least squares fitting, wherein the deviation trend prediction curve includes a transverse deviation component and a longitudinal deviation component.

[0146] In this step, the periodic offset refers to the positioning deviation of the color mark that appears repeatedly in the continuous printing cycle (such as ±0.12mm in the longitudinal direction); the offset trend prediction curve refers to the mathematical model for predicting future offset changes generated by fitting historical data, which includes prediction components in two dimensions: horizontal (left and right) and vertical (front and back).

[0147] In this example, a high-speed CCD camera is used to detect the real-time offset of the cyan color mark (vertical +0.12mm). This is then compared to the preset tolerance range (±0.1mm) to determine if it exceeds the limit. Next, the offset data for the most recent 10 printing cycles (+0.08mm, +0.10mm, +0.12mm, etc.) are extracted and a two-dimensional data set (cycle number, offset) is constructed in time series. Next, the least squares method is used to fit a linear function curve (y = 0.02x + 0.07, R 2 =0.95), and decomposed into a longitudinal component (slope 0.02mm / cycle) and a transverse component (residual fluctuation ±0.03mm). Finally, the predicted curve parameters are written into the compensation decision database.

[0148] Step 602: decompose the first dynamic compensation signal into a roller angle compensation component, and decompose the second dynamic compensation signal into a tension fluctuation compensation component;

[0149] In this step, the roller angle compensation component refers to the control amount in the first dynamic compensation signal used to correct the axial angle deviation of the printing roller (e.g., a +1.15° axial adjustment); the tension fluctuation compensation component refers to the control amount in the second dynamic compensation signal used to adjust the tension of the substrate (e.g., a +0.8N tension increase). These two components correspond to the longitudinal and lateral correction dimensions of the register error, respectively.

[0150] In this embodiment, first, a first dynamic compensation signal (+12.8° axial compensation) is acquired, and its axial adjustment amount is extracted as the longitudinal compensation component (+12.8°). Next, a second dynamic compensation signal (+0.8N tension compensation) is acquired from step 405, and its tension adjustment amount is extracted as the lateral compensation component (+0.8N). Next, the two components are mapped to the X-axis (longitudinal) and Y-axis (lateral) of the three-dimensional compensation space, respectively, and correlated with the Z-axis component (+0.14mm) of the offset trend prediction curve generated in step 601. Finally, a compensation component dataset with direction marks is generated (longitudinal +12.8°, lateral +0.8N).

[0151] Step 603: Establish a three-dimensional compensation space coordinate system, wherein the first coordinate axis of the three-dimensional compensation space coordinate system represents the roller angle compensation component, the second coordinate axis represents the tension fluctuation compensation component, and the third coordinate axis represents the predicted offset of the offset trend prediction curve;

[0152] In this step, the three-dimensional compensation space coordinate system refers to a three-dimensional mathematical model used to comprehensively characterize the relationship between roller angle compensation, tension compensation and predicted offset, wherein the first coordinate axis (X-axis) corresponds to the longitudinal compensation dimension, the second coordinate axis (Y-axis) corresponds to the lateral compensation dimension, and the third coordinate axis (Z-axis) represents the future offset prediction dimension.

[0153] In this embodiment, first, the decomposed roller angle compensation component (+12.8°) is obtained and mapped to the X-axis, and the tension fluctuation compensation component (+0.8N) is mapped to the Y-axis. Secondly, the predicted offset (+0.14mm) of the next cycle in the offset trend prediction curve generated in step 601 is projected to the Z-axis. Then, a coordinate system parameterization model is established, and the X-axis dimension is set to angle (°), the Y-axis dimension is set to tension (N), and the Z-axis dimension is set to displacement (mm). Finally, the calibration parameters (X: +12.8°, Y: +0.8N, Z: +0.14mm) and the directional relationship of each component in the three-dimensional space are stored.

[0154] Step 604: Calculate the vector angle between the roller angle compensation component and the tension fluctuation compensation component in the three-dimensional compensation space coordinate system, and start the collaborative compensation mode when the vector angle is less than a preset threshold.

[0155] In this step, the vector angle refers to the directional difference angle between the two vectors of the roller angle compensation component and the tension fluctuation compensation component in the three-dimensional compensation space; the preset threshold refers to the critical angle value (such as 45°) for determining whether to enable collaborative compensation; the collaborative compensation mode refers to the joint optimization compensation strategy enabled when the directions of the two compensation components are converging.

[0156] In this embodiment, first, the X-axis roller angle compensation component vector (+12.8°, 0, 0) and the Y-axis tension fluctuation compensation component vector (0, +0.8N, 0) are extracted from the three-dimensional compensation space. Secondly, the projection vector angle of the two vectors in the XY plane is calculated (formula: θ = arccos (X·Y / (|X||Y|))), and the angle value is 38°. Then, compared with the preset threshold (45°), the collaborative compensation mode is triggered because 38° < 45°. Finally, a mode start instruction is generated and associated with the current space coordinate parameters (X: +12.8°, Y: +0.8N, Z: +0.14mm), and written into the collaborative control instruction queue.

[0157] Step 605 , in the collaborative compensation mode, weightedly fusing the roller angle compensation component, the tension fluctuation compensation component, and the predicted offset to generate an initial composite compensation vector;

[0158] In this step, weighted fusion refers to the process of assigning weights to multi-dimensional compensation components according to preset rules and performing superposition calculations; the initial composite compensation vector refers to a set of multi-dimensional control instructions after integrating roller angle compensation, tension compensation and predicted offset.

[0159] In this embodiment, first, the roller angle compensation component (+12.8°), tension fluctuation compensation component (+0.8N), and predicted offset (+0.14mm) are read from the three-dimensional compensation space. Secondly, a weight distribution rule is set based on the current printing speed (20,000 sheets / hour): the tension weight in the high-speed stage accounts for 70%, the roller angle accounts for 20%, and the predicted offset accounts for 10%. Next, the contribution value of each component is calculated according to the weight (tension: +0.8N×70%=+0.56N; roller angle: +12.8°×20%=+2.56°; predicted offset: +0.14mm×10%=+0.014mm). Finally, the three are synthesized in three-dimensional space to form the initial composite compensation vector (X:+2.56°, Y:+0.56N, Z:+0.014mm).

[0160] Step 606 , dynamically normalizing the initial composite compensation vectors to balance the contribution of each initial composite compensation vector in the high-speed printing stage and the low-speed registration stage, and outputting composite compensation instructions;

[0161] In this step, dynamic normalization processing refers to the process of converting the multidimensional compensation vector into a unified dimension according to the characteristics of different printing stages; maintaining the balance of contribution refers to adapting the intensity of each compensation component in the high-speed / low-speed stage to the current working conditions through weight adjustment.

[0162] In this example, the initial composite compensation vector (X: +2.56°, Y: +0.56N, Z: +0.014mm) and the current printing speed (20,000 sheets / hour) are first read. Next, normalization factors are calculated for each component: the axial angle factor (1° / mm) and the tension factor (3.5N / mm), converting the X / Y axis components into displacement dimensions (X: +2.56° → +0.18mm, Y: +0.56N → +0.16mm). Next, a balancing rule is set based on the speed state: at high speeds, the tension component weight is increased to 60%, the roller angle weight is reduced to 30%, and the predicted offset weight is 10%; at low speeds, the angle weight is increased to 70%. Subsequently, the compensation amount is recalculated based on the weights (at high speeds: 0.18mm × 30% + 0.16mm × 60% + 0.014mm × 10% = 0.15mm). Finally, a composite compensation instruction is output (axial adjustment +0.15mm equivalent angle, tension increase +0.16mm equivalent value).

[0163] Because a simple weighted average fusion method is difficult to adapt to changes in compensation requirements at different printing speeds, as another embodiment, according to step 604, in the collaborative compensation mode, the roller angle compensation component, the tension fluctuation compensation component, and the predicted offset are weightedly fused to generate an initial composite compensation vector, including:

[0164] Step 701: Obtain a speed weight coefficient corresponding to the current printing speed from the operation control system of the multi-color offset printing unit;

[0165] In this step, the speed weight coefficient refers to the compensation component weight distribution parameter dynamically adjusted according to the printing speed, which is used to reflect the priority relationship of each compensation component in different speed stages (such as the tension compensation weight is higher in the high-speed stage).

[0166] In this example, the current printing speed (20,000 sheets / hour) is first read from the printing unit PLC controller. Next, a preset speed-weight mapping table is queried (the high-speed range of 20,000-25,000 sheets / hour corresponds to a tension weight of 0.7 and a roller angle weight of 0.3). The current speed weight coefficient (tension 0.7, angle 0.3) is extracted. This coefficient is then associated with the compensation mode label (high speed / low speed) and written into the weight parameter library. Finally, the coefficient is transmitted in real time to the compensation fusion module via the data bus.

[0167] Step 702: Calculate the synergy factor based on the angle between the vectors.

[0168] In this step, the synergy factor refers to a dimensionless parameter that reflects the degree of synergy between the roller angle compensation and the tension compensation components, and its value is negatively correlated with the vector angle (the smaller the angle, the higher the synergy).

[0169] In this embodiment, the vector angle (38°) and the preset maximum synergy angle (45°) are first obtained. Next, the synergy factor is calculated using a linear mapping formula (1-included angle / maximum angle = 1-38 / 45 ≈ 0.82). Next, the synergy factor (0.82) is associated with the current printing speed label (high speed 20,000 sheets / hour) and stored in the synergy parameter library. Finally, the synergy factor is transmitted in real time via the data bus to the weighted fusion module in step 605.

[0170] Step 703: Decompose the predicted offset into a transverse prediction component and a longitudinal prediction component, and establish mapping relationships with the tension fluctuation compensation component and the roller angle compensation component respectively;

[0171] In this step, the lateral prediction component refers to the predicted deviation amount perpendicular to the printing direction in the offset trend prediction curve (such as left and right offset ±0.05mm); the longitudinal prediction component refers to the predicted deviation amount along the printing direction (such as front and back offset ±0.12mm); the mapping relationship refers to the corresponding rule that establishes a mathematical association between the mechanical compensation component and the predicted offset dimension (such as the longitudinal prediction component is associated with roller angle compensation, and the lateral prediction component is associated with tension compensation).

[0172] In this embodiment, first, the longitudinal component (+0.14mm) and the transverse residual component (+0.05mm) are extracted from the offset trend prediction curve. Secondly, a linear mapping between the longitudinal prediction component and the roller angle compensation component is established (1° axial adjustment ≈ 0.1mm longitudinal offset correction), and a nonlinear mapping between the transverse prediction component and the tension compensation component is established (1N tension adjustment ≈ 0.06mm transverse offset correction). Then, the mapping coefficients (longitudinal 10mm / °, transverse 16.67N / mm) are written into the compensation relationship database. Finally, a mapping parameter set with direction identification is generated (longitudinal: +0.14mm→+1.4°; transverse: +0.05mm→+0.83N)

[0173] Step 704: Using a dynamic weighting algorithm, the speed weight coefficient and the synergy factor are used as input parameters to calculate the real-time fusion weights of the roller angle compensation component, the tension fluctuation compensation component, the transverse prediction component, and the longitudinal prediction component.

[0174] In this step, the dynamic weighted algorithm refers to an adaptive calculation method that dynamically adjusts the weights of each compensation component according to the operating parameters; the real-time fusion weight refers to the real-time action proportional coefficient assigned to the roller angle compensation, tension compensation and prediction components after dynamic calculation.

[0175] In this embodiment, first, the velocity weight coefficient (tension 0.7, angle 0.3) and the synergy factor (0.82) are called. Next, a weight calculation model is established: tension-based weight = velocity weight × synergy factor (0.7 × 0.82 = 0.574), angle-based weight = 0.3 × 0.82 = 0.246. Next, the lateral prediction component (+0.05 mm) is converted into an equivalent tension weight (0.05 × 16.67 / 0.8 N = 1.04) according to the tension mapping relationship (16.67 N / mm), and the longitudinal prediction component (+0.14 mm) is converted into an equivalent angle weight (0.14 × 10 / 12.8° = 0.11) according to the angle mapping relationship (10 mm / °). Finally, the four categories of weights are normalized (sum = 0.574 + 0.246 + 1.04 + 0.11 = 1.97) to generate real-time fusion weights (tension 29.1%, angle 12.5%, horizontal prediction 52.8%, vertical prediction 5.6%).

[0176] Step 705 , performing a weighted sum operation on the roller angle compensation component, the tension fluctuation compensation component, the transverse prediction component, and the longitudinal prediction component according to the real-time fusion weight to generate a three-dimensional compensation vector;

[0177] In this step, the weighted summation operation refers to the mathematical calculation of proportional superposition of each compensation component according to the real-time fusion weight; the three-dimensional compensation vector refers to the final compensation instruction set containing three dimensions: longitudinal (X-axis), lateral (Y-axis) and predicted (Z-axis).

[0178] In this example, the real-time fusion weights are first obtained (roller angle 12.5%, tension 29.1%, lateral prediction 52.8%, longitudinal prediction 5.6%). Next, the mapped component values are read (roller angle +3.96°, tension +1.39N, lateral prediction +0.83N, longitudinal prediction +1.4°). Next, the contribution of each component is calculated based on the weights: roller angle (+3.96° × 12.5% = +0.495°), tension (+1.39N × 29.1% = +0.404N), lateral prediction (+0.83N × 52.8% = +0.438N), and longitudinal prediction (+1.4° × 5.6% = +0.078°). Finally, the longitudinal dimension (X-axis) is merged (+0.495°+0.078°=+0.573°), the transverse dimension (Y-axis) is merged (+0.404N+0.438N=+0.842N), and the predicted offset (Z-axis +0.14mm) is retained to generate a three-dimensional compensation vector (X: +0.573°, Y: +0.842N, Z: +0.14mm).

[0179] Step 706 , performing amplitude constraint processing on the three-dimensional compensation vector, and generating an initial composite compensation vector when the component values of the constrained three-dimensional compensation vector do not exceed a preset maximum compensation limit;

[0180] In this step, amplitude constraint processing refers to limiting the values of each component of the three-dimensional compensation vector to ensure that the compensation amount does not exceed the safety allowable range of the equipment; the maximum compensation limit refers to the preset single maximum allowable value of the roller angle adjustment, tension adjustment and offset correction (such as angle ±2.0°, tension ±1.5N).

[0181] In this embodiment, first, read the values of each component of the three-dimensional compensation vector (X: +0.573°, Y: +0.842N, Z: +0.14mm). Secondly, call the preset limit parameters: single maximum compensation of the roller angle ±0.5°, tension adjustment ±0.8N, predicted offset correction ±0.2mm. Then, limit the X-axis component (+0.573°>+0.5°, corrected to +0.5°), the Y-axis component (+0.842N>+0.8N, corrected to +0.8N), and the Z-axis component (+0.14mm is within ±0.2mm, keep the original value). Finally, generate the constrained initial composite compensation vector (X: +0.5°, Y: +0.8N, Z: +0.14mm) and associate the limit mark.

[0182] Figure 2 The present application provides a schematic structural diagram of a dynamic compensation system for register error of a four-color offset printing press based on multi-sensor fusion, as shown in FIG. Figure 2 As shown, the system includes:

[0183] The acquisition module 21 is used to collect the rotation phase of the printing cylinder, the tension fluctuation value of the substrate, and the periodic offset of the color mark generated during the operation of the multi-color offset printing unit;

[0184] a matching module 22 for matching the rotation phase of the printing cylinder with a preset reference phase in real time to generate a first dynamic compensation signal;

[0185] A generating module 23 is configured to generate a second dynamic compensation signal according to a degree of deviation between the tension fluctuation of the substrate and a preset tension range;

[0186] The analysis module 24 is configured to, when detecting that the periodic offset of the color mark exceeds a preset tolerance range, superimpose and analyze the periodic offset with the first dynamic compensation signal and the second dynamic compensation signal, and generate a composite compensation instruction according to the analysis result.

[0187] Figure 2 The four-color offset printing press registration error dynamic compensation system based on multi-sensor fusion can be performed Figure 1The implementation principles and technical effects of the multi-sensor fusion-based dynamic compensation method for a four-color offset printing press are not detailed here. The specific manner in which each module and unit performs operations in the multi-sensor fusion-based dynamic compensation system for a four-color offset printing press in the aforementioned embodiment has been described in detail in the related embodiments and will not be further elaborated here.

[0188] In one possible design, Figure 2 The four-color offset printing press registration error dynamic compensation system based on multi-sensor fusion of the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0189] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0190] The processing component 32 is used for the above Figure 1 The embodiment provides a dynamic compensation method for registration error of a four-color offset printing press based on multi-sensor fusion.

[0191] The processing component 32 may include one or more processors to execute computer instructions to perform all or part of the steps in the above method. Of course, the processing component may also be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above method.

[0192] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0193] Of course, a computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0194] The input / output interface provides an interface between the processing component and the peripheral interface module, which can be an output device, an input device, etc.

[0195] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0196] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0197] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The illustrated embodiment is a dynamic compensation method for registration error of a four-color offset printing press based on multi-sensor fusion.

[0198] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0199] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0200] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiments.

[0201] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A dynamic compensation method for registration error of a four-color offset printing press based on multi-sensor fusion, characterized in that: include: Collect the rotation phase of the printing cylinder, the tension fluctuation value of the substrate, and the periodic offset of the color mark generated during the operation of the multi-color offset printing unit; Matching the rotation phase of the printing cylinder with a preset reference phase in real time to generate a first dynamic compensation signal; generating a second dynamic compensation signal according to a degree of deviation between the tension fluctuation of the substrate and a preset tension range; When it is detected that the periodic offset of the color mark exceeds a preset tolerance range, the periodic offset is superimposed and analyzed with the first dynamic compensation signal and the second dynamic compensation signal, and a composite compensation instruction is generated according to the analysis result.

2. The method according to claim 1, characterized in that Matching the rotation phase of the printing cylinder with a preset reference phase in real time to generate a first dynamic compensation signal includes: Calculating a dynamic phase difference between the rotation phase of the printing cylinder and a preset reference phase; Using a preset sliding window to extract the gradient characteristics of the dynamic phase difference in the current printing cycle; Determining an increasing weight of the compensation amplitude as the phase difference increases based on a nonlinear relationship model between the rate of change of the gradient feature and a preset phase compensation curve; Performing a product operation on the dynamic phase difference and the incremental weight to generate an initial compensation amount; The initial compensation amount is input into a pre-trained phase smoothing filter, and the high-frequency oscillation component is suppressed according to the cumulative trend of the cylinder angle deviation in adjacent printing cycles, and a first dynamic compensation signal is output.

3. The method according to claim 2, characterized in that Calculating a dynamic phase difference between the rotation phase of the printing cylinder and a preset reference phase includes: monitoring a rotational position signal of the printing cylinder and decomposing the rotational position signal into an instantaneous deviation between a current phase angle and a reference phase angle; Performing multi-cycle rolling sampling on the instantaneous deviation, and quantifying the instantaneous fluctuation value of the phase jump by a discrete integrator with a preset window length being an integer multiple of the circumference of the printing cylinder; Calculating the cumulative offset of the printing cylinder angle deviation according to the change direction and amplitude of the instantaneous fluctuation value during the continuous printing cycle; The instantaneous fluctuation value and the accumulated offset are vector-superimposed to generate a dynamic phase difference.

4. The method according to claim 1, wherein Generating a second dynamic compensation signal according to the degree of deviation between the tension fluctuation of the substrate and a preset tension range includes: Real-time monitoring of tension fluctuations in the substrate's running path, and performing mean filtering on the tension fluctuations within a preset time window to generate dynamic tension fluctuation values; Performing a difference operation on the dynamic tension fluctuation value and the upper and lower limit thresholds of the preset tension range to obtain an instantaneous deviation coefficient; Extracting a characteristic vector of the tension change direction according to the historical fluctuation trend of the dynamic tension fluctuation value, and generating a trend factor by rolling and accumulating the characteristic vector over multiple periods; Performing nonlinear mapping between the trend factor and a preset tension compensation sensitivity parameter to generate an elastic weight that varies with printing speed, wherein the elastic weight is fixed to a constant value in a low-speed registration stage; The instantaneous deviation coefficient and the elastic weight are weighted and superimposed to generate an initial tension compensation amount, and the initial tension compensation amount is input into a preconfigured tension compensation filter. The tension compensation filter is used to suppress low-frequency interference components according to the tension coupling relationship between adjacent printing units, and a second dynamic compensation signal is output.

5. The method according to claim 4, characterized in that The trend factor is nonlinearly mapped with a preset tension compensation sensitivity parameter to generate an elastic weight that varies with printing speed, including: Normalizing the trend factor to generate a standardized trend value, wherein a value range of the standardized trend value matches the physical dimension of the tension compensation sensitivity parameter; Determine the speed adjustment coefficient of tension compensation based on the proportional relationship between the current printing speed of the multi-color offset printing unit and the preset speed threshold; performing a product operation on the normalized trend value and the tension compensation sensitivity parameter to generate an initial mapping factor, and inputting the initial mapping factor into a preconfigured dynamic scaling module; The dynamic scaling module performs nonlinear scaling on the initial mapping factor according to the speed adjustment coefficient to generate an intermediate elastic coefficient; The intermediate elastic coefficient is superimposed on a preset elastic reference value to generate an elastic weight, wherein the output direction of the elastic weight is determined by the sign of the characteristic vector of the tension change direction.

6. The method according to claim 1, characterized in that When it is detected that the periodic offset of the color mark exceeds a preset tolerance range, the periodic offset is superimposed and analyzed with the first motion compensation signal and the second motion compensation signal, and a composite compensation instruction is generated according to the analysis result, including: Detecting the periodic offset of the color mark, and when the periodic offset exceeds the preset tolerance range, extracting the change trajectory of the periodic offset in the continuous printing cycle, and generating a deviation trend prediction curve by least squares fitting, wherein the deviation trend prediction curve includes a transverse deviation component and a longitudinal deviation component; decomposing the first dynamic compensation signal into a roller angle compensation component and decomposing the second dynamic compensation signal into a tension fluctuation compensation component; Establishing a three-dimensional compensation space coordinate system, wherein a first coordinate axis of the three-dimensional compensation space coordinate system represents the roller angle compensation component, a second coordinate axis represents the tension fluctuation compensation component, and a third coordinate axis represents the predicted offset of the offset trend prediction curve; In the three-dimensional compensation space coordinate system, calculating the vector angle between the roller angle compensation component and the tension fluctuation compensation component, and starting the collaborative compensation mode when the vector angle is less than a preset threshold; In the collaborative compensation mode, the roller angle compensation component, the tension fluctuation compensation component and the predicted offset are weightedly fused to generate an initial composite compensation vector; The initial composite compensation vectors are dynamically normalized to balance the contribution of the initial composite compensation vectors in the high-speed printing stage and the low-speed registration stage, and a composite compensation instruction is output.

7. The method according to claim 6, characterized in that In the collaborative compensation mode, the roller angle compensation component, the tension fluctuation compensation component, and the predicted offset are weightedly fused to generate an initial composite compensation vector, including: Obtaining a speed weight coefficient corresponding to the current printing speed from the operation control system of the multi-color offset printing unit; Calculating a synergy factor based on the magnitude of the vector angle; Decomposing the predicted offset into a transverse prediction component and a longitudinal prediction component, and establishing mapping relationships with the tension fluctuation compensation component and the roller angle compensation component respectively; A dynamic weighting algorithm is used to calculate the real-time fusion weights of the roller angle compensation component, the tension fluctuation compensation component, the transverse prediction component, and the longitudinal prediction component by taking the speed weight coefficient and the synergy factor as input parameters; performing a weighted sum operation on the roller angle compensation component, the tension fluctuation compensation component, the transverse prediction component, and the longitudinal prediction component according to the real-time fusion weight to generate a three-dimensional compensation vector; Amplitude constraint processing is performed on the three-dimensional compensation vector, and when the component values of the three-dimensional compensation vector after constraint do not exceed a preset maximum compensation limit, an initial composite compensation vector is generated.

8. A dynamic compensation system for registration error of a four-color offset printing press based on multi-sensor fusion, characterized in that: include: The acquisition module is used to collect the rotation phase of the printing cylinder, the tension fluctuation value of the printing material, and the periodic offset of the color mark generated during the operation of the multi-color offset printing unit; a matching module, configured to match the rotation phase of the printing cylinder with a preset reference phase in real time to generate a first dynamic compensation signal; a generating module, configured to generate a second dynamic compensation signal according to a degree of deviation between the tension fluctuation of the substrate and a preset tension range; The analysis module is used to, when detecting that the periodic offset of the color mark exceeds a preset tolerance range, superimpose and analyze the periodic offset with the first dynamic compensation signal and the second dynamic compensation signal, and generate a composite compensation instruction according to the analysis result.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the dynamic compensation method for registration error of a four-color offset printing machine based on multi-sensor fusion as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that A computer program is stored, and when the computer program is executed by a computer, the method for dynamic compensation of registration error of a four-color offset printing press based on multi-sensor fusion as claimed in any one of claims 1 to 7 is implemented.

Images