System and method for predicting inoperative inkjets within printheads in inkjet printer

Abstract

To provide a new method of operating an inkjet printer while predicting the occurrences of inoperative inkjets to determine when printhead purging should be performed before image quality is adversely impacted.SOLUTION: A method of operating an inkjet printer indicates a need for a remedial printhead operation by predicting the number of inoperative inkjets and the locations of the inoperative inkjets in at least one printhead in the inkjet printer at a predetermined time. The prediction is made using Markov chain Monte Carlo models that correspond to different ranges of area coverage density for inkjet areas of a printhead.SELECTED DRAWING: Figure 4A

Description

[Technical field]

[0001] The present disclosure relates to printheads that eject liquid ink to form ink images on a substrate as the substrate passes the printhead, and more particularly to a method for predicting the occurrence of inoperable inkjets in such printheads. [Background technology]

[0002] Inkjet printers eject liquid ink droplets from a printhead to form ink images on a receiving surface that passes through the printer. The printhead includes a number of inkjets arranged in some type of array. Each inkjet has a thermal or piezoelectric actuator coupled to a printhead driver. The printhead controller generates firing signals corresponding to ink image content data for creating an ink image on a medium that passes through the printer. The actuators in the printhead are positioned relative to ink chambers in the printhead, and when the actuators respond to the firing signals, the actuators expand into the ink chambers to eject ink droplets onto the passing medium to create an ink image that corresponds to the ink image content data used to generate the firing signals.

[0003] Inkjets, particularly those in printheads that eject water-based inks, need to be fired periodically to help prevent the ink from drying out in the nozzles formed in the faceplate of the printhead. If the viscosity of the ink increases too much, the probability of inkjet failure increases substantially. During printing of a print job, test pattern images are printed on the sheet at predetermined intervals to evaluate the operational status of the inkjets. Optical sensors generate digital image data of these test pattern images, which is analyzed by the printer controller to determine which inkjets, if any, were actually operated to eject ink in the test pattern, and if the inkjets ejected ink droplets, whether the ejected droplets had the proper mass and landed where the droplets were supposed to land. Any inkjet nozzle that does not eject the ink droplets it is supposed to eject, or that ejects droplets that do not have the correct mass, or that land in the wrong location, is referred to herein as an inoperable inkjet. The controller stores data in a database operatively connected to the controller that identifies the inoperable inkjets in each printhead. These sheets with the test pattern printed on them are sometimes called run-time missing inkjet (RTMJ) sheets, and these sheets are discarded from the print job's output.

[0004] Inoperable inkjets can cause streaks in the ink images produced by an inkjet printer. The number of inoperable inkjets in a printhead typically increases over time, and the printhead must be purged, at some repetitive interval, to recover the inoperable inkjets in order to maintain an adequate level of quality in the ink images. Methods of detecting inoperable inkjets from images of test patterns printed on RTMJ sheets during a print job are time consuming, waste ink, and affect the overall production rate and cost of the inkjet printer. It would be beneficial to be able to predict the occurrence of inoperable inkjets without relying on printing test patterns on RTMJ sheets and analyzing image data of the test patterns on the RTMJ sheets. Summary of the Invention

[0005] A new method of operating an inkjet printer predicts the occurrence of inoperable inkjets to determine when a printhead purge should be performed before image quality is adversely affected. The method includes predicting a number of inoperable inkjets and locations of the inoperable inkjets in at least one printhead in the inkjet printer at a given time, and generating a signal indicating that at least one printhead requires corrective action when the number of inoperable inkjets exceeds a given threshold or the locations of the inoperable inkjets prevent implementation of inoperable inkjet compensation.

[0006] A new inkjet printer predicts the occurrence of inoperable inkjets to determine when to perform a printhead purge before image quality is adversely affected. The inkjet printer includes at least one printhead having a plurality of inkjets and a controller operatively connected to the printhead. The controller is configured to predict a number of inoperable inkjets and locations of the inoperable inkjets in at least one printhead in the inkjet printer at a given time, and to generate a signal indicating that at least one printhead requires corrective action when the number of inoperable inkjets exceeds a predetermined threshold or the locations of the inoperable inkjets prevent implementation of inoperable inkjet compensation. [Brief description of the drawings]

[0007] The above aspects and other features of operating an inkjet printer to predict the occurrence of an inoperable inkjet so that a printhead purge can be performed before image quality is adversely affected are explained in the following description in conjunction with the accompanying drawings. [Figure 1A] An inkjet printer is described that uses an inoperable inkjet compensation scheme and the number of inoperable inkjets and the locations of the inoperable inkjets predicted by a Markov chain Monte Carlo model to determine when corrective printhead maintenance should be performed before image quality is adversely affected. [Figure 1B] FIG. 1B is a diagram of the print zone in the printer of FIG. 1A. [Diagram 2] 1 depicts the distribution of inoperable inkjets among three printheads of a printhead module before and after a print job. [Diagram 3] 1 is a graph of the number of inoperable inkjets occurring at four different area coverage percentage ranges. [Figure 4A]1 illustrates a process flow for generating a model for predicting inoperable inkjets in a printhead over time, and a process flow for using the model in a printer. [Figure 4B] 4B is a table showing different area coverage density ranges for each print head shown in FIG. 4A. [Diagram 5] 1 depicts a first-order Markov chain Monte Carlo (MCMC) model that was used to generate transition probabilities between two states of the inkjet: inoperative and operational. [Figure 6] FIG. 5 illustrates two example sampling strategies that can be used to realize online prediction of the model shown in FIG. [Figure 7A] FIG. 7 depicts predicted results of inoperative inkjet counts using the two sampling strategies shown in FIG. 6. [Figure 7B] 1 is a graph comparing the results of two sampling strategies for different settings. [Figure 8A] 6 depicts a strategy for modeling spikes corresponding to print job interruptions in the standard model of FIG. 5. [Figure 8B] 6 is a graph showing the results of an experiment used to determine when the effect of the spikes disappears from the standard model of FIG. 5. [Figure 9A] 8A depicts a plot comparing the results of using only the standard model of FIG. 5 prediction with ground truth and the results of combining the standard model of FIG. 5 with the spike modeling of FIG. 8A against the ground truth. [Figure 9B] FIG. 9B is a graph showing the mean absolute prediction error (MAE) for the model results shown in FIG. 9A. [Figure 10] 1 shows the equation used to predict whether a grid has an inoperable inkjet within an m×m grid, where m is a tunable parameter, and a comparison of the predicted map to a ground truth map that identifies grids that have inoperable inkjets. [Figure 11]13 illustrates how evaluation scores are generated for printheads that jet different colors of ink at a given resolution at each predicted time during a print job. [Figure 12A] 13 shows a graph of the average F1 score over prediction time. [Figure 12B] 1 shows a graph of mean average precision (mAP) measurements over prediction time. [Figure 13] FIG. 1B is a flow diagram of a process used by the controller of the inkjet printer of FIG. 1A to predict the occurrence of an inoperable inkjet, determine a compensation scheme, and determine when corrective printhead maintenance should be performed before image quality is adversely affected. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] For a general understanding of the environment of the systems and methods, and details of the systems and methods disclosed herein, reference is made to the drawings, in which like reference numbers are used throughout the drawings to designate like elements. As used herein, the term "inkjet printer" encompasses any device that produces ink images on a medium by operating inkjets in a printhead to eject ink droplets toward a medium passing the printhead. As used herein, the term "process direction" refers to the direction of movement of the medium in which the ink image is formed, and the term "cross-process direction" is the direction substantially perpendicular to the process direction along the surface of the medium.

[0009] The printer and method described below use machine learning techniques to develop a spatio-temporal model to predict when and where an inoperable inkjet is likely to occur. A successful predictive system helps the inkjet printer's controller to operate the inkjet printer more intelligently during a customer job. Empirical digital image data of previously printed images and analysis of that digital image data to identify inoperable inkjet suggests that the distribution of inoperable inkjet in a printhead varies with respect to the ink color jetted by the printhead and the ink coverage area density in the image printed by the printhead. In addition, these data indicate that inkjets in the vicinity of an inoperable inkjet are more likely to become inoperable before any corrective action is taken. These propositions were verified by correlating the identified inoperable inkjet with typical customer job parameters as a function of time. The customer job parameters include, but are not limited to, image characteristics such as whether the printed portion of the image was solid, text, office graphics, blank, etc. FIG. 2 illustrates such a visualization of an inoperable inkjet in a black ink jetting printhead of a printhead module during the printing process in the form of a printhead map. The legend to the right of the printhead maps shows symbols indicating inoperable inkjets, operable inkjets, and no inkjets on the printhead faceplates. The three printheads are arranged in a printhead module as depicted to the right of the legend. The leftmost three printhead maps depict the locations of the inoperable inkjets on the three printhead faceplates at the start of a print job, and the rightmost three printhead maps depict the locations of the inoperable inkjets on the three printhead faceplates at the end of a print job.

[0010] 3 is a graph of experimental data showing the number of inoperable inkjets as a function of time for four different levels of ink area coverage density: area coverage density from 0% up to 25%, area coverage density from 25% up to 50%, area coverage density from 50% up to 75%, and area coverage density from 75% to 100%. The graph shows that inkjets used to print lower density areas have a higher number of inoperable inkjets because the inkjets are used less frequently. These blocks for quantizing printhead coverage areas into sequential areas, and the construction of a predictive model for each block, are merely exemplary, as other blocks with corresponding predictive models are possible.

[0011] These graphs show that the occurrence of inoperable inkjets in a printhead can be modeled using probabilistic and stochastic methods. The system and method described below models the evolution of the occurrence of inoperable inkjets in a printhead during a print job and predicts the state of the inkjets at future times, i.e., operational or inoperable, at both the printhead level and the nozzle level. At the printhead level, the task of predicting the number of inoperable inkjets over time is based on the distribution of ink area coverage density produced by each printhead in the printed image. At the nozzle level, the likelihood of individual nozzles transitioning from operational to inoperable, as well as the nozzles in a small neighborhood around each nozzle, is predicted using a model developed using digital image data of media printed during previously executed print jobs on the inkjet printer. This digital image data of previously printed media is generated by an optical system used to analyze test patterns printed on RTMJ sheets. An online learning system or model has been developed that predicts the number of inoperable inkjets at future times during a print job based on the inoperable inkjets data determined from this digital image data. This model is used during the printing process by retraining the model with the latest area coverage density data derived from the image content data used to operate the print head to better adapt to changes that occur in the inkjet transitions.

[0012] FIG. 4A illustrates an overall training and inference pipeline 400 for generating inoperative inkjet predictive models using an incoming stream of area coverage density data derived from image content data used to operate the inkjets in a printhead. A predictive model is used for each ink color and different area coverage density ranges for each printhead, as shown in table 404 in FIG. 4B. Process 408 in FIG. 4A illustrates that as new image content data is evaluated, every inkjet is mapped to its corresponding model based on a calculation of the average area coverage density printed by the inkjet since the last prediction. Also, the new area coverage density data is added to the corresponding model, and the predictive model for the corresponding model is updated. As used herein, the term "predictive model" refers to a number of programmed instructions that, when executed, identify the operating state of each inkjet in a region of the printhead using the previous operating state of the inkjets in that region.

[0013] 5 shows the predictive model used to predict the transition of inkjet states between operational and inoperable. This is a first-order Markov Chain Monte Carlo (MCMC) method that generates transition probabilities between the two states inoperable and operational. In the model, a "0" represents an inkjet that is operational and a "1" represents an inkjet that is inoperable, which may be referred to herein as a missing inkjet or MJ. The transition probabilities between the two states are parameters in the model. The symbols are used to represent the transition matrix, where:

[0014]

number

[0015]

number

[0016]

number

[0017] The plot on the left of FIG. 7A shows the prediction results of inoperable inkjet count using two sampling strategies. At K=5, the mean absolute error (MAE) is about 7 inoperable inkjet and the percentage error (MAPE) is about 10%. In general, the prediction tracks the ground truth well as the printing time increases. The graph in FIG. 7B shows the comparison of the two sampling strategies and the results of the average MAE with different settings and varying K. The results show that the single sampling strategy achieves a lower MAE and higher accuracy. Since the single sampling strategy is also easier to schedule during the printing process and requires about half the input data than the double sampling strategy, the single sampling strategy is used in the prediction model to achieve improved ink usage and time consumption. This single sampling strategy is used in the prediction model described in the remainder of this specification, and this model is referred to as the "standard MJ" model.

[0018] The Standard MJ model is valid as long as the printer is operational. However, unscheduled print interruptions occur. Print interruptions, such as paper jams, are often unavoidable during the printing process and affect the inkjet status state transitions. Inkjet status state transitions that occur after a print interruption do not follow the previous transition behavior, and inoperable inkjet counts often increase after a print interruption. Thus, spikes occur in the time series data, and these spikes require adjustment of the Standard MJ model. A strategy for modeling spikes in the Standard MJ model is shown in FIG. 8A. When an unscheduled interruption occurs, an additional spike model is updated and used to predict inoperable inkjets independently of the Standard MJ model. Immediately after the print interruption, inoperable inkjet status data is collected at a time after the interruption and at time +1. The transitions between the two states are used to update the spike model. The spike model is used to predict the subsequent time (defined as the data collection time affected by the spike). After that time, use of the Standard MJ resumes. The graph in FIG. 8B shows the results of the experiment used to determine the value of . The graph shows that there is a minimum when the prediction error is equal to 2. This error minimization indicates that the number of times affected by a print interruption is approximately 2.

[0019] The plot in FIG. 9A compares the results of using only the standard MJ model prediction with ground truth (GT) with the results of combining the standard MJ model with spike data modeling against the ground truth. In the graph in FIG. 9A, E is the average MAE, while E(%) is the average MAPE. This comparison shows that the simultaneous addition of spike modeling in the standard MJ model significantly reduces the prediction error. The best prediction results were achieved using a single sampling strategy and additional spike modeling, as shown by the spike line in the graph in FIG. 9B. In FIG. 9B, the mean absolute prediction error (MAE) is shown for different values ​​of K. In this graph, a value of K=7 keeps the prediction error at about 5% in MAE. These results show that the standard MJ model using additional spike modeling can replace 6 out of 7 (about 85%) of the conventional inoperable inkjet detections made by analyzing test patterns on RTMJ sheets with an MAE of about 5.

[0020] The model described thus far predicts the likelihood of inkjet counts during a print job. Such predictions help the printer schedule corresponding actions to prevent the appearance of streaks in the printed image and ensure proper image quality. Identifying which inkjets will be inoperable during a print job is equally important, as neighboring inkjets can be used to compensate for the lack of ink that would have been jetted by the inoperable inkjet. Identifying which specific inkjets will be inoperable is a highly probabilistic process, and the identification is difficult to predict with a high degree of certainty at the inkjet level. An alternative goal is to localize areas of the printhead that are prone to inoperable inkjets.

[0021] The MCMC model described above is capable of predicting inoperable inkjet counts during a print job. To extend this model so that it can predict printhead regions where inkjets will be inoperable, the model is modified to consider the probability of an inkjet being inoperable with respect to different area coverage densities. Four types of area coverage (AC) densities are defined in the range of 0-100% AC. For each inkjet in the ith row and jth column of the printhead, the area coverage for future prints is calculated and the coverage density for the inkjet is mapped to its corresponding AC density type. The corresponding transition probabilities are

[0022]

number

[0023]

number

[0024]

number

[0025] This approach provides a probability that an inkjet will become inoperable for each type of area coverage density, and the probabilities are mapped to each location (,) on the printhead. In addition, the probability of each inkjet transitioning to inoperable also depends on the state of neighboring inkjets. Thus, the printhead is divided into a grid having a size of: As used herein, the term "grid" or "grid area" or "area of ​​a grid" refers to an arrangement of a predetermined number of inkjets around a given inkjet location in the printhead. Within each grid, the probability of at least one inkjet transitioning to inoperable is calculated using the following formula:

[0026]

number

[0027] A prediction of a grid having inoperable inkjets is generated using a 0.5 threshold applied to the probability as shown in the equation of FIG. 10, although other thresholds may be used. In this manner, a prediction of inoperable inkjets occurring within an m×m grid (m is an adjustable parameter) may be generated. To evaluate the performance of the modified model, the prediction map of the modified model is compared to a ground truth map such as that shown in FIG. 10. Because the prediction map is generated on a lower resolution printhead map, the resolution of the ground truth map is also reduced to determine whether there is at least one inoperable inkjet within the grid. In addition, different sized grids may be used to generate maps of different resolutions, and a sliding window may also be used with different sized grids on the original map to compute the prediction and ground truth maps for comparison.

[0028] The inoperable inkjets are sparse in the printhead map since there are only a few dozen or a few hundred inoperable inkjets in an inkjet printhead having 16,632 inkjets. Because the data is imbalanced between inoperable and operational inkjets, the F1 score, defined by the following formula, is used to evaluate the performance of the predictions:

[0029]

number

[0030] FIG. 11 shows the F1 scores for four printheads (CMYK) with original resolution at each prediction time during the print job. The left side of the figure is a plot showing the results for K=1, and the right side of the figure is a plot showing the results for K=5. The vertical lines in the background in the plots are times where the ground truth is captured in the data and the F1 scores are not applicable. In addition, since at the start of the print job there are no inoperable inkjets in the printhead map and the F1 scores are not applicable at this time point, the F1 scores are labeled as 0 at these time points in the plots. From a visual comparison of these plots, the predictions in black and magenta achieve higher F1 scores on average, implying that the location of the inoperable inkjets may be more dependent on the area coverage density for these two colors of the printhead. In addition, from this comparison, a larger K value negatively impacts the model's performance. To optimize the value of K, the average F1 scores are shown in FIG. 12A over the prediction times, and the mean average precision (mAP) is shown in FIG. 12B over the prediction times. Different lines show the performance of the model at different resolutions. As expected, lower resolutions yield more accurate results. On the other hand, when K is greater than 3, both F1-score and mAP seem to decrease more rapidly. Therefore, the optimal K is 3. In conclusion, two-thirds of conventional inspection runs can be replaced by using the online MCMC model to locate inoperable inkjets on a 240 dpi printhead map with an F1-score of about 0.7.

[0031] The description of the MCMC model presented above demonstrates that prediction of inoperable inkjets at the printhead and nozzle levels is possible. The main factors the model predicts are the area ink coverage distribution in the printed image and the interaction of the inkjet with its neighboring inkjets. The model shows that its predictive capabilities are sufficient for the model to be used in inkjet printers to schedule inoperable inkjet detection and corrective actions when the predicted results indicate a negative impact on image quality. As mentioned above, the model is able to predict inoperable inkjets within a 5×5 neighborhood with an F1 score of 0.7, although other grid sizes are possible.

[0032] FIG. 1A depicts a high-speed color inkjet printer 10 configured with program instructions stored in a memory operatively connected to a controller 80 that, when executed, implements an MCMC model that predicts the number and location of inoperable inkjets in a printhead of the printer, so that the controller can determine whether image quality is impaired to the extent that corrective maintenance action is required. As used herein, the term "corrective action" refers to an action taken on a printhead that restores the inkjets in the printhead to an operable condition. As illustrated, the printer 10 is a printer that forms ink images directly on the surface of a media sheet retrieved from one of media sheet supplies S1 or S2, the sheet S being moved through the printer 10 by the controller 80 operating one or more actuators 40 that are operatively connected to rollers or at least one drive roller of a conveyor 52, the conveyor 52 comprising a portion of a media transport 42 that passes through a print zone PZ (shown in FIG. 1B) of the printer. In one embodiment, each printhead module has only one printhead having a width corresponding to the width of the widest media in the cross-process direction that can be printed by the printer. In other embodiments, the printhead module has multiple printheads, each printhead having a width less than the width of the widest media in the cross-process direction that the printer can print. In these modules, the printheads are arranged in a staggered printhead array that allows media wider than a single printhead to be printed. In addition, printheads within or between modules can also be combined such that the density of droplets jetted in the cross-process direction by a printhead can be greater than the minimum spacing between inkjets in a printhead in the cross-process direction. Although the printer 10 is depicted with only two media sheet supplies, the printer can be configured with three or more sheet supplies, each containing a different type or size of media.

[0033] The print zone PZ in the printer 10 of FIG. 1A is shown in FIG. 1B. The print zone PZ has a length in the process direction equal to the distance from the first inkjet that the sheet passes in the process direction to the last inkjet that the sheet passes in the process direction, and a width that is the maximum distance between the outermost inkjet on either side of the print zone that are directly opposite each other in the cross-process direction. Each printhead module 34A, 34B, 34C, and 34D shown in FIG. 1B has three printheads 204 mounted on one of the printhead carrier plates 316A, 316B, 316C, and 316D, respectively. The printheads of each module jet ink of the same color, which means that in the printer 10, the printhead of module 34A jets cyan ink, the printhead of module 34BA jets magenta ink, the printhead of module 34C jets yellow ink, and the printhead of module 34D jets black ink. The printhead 204 on the left side of the module in the process direction is referred to herein as the inner printhead, the printhead 204 on the right side of the module in the process direction is referred to herein as the outer printhead, and the printhead 204 between the inner and outer printheads is referred to as the central printhead.

[0034] As shown in FIG. 1A, the printed image passes under the image dryer 30 after the ink image is printed on the sheet S. The image dryer 30 may include an infrared heater, a heated air blower, an air return, or a combination of these components to heat the ink image and at least partially fix the image to the web. The infrared heater applies infrared heat to the printed image on the surface of the web to evaporate water or solvent in the ink. The heated air blower uses a fan, or other pressurized air source, to direct heated air over the ink to supplement the evaporation of water or solvent from the ink. Air is then collected and exhausted by the air return to reduce interference of the dryer airflow with other components in the printer.

[0035] The dual path 72 is provided to receive the sheet from the transport system 42 after the substrate has been printed and to move the sheet by rotation of rollers in a direction opposite to the direction of movement past the print head. At a position 76 in the dual path 72, the substrate can be inverted so that it can join the job stream being carried by the media transport system 42. The controller 80 is configured to selectively flip the sheet. That is, the controller 80 can operate the actuator to invert the sheet so that the reverse side of the sheet can be printed, or the controller 80 can operate the actuator so that the sheet is returned to the transport path without inverting the sheet so that the printed side of the sheet can be printed again. Access to the dual path 72 is provided by movement of a pivot member 88. The rotation of the pivot member 88 is controlled by the controller 80, which selectively operates the actuator 40 operably connected to the pivot member 88. When the pivot member 88 is rotated counterclockwise as shown in FIG. 1A, the substrate from the media transport 42 is diverted to the dual path 72. Rotating pivot member 88 in a clockwise direction from the turnaround position closes access to dual path 72 so that substrates on the media transport move to receptacle 56. Another pivot member 86 is positioned between location 76 on dual path 72 and media transport 42. When controller 80 operates an actuator to rotate pivot member 86 in a counterclockwise direction, substrates from dual path 72 join the job stream on media transport 42. Rotating pivot member 86 in a clockwise direction closes dual path access to media transport 42.

[0036] As further shown in FIG. 1A, the printed sheets of media S that are not diverted to the dual path 72 are conveyed by the media transport to the sheet bin 56 where they are collected. Before the printed sheets reach the bin 56, they pass an optical sensor 84. The optical sensor 84 generates image data of the printed sheets, which is analyzed by the controller 80. The controller 80 is configured to identify inoperable inkjets in the printed image of the test pattern on the RTMJ sheet inserted in the print job and generate a printhead map for each printhead in the print zone. The RTMJ sheet is discarded from the output of the print job. To identify inoperable inkjets, the test pattern image is analyzed by the controller 80 to determine which inkjets, if any, that were operated to eject ink in the test pattern actually ejected ink, and, if the inkjets ejected ink droplets, whether the ink droplets landed at their intended locations with the proper mass. Any inkjet that does not eject the ink droplets it is supposed to eject, or that ejects droplets that do not have the correct mass, or that land in an incorrect location is identified as an inoperable inkjet. The controller 80 uses the identified inoperable inkjet to generate a printhead map and generates an index using the printhead map. The printhead map index is compared to the indexes of the clusters stored in a database 92 operatively connected to the controller. The highest similarity score between the printhead index and one of the indexes stored in the dictionary 212 identifies the cluster that is most similar to the generated printhead map. The known causes and solutions stored in association with the identified index from the dictionary are used to diagnose problems in the printer 10, as described in more detail below. The optical sensor can be a digital camera, an array of LEDs, and a photodetector, or other device configured to generate digital image data of a passing surface.As previously mentioned, the media transport also includes a dual path that can invert a sheet and return it to the transport before the printhead module so that the other side of the sheet can be printed. Although Figure 1A shows printed sheets being collected in a sheet bin, the sheets can be directed to other processing stations (not shown) that perform operations such as folding, collating, stitching, and stapling the media sheets.

[0037] The operation and control of the various subsystems, components and functions of the machine or printer 10 is performed with the aid of a controller or electronic subsystem (ESS) 80. The ESS or controller 80' is operatively connected to the printhead modules 34A-34D (and thus the printheads), the actuators 40 and the components of the dryer 30. The ESS or controller 80 is, for example, a self-contained computer having a central processor unit (CPU) with electronic data storage and a display or user interface (UI) 50. The ESS or controller 80 includes, for example, sensor input and control circuits, as well as pixel arrangement and control circuits. In addition, the CPU reads, captures, prepares and manages the flow of image data between an image input source, such as a scanning system or an on-line or workstation connection (not shown), and the printhead modules 34A-34D. The ESS or controller 80 is thus the main multitasking processor for operating and controlling all other machine subsystems and functions, including the printing process.

[0038] The controller 80 can be implemented using a general-purpose or dedicated programmable processor that executes program instructions. Instructions and data required to perform the programmed functions can be stored in a memory associated with the processor or the controller. The processors, their memories, and interface circuits configure the controller to perform the operations described below. These components can be provided on a printed circuit card or as circuits in an application specific integrated circuit (ASIC). Each of the circuits can be implemented in a separate processor, or multiple circuits can be implemented on the same processor. Alternatively, the circuits can be implemented with individual components or circuits provided in a very large scale integrated (VLSI). The circuits described herein can also be implemented with a combination of processors, ASICs, individual components, or VLSI circuits.

[0039] During operation, ink image content data for the ink image to be generated is sent to the controller 80 from either the scanning system or an on-line or workstation connection. The ink image content data is processed to generate inkjet ejector firing signals that are sent to the printheads in modules 34A-34D. Along with the image content data, the controller receives print job parameters that identify the weight of the media, the dimensions of the media, the print speed, the type of media, the ink area coverage to be generated on each side of each sheet, the location of the image to be generated on each side of each sheet, the color of the media, the media fiber orientation for fibrous media, the temperature and humidity of the print zone, the moisture content of the media, and the manufacturer of the media. As used in this document, the term "print job parameters" refers to non-image content data for a print job and the term "ink image content data" refers to digital data that identifies the color and amount of each pixel that forms the ink image to be printed on the media sheet.

[0040] A process 1300 for identifying the number and location of inoperable inkjets in a printhead of a printer using a predictive MCMC model is shown in FIG. 13. In the description of a process, a statement that the process performs a certain task or function refers to a controller or general-purpose processor executing program instructions stored in a non-transitory computer-readable storage medium operatively connected to the controller or processor to manipulate data or operate one or more components in the printer to perform the task or function. The controller 80 described above can be such a controller or processor. Alternatively, a controller can be implemented with two or more processors and associated circuits and components, each configured to form one or more tasks or functions described herein. In addition, the steps of the method can be performed in any feasible chronological order, regardless of the order shown in the figures or the order in which the process is described.

[0041] The process 1300 begins by receiving image content data for a print job (block 1304). At a predetermined prediction time (block 1308), the process identifies area coverage densities for a predetermined number of grids in each print head from the image content data that will be used to operate the inkjets in the print head to form an ink image on a medium from a previous time in the print job to the predetermined prediction time (block 1312). The identified area coverage densities for each grid are used to select a predictive model (block 1316). The selected predictive model uses the operational state of each inkjet in the grid at a previous time in the print job to identify the operational state of each inkjet in the grid at the predetermined prediction time and the predicted location of each inoperable inkjet (block 1320). The predicted number and location of inoperable inkjets in the print head are stored in an inoperable inkjet database (1324). If the number of inoperable inkjets exceeds a predetermined threshold (1328), a signal is generated that corrective print head maintenance, such as a purge, is required and printing is stopped (block 1332). In addition, if the location of the inoperable inkjet prevents implementation of the inoperable inkjet compensation scheme (block 1336), a signal is generated that corrective printhead maintenance, such as purging, is required and printing is stopped (block 1332). The process determines if the print job is finished (block 1340), and if so, the process stops. If not, the process continues until the next predicted time occurs (block 1308). As used herein, the term "inoperable inkjet compensation scheme" refers to a technique used to distribute ink droplet ejections from an inoperable inkjet to operational inkjets adjacent to the inoperable inkjet.

[0042] It will be appreciated that various of the above-disclosed and other features and functions, or alternatives thereof, may be desirably combined into many other different systems or applications. Various presently unforeseen substitutions, modifications, variations, or improvements may subsequently occur to those skilled in the art which are also intended to be encompassed by the following claims.

Claims

1. A method for operating an inkjet printer, To predict the number of non-functioning inkjet printheads in at least one printhead within the inkjet printer and the location of the non-functioning inkjet printheads within a predetermined time, A method comprising: generating a signal indicating that at least one print head requires corrective action when the predicted number of malfunctioning inkjet prints exceeds a predetermined threshold, or when the location of the malfunctioning inkjet prints prevents the implementation of malfunctioning inkjet compensation.

2. Identifying the regional coverage density for each grid of the plurality of grids of the at least one print head, Using the identified regional coverage density for each grid, a prediction model is selected from multiple prediction models for each grid. The method according to claim 1, further comprising using the selected prediction model to predict the number of inoperable inkjet printheads and the locations of the inoperable inkjet printheads.

3. The method according to claim 2, wherein each grid has the same region.

4. The plurality of prediction models selected from the prediction models are: A predictive model for the identified area coverage density from 0 percent to 25 percent of the region of the grid, A predictive model for the identified area coverage density of 25 percent to 50 percent of the region of the grid, A predictive model for the identified area coverage density of 50 percent to 75 percent of the area in the grid, The method according to claim 3, further comprising: a predictive model for identified area coverage density from 75 percent to 100 percent of the area of ​​the grid.

5. The method according to claim 4, wherein each prediction model is a Markov chain Monte Carlo (MCMC) model.

6. The method according to claim 5, wherein each MCMC model is trained using digital image data of at least one ink image previously printed by the at least one print head.

7. The method according to claim 6, wherein each MCMC model uses a probability threshold to predict the number of non-functioning inkjet prints and the location of the non-functioning inkjet prints within the grid.

8. The method according to claim 7, wherein the probability threshold is 0.

5.

9. The method according to claim 8, wherein the area of ​​the grid corresponds to a 5x5 pattern of inkjet.

10. When the number of non-functioning inkjet printers in at least one print head exceeds the predetermined threshold, the operation of at least one print head is stopped. The method according to claim 1, further comprising:

11. It is an inkjet printer, A print head having multiple inkjet inks, A controller operably connected to the print head, The controller is equipped with, The number of inoperable inkjet printheads in at least one printhead within the inkjet printer and the location of the inoperable inkjet printheads are predicted within a predetermined time. An inkjet printer configured to generate a signal indicating that corrective action is needed when the predicted number of malfunctioning inkjet prints exceeds a predetermined threshold, or when the location of the malfunctioning inkjet prints prevents the implementation of malfunctioning inkjet compensation.

12. The aforementioned controller, Identify the area coverage density for each grid of the plurality of grids of the at least one print head, Using the identified regional coverage density for each grid, a prediction model is selected from multiple prediction models for each grid. The inkjet printer according to claim 11, further configured to use the selected prediction model to predict the number of inoperable inkjet printheads and the location of the inoperable inkjet printheads.

13. The inkjet printer according to claim 12, wherein each grid has the same area.

14. The aforementioned controller, A predictive model for the identified area coverage density from 0 percent to 25 percent of the region in the grid, A predictive model for the identified area coverage density from 25 percent to 50 percent of the area in the grid, A predictive model for the identified area coverage density from 50 percent to 75 percent of the area in the grid, and The inkjet printer according to claim 13, further configured to select a predictive model from among predictive models for identified area coverage densities from 75 percent to 100 percent of the area of ​​the grid.

15. The inkjet printer according to claim 14, wherein each prediction model is a Markov chain Monte Carlo (MCMC) model.

16. The inkjet printer according to claim 15, wherein each MCMC model is trained using digital image data of at least one ink image previously printed by the at least one print head.

17. The inkjet printer according to claim 16, wherein each MCMC model uses a probability threshold to predict the number of non-functioning inkjets and the location of the non-functioning inkjets within the grid.

18. The inkjet printer according to claim 17, wherein the probability threshold is 0.

5.

19. The inkjet printer according to claim 18, wherein the area of ​​the grid corresponds to a 5x5 pattern of inkjet.

20. The aforementioned controller, The inkjet printer according to claim 11, further configured to stop the operation of the at least one print head when the number of inoperable inkjet particles in the at least one print head exceeds a predetermined threshold.

Images