Random offset compensation method and system for inkjet printhead clogged nozzle

Abstract

The present application relates to a kind of random offset compensation method and system of ink-jet printing nozzle hole blocking, first in the both ends of nozzle array reserve spare channel provides offset margin, when generating random offset amount each layer printing, the whole translation of working nozzle window is carried out, the same defect position of hole blocking is dispersed layer by layer to form mesh interlaced distribution, and macroscopic continuous crack is eliminated.Simultaneously, introduce the neighborhood compensation weight mechanism based on distance reciprocal attenuation, the gray scale driving information of hole blocking pixel is distributed according to weight to the normal nozzle of both sides, the ink drop is expanded and spread to the direction of hole blocking by increasing neighborhood driving pulse, and the active repair of material density in hole blocking area is realized in single layer, so that the defect position is dispersed between layers while reducing the severity of single layer defect, and the density of overall forming piece is improved.

Description

Technical Field

[0001] This invention relates to the field of inkjet printing control technology, and in particular to a method and system for compensating for random offset of inkjet printhead clogging. Background Technology

[0002] In the field of inkjet additive manufacturing using particulate inks, printhead arrays typically consist of hundreds to thousands of physical nozzles. Functional ink droplets are precisely deposited onto a powder bed or substrate layer through layer-by-layer scanning to construct three-dimensional parts. However, due to the presence of solid particles in the particulate ink, some nozzles inevitably become clogged and fail during long-cycle continuous printing operations. When a nozzle becomes clogged, a material deposition gap will continuously appear at the fixed physical location corresponding to it in the sub-scanning direction in each layer of printing. As this gap accumulates layer by layer, the point-like defect at this fixed location extends along the stacking direction, eventually forming a macroscopically visible continuous linear crack, severely weakening the structural density and mechanical strength of the formed part in that area.

[0003] To address the spatial continuity issue of the aforementioned hole-clogging defects, one feasible approach is to reserve spare channels at both ends of the nozzle array and apply a random offset to the actual working nozzle window in the sub-scanning direction during each layer's printing. This disperses the defect location of the same hole in different layers to different physical coordinates, eliminating the conditions for the formation of macroscopic continuous cracks. However, this random offset strategy only solves the spatial continuity problem of defects in the interlayer direction. Within each independent printing layer, the hole-clogging area still suffers from 100% material loss—that is, the grayscale driving information carried by the corresponding pixel is completely discarded and not received in any form by any adjacent physical channel. After multiple layers accumulate, the integral value of the material volume density in the area corresponding to the hole-clogging area remains systematically lower than the theoretical value, leading to a decrease in the local density of the molded part. Therefore, how to actively repair the local material density of the hole-clogging area within a single layer plane while maintaining the interlayer defect dispersion effect has become an urgent technical problem to be solved.

[0004] Therefore, an optimized random offset compensation scheme for inkjet printhead clogging is desired. Summary of the Invention

[0005] To address the aforementioned technical problems, this application is proposed. Embodiments of this application provide a method and system for compensating for random offset in inkjet printer nozzle clogging.

[0006] According to one aspect of this application, a random offset compensation method for inkjet printer nozzle clogging is provided, comprising: Step S1: Based on the nozzle array specification parameters including the total number of nozzles and the nozzle spacing, determine the nozzle droplet ejection status of each nozzle, including the nozzle number and its corresponding ejection amplitude, to obtain a nozzle blockage location identification map. Step S2: Based on the total number of nozzles and the preset number of spare nozzles at the top and bottom ends, the effective printing range of the nozzle array corresponding to the nozzle blockage location marking diagram is defined and the offset margin is estimated to obtain the effective working nozzle range and the allowable offset range of the sub-scanning direction. Step S3: Based on the current printing layer number, perform uniformly distributed random integer sampling within the allowable offset range in the sub-scanning direction to obtain the number of offset holes in the current layer; Step S4: The effective working nozzle range is translated as a whole using the number of offset holes in the current layer, and the mapping between pixels and nozzles is recalculated between the current layer slice image data and the translated nozzle range. The excitation state of the corresponding pixel of the blocked nozzle is set to skipped in combination with the hole blockage location identification map, and the nozzle excitation sequence after the current layer offset is generated. Step S5: Drive the nozzle to scan and print the current layer according to the nozzle excitation sequence after the current layer offset, and repeat steps S3 to S4 for each subsequent layer, so that the defect positions of the same hole are randomly dispersed in the layer-by-layer printing to form a mesh-like interwoven distribution.

[0007] According to another aspect of this application, a random offset compensation system for inkjet printhead clogging is provided, comprising: The nozzle status detection module is used to determine the nozzle droplet ejection status of each nozzle based on the nozzle array specification parameters including the total number of nozzles and the nozzle spacing, and to obtain a nozzle blockage location identification map. The offset margin calculation module is used to define the effective printing range and estimate the offset margin of the nozzle array corresponding to the hole blockage location marking diagram based on the total number of nozzles and the preset number of spare nozzles at the upper and lower ends, so as to obtain the effective working nozzle range and the allowable offset range of the sub-scanning direction. The random offset generation module is used to perform uniformly distributed random integer sampling within the allowable offset range in the sub-scanning direction based on the current printing layer number to obtain the number of offset holes in the current layer. The mapping reconstruction and sequence generation module is used to perform an overall translation of the effective working nozzle range using the number of offset holes in the current layer, and to recalculate the mapping between pixels and nozzles using the current layer slice image data and the translated nozzle range. It also combines the hole blockage location identification map to set the excitation state of the corresponding pixel of the blocked nozzle to skip, and generates the nozzle excitation sequence after the current layer offset. The layer-by-layer printing execution module is used to drive the nozzle to scan and print the current layer according to the nozzle excitation sequence after the current layer offset. It also repeats the random offset generation module to the mapping reconstruction and sequence generation module for each subsequent layer, so that the defect positions of the same hole are randomly dispersed in the layer-by-layer printing to form a mesh-like interwoven distribution.

[0008] Compared with existing technologies, this invention proposes a random offset compensation method for inkjet printer nozzle clogging. It achieves systematic suppression of nozzle clogging defects by constructing a two-dimensional collaborative compensation architecture of inter-layer random offset dispersion and intra-layer neighborhood gain compensation. First, the ejection state of each nozzle is detected, and a clogging location identification map is established. Spare channels are reserved at both ends of the nozzle array to provide offset margin. During each layer of printing, a random offset is generated based on the layer number to shift the effective working nozzle window as a whole. This causes the defect location of the same clogging nozzle to be randomly dispersed in a mesh-like distribution during layer-by-layer printing, eliminating macroscopic continuous cracks. Based on this, to address the issue of full material loss at the hole-blocking location within a single layer, a neighborhood compensation weight allocation mechanism based on inverse distance decay is introduced. This mechanism distributes the grayscale driving information carried by the hole-blocking pixel to the normal nozzles in the neighboring areas on both sides according to the decay weight. By increasing the driving pulse width of the neighboring nozzles, the ink droplets expand and spread towards the hole-blocking direction, thereby achieving active repair of local material density in the hole-blocking area within the single-layer plane. This reduces the severity of single-point defects in each layer while dispersing the spatial location of defects between layers, thus improving the overall density and mechanical strength of the molded part. Attached Figure Description

[0009] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0010] Figure 1 This is a flowchart of a random offset compensation method for inkjet printer nozzle clogging according to an embodiment of this application; Figure 2 This is a data flow diagram illustrating a random offset compensation method for inkjet printer nozzle clogging according to an embodiment of this application; Figure 3 This is a flowchart illustrating a random offset compensation method for inkjet printer nozzle clogging according to an embodiment of this application. The method involves defining the effective printing range and estimating the offset margin of the nozzle array corresponding to the clogging position identification diagram based on the total number of nozzles and the preset number of spare nozzles at the upper and lower ends, so as to obtain the effective working nozzle range and the allowable offset range of the sub-scanning direction. Figure 4 This is a flowchart illustrating a random offset compensation method for inkjet printer nozzle clogging according to an embodiment of this application, which involves uniformly distributing random integer sampling within the allowable offset range in the sub-scanning direction based on the current printing layer number to obtain the number of offset holes in the current layer. Figure 5The flowchart of a random offset compensation method for inkjet printer nozzle clogging according to an embodiment of this application is as follows: the effective working nozzle range is shifted as a whole by using the number of offset nozzles in the current layer, and the pixel-to-nozzle mapping is recalculated by the current layer slice image data and the shifted nozzle range. The excitation state of the corresponding pixel of the blocked nozzle is set to skipped in combination with the clogging position identification map, and the excitation sequence of the nozzle after the current layer offset is generated. Figure 6 This is a flowchart illustrating a random offset compensation method for inkjet printer nozzle clogging according to an embodiment of this application, which involves gain reconstruction of the pixel-nozzle matching mapping table based on neighborhood attenuation compensation to generate the nozzle excitation sequence after the current layer offset. Figure 7 This is a block diagram of a random offset compensation system for inkjet printer nozzle clogging according to an embodiment of this application. Detailed Implementation

[0011] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0012] In inkjet additive manufacturing with particulate ink, clogging of individual nozzles in the printhead creates continuous material deposition gaps at fixed positions in the sub-scanning direction. These gaps accumulate layer by layer, forming macroscopic continuous cracks that penetrate the stacking direction, severely weakening the structural density and mechanical strength of the molded part. While randomly shifting the defect locations layer by layer in the interlayer direction can eliminate the spatial continuity of the cracks, the clogging location within each independent layer still retains a complete material loss, resulting in a systematically low volumetric density in that region after multi-layer integration. Therefore, this application proposes a random offset compensation method for printhead clogging. This method, in the interlayer dimension, provides physical offset margins by reserving spare channels at the top and bottom of the nozzle array. During each layer printing, a random integer offset is generated within the allowable offset range based on the layer number. The effective working nozzle window is then translated as a whole, and the mapping relationship between pixels and nozzles is re-established. This allows the defect locations of the same clogging point in different layers to be randomly dispersed, forming a mesh-like interwoven distribution, fundamentally eliminating the conditions for the formation of macroscopic continuous cracks. In the intra-layer dimension, a neighborhood compensation weight allocation mechanism based on inverse distance decay is introduced. The grayscale driving information carried by the blocked pixel is quantitatively distributed to the normal nozzles in the neighboring areas on both sides according to the normalized weight. By increasing the driving pulse width of the neighboring channel, the ink droplet expands and spreads in the direction of the blocked hole, and the amplitude is limited and truncated with the hardware safety upper limit. The local material density of the blocked area is actively repaired in the single-layer plane, so that the single-layer missing rate is reduced from the full amount to an adjustable ratio controlled by the compensation gain coefficient, thereby improving the density and mechanical properties of the blocked area of ​​the molded part as a whole.

[0013] Figure 1 This is a flowchart of a random offset compensation method for inkjet printer nozzle clogging according to an embodiment of this application. Figure 2 This is a data flow diagram illustrating a random offset compensation method for inkjet printer nozzle clogging according to an embodiment of this application. Figure 1 and Figure 2 As shown, a random offset compensation method for inkjet printer nozzle clogging according to an embodiment of this application includes the following steps: S1, based on the nozzle array specification parameters including the total number of nozzles and the nozzle spacing, performing nozzle droplet ejection detection signals including each nozzle number and its corresponding ejection amplitude to determine the nozzle ejection state one by one to obtain a clogging position identification map; S2, based on the total number of nozzles and the preset number of spare nozzles at the upper and lower ends, performing effective printing interval delineation and offset margin estimation on the nozzle array corresponding to the clogging position identification map to obtain the effective working nozzle interval and the allowable offset range in the sub-scanning direction; S3, based on the current printing layer number, within the allowable offset range in the sub-scanning direction... S4. Uniformly distributed random integer sampling is performed within the offset range to obtain the number of offset holes in the current layer; S5. The effective working nozzle range is translated as a whole using the number of offset holes in the current layer, and the mapping between pixels and nozzles is recalculated between the current layer slice image data and the translated nozzle range. The excitation state of the corresponding pixel of the blocked nozzle is set to skipped in combination with the hole blockage location identification map, and the current layer offset nozzle excitation sequence is generated; S6. The nozzle is driven to perform current layer scanning and printing according to the current layer offset nozzle excitation sequence, and steps S3 to S4 are repeated for each subsequent layer, so that the defect position of the same blocked hole is randomly dispersed in the layer-by-layer printing to form a mesh-like interwoven distribution.

[0014] Specifically, in step S1, based on the printhead array specifications including the total number of nozzles and the nozzle spacing, the nozzle droplet ejection detection signal, which includes the nozzle number and its corresponding ejection amplitude, is used to determine the ejection status of each nozzle individually to obtain a clogging location identification map. It should be noted that, due to the randomness and concealment of nozzle clogging during inkjet additive manufacturing with particulate ink, if the ejection function status of each nozzle is not detected one by one and the physical location of the clogging nozzle is not accurately located before the printing job is executed, subsequent offset compensation will lack a clear target object and spatial reference. Therefore, the technical solution of this application first uses the printhead array specifications including the total number of nozzles and the nozzle spacing to determine the ejection status of each nozzle droplet ejection detection signal, which includes the nozzle number and its corresponding ejection amplitude, to obtain a clogging location identification map. Through the above processing, all malfunctioning clogging nozzles in the printhead can be completely identified in the form of numbers and physical coordinates, providing an accurate spatial distribution reference for subsequent offset margin calculation and excitation sequence generation.

[0015] More specifically, in a specific example of this application, the nozzle array specifications record the total number of nozzles N and the physical distance P between the centers of adjacent nozzles in the sub-scanning direction. The nozzle droplet ejection detection signal is multi-channel pulse response waveform data obtained by performing test pattern ejection and acquiring it hole by hole by optical sensor before the printing job starts.

[0016] In sub-step S11, the droplet ejection detection signal from the nozzles is subjected to multi-channel amplitude separation and sequence reconstruction according to the total number of nozzles N. Specifically, for the original detection waveform data of the N physical channels, the signal of each channel is separated in time according to the physical number index of the nozzle, and the maximum amplitude of the pulse response of each channel is extracted as the ejection response feature of that nozzle. Then, the amplitudes of the N channels are serialized and reconstructed according to the physical number order to form a 1×N ejection amplitude feature vector for each nozzle.

[0017] In sub-step S12, based on a preset normal injection judgment threshold, the injection amplitude feature vectors of each nozzle are compared with the injection amplitude threshold and classified according to their status. Specifically, each element in the injection amplitude feature vector of each nozzle is traversed one by one, and its amplitude is compared with the preset normal injection judgment threshold. If the response amplitude of the nozzle is greater than or equal to the threshold, its working status is determined to be normal and a classification label value of 1 is assigned. If the response amplitude is lower than the threshold, it is determined to be blocked and a classification label value of 0 is assigned. The classification label values ​​of all N nozzles are encapsulated into a nozzle status classification vector in the original number order.

[0018] In sub-step S13, the nozzle spacing P in the nozzle array specifications is used to perform physical coordinate mapping and map construction on the nozzles marked as blocked in the nozzle state classification vector. Specifically, all elements with a label value of 0 in the nozzle state classification vector are retrieved and their corresponding nozzle physical numbers are extracted. The physical number of each blocked nozzle is multiplied by the nozzle spacing P to obtain its absolute physical position coordinates in the sub-scanning direction. A one-to-one mapping relationship is established between the number of each blocked nozzle and its physical position coordinates to construct a blocked nozzle position identification map.

[0019] Specifically, in step S2, based on the total number of nozzles and the preset number of spare nozzles at the top and bottom, the effective printing range and offset margin are estimated for the nozzle array corresponding to the nozzle blockage location marker to obtain the effective working nozzle range and the allowable offset range in the sub-scanning direction. It should be noted that, given that the implementation of random offset compensation requires the printhead to have physical movement margin in the sub-scanning direction, if all nozzles are directly used for image printing, there will be no redundant channels available for offsetting, and the offset operation will cause the image edge area to exceed the nozzle coverage area, resulting in printing defects. Based on this, the technical solution of this application further determines the effective printing range and offset margin for the nozzle array corresponding to the nozzle blockage location marker based on the total number of nozzles and the preset number of spare nozzles at the top and bottom, to obtain the effective working nozzle range and the allowable offset range in the sub-scanning direction. Through the above processing, the effective working range carrying the image content and the spare range providing offset margin at both ends can be clearly defined in the nozzle array, providing a definite numerical boundary constraint for the subsequent generation of layer-by-layer random offsets.

[0020] Figure 3 This document describes a method for random offset compensation of inkjet printer nozzle clogging according to an embodiment of this application. It involves defining the effective printing range and estimating the offset margin of the nozzle array corresponding to the clogging location marker diagram based on the total number of nozzles and a preset number of spare nozzles at the upper and lower ends, to obtain the effective working nozzle range and the allowable offset range in the sub-scanning direction. Figure 3 As shown, step S2 includes: S21, setting the number of spare nozzles reserved at both ends according to the total number of nozzles and the equipment boundary tolerance requirements, and constructing the allowable offset range of the sub-scanning direction with the number of spare nozzles as the upper limit; S22, determining the number of effective working nozzles based on the total number of nozzles and the number of spare nozzles at both ends, and calculating the effective printing width and defining the effective working nozzle interval in combination with the nozzle spacing; S23, traversing the nozzle blockage position identification diagram for each blocked nozzle number, verifying whether it falls within the effective working nozzle interval to confirm the necessity of random offset compensation, and outputting the effective working nozzle interval, effective printing width, and allowable offset range of the sub-scanning direction.

[0021] In step S21, the number of spare nozzles reserved at both the top and bottom ends is set according to the total number of nozzles and the equipment boundary tolerance requirements, and the allowable offset range in the sub-scanning direction is constructed with the number of spare nozzles as the upper limit. It should be noted that since the physical implementation of random offset relies on the existence of redundant channels at both ends of the nozzle array that do not participate in image carrying, the number of spare nozzles directly determines the limit amplitude of the offset operation. If the number is too small, the offset range is limited and the defect dispersion effect is insufficient; if the number is too large, the effective printing width is narrowed and the printing efficiency decreases. Based on this, the technical solution of this application further sets the number of spare nozzles reserved at both the top and bottom ends according to the total number of nozzles and the equipment boundary tolerance requirements, and constructs the allowable offset range in the sub-scanning direction with the number of spare nozzles as the upper limit. Through the above processing, a clear integer interval boundary constraint can be provided for the subsequent sampling of random offsets layer by layer.

[0022] More specifically, in a specific example of this application, the total number of nozzles N in the printhead array is 1000. Based on the mechanical positioning accuracy and boundary tolerance requirements of the device in the sub-scanning direction, 50 nozzles are reserved at the upper and lower ends of the nozzle array as spare channels that do not participate in regular image printing; that is, the number of spare nozzles R is set to 50. These 50 spare nozzles do not carry any slice image content in the nominal state; their purpose is to provide physical space margin for the positive or negative translation of the nozzle window in the sub-scanning direction. Based on this, using the number of spare nozzles R as the upper limit of the offset amplitude, a closed interval [-R, +R], i.e., [-50, +50], is constructed for the allowable offset range in the sub-scanning direction. This indicates that during each layer of printing, the effective working nozzle window is allowed to offset a maximum of 50 nozzle spacings in the positive direction and a maximum of 50 nozzle spacings in the negative direction relative to the nominal position in the sub-scanning direction. Any integer value within this interval corresponds to a legal nozzle window translation state, and the translated working area will not exceed the physical boundary of the printhead array.

[0023] In step S22, the number of effective working nozzles is determined based on the total number of nozzles and the number of spare nozzles at both ends. The effective printing width and the effective working nozzle range are then calculated and defined based on the nozzle spacing. It should be noted that since the number of spare nozzles has been determined, it is necessary to further clarify the remaining number of nozzles capable of carrying image content and the physical printing area they cover, so that the pixel column number of subsequent sliced ​​image data can accurately correspond one-to-one with the effective working nozzles. Based on this, the technical solution of this application further determines the number of effective working nozzles based on the total number of nozzles and the number of spare nozzles at both ends, and calculates the effective printing width and defines the effective working nozzle range based on the nozzle spacing. Through the above processing, a definite channel number boundary and physical width reference can be provided for the subsequent recalculation of pixel-to-nozzle mapping.

[0024] More specifically, in a specific example of this application, the total number of nozzles N is 1000, the number of spare nozzles R at both the top and bottom ends is 50, and the nozzle spacing P is 42.3 micrometers. First, subtract twice the number of spare nozzles at the top and bottom ends from the total number of nozzles to obtain the effective number of working nozzles M that actually participate in image printing under nominal conditions, i.e., M = N - 2R = 1000 - 2 × 50 = 900. Then, multiply the effective number of working nozzles M by the nozzle spacing P to obtain the effective printing width W, i.e., W = M × P = 900 × 42.3 = 38070 micrometers. This value represents the maximum image printing range that the printhead can cover in the sub-scanning direction, and the number of pixel columns of the current layer slice image data corresponds to this width. Based on this, the effective working nozzle range under nominal conditions is defined according to the number of spare nozzles R and the total number of nozzles N. The starting nozzle number is R+1, i.e., nozzle number 51, and the ending nozzle number is NR, i.e., nozzle number 950, thus forming a closed interval [51, 950] as the effective working nozzle range. The 900 continuous physical channels within this range carry all slice image content when no offset is applied, while the 50 spare channels on each side of the range remain silent under nominal conditions.

[0025] In step S23, the blockage nozzle numbers in the blockage location identification diagram are traversed, and it is verified whether they fall within the effective working nozzle range to confirm the necessity of random offset compensation. The effective working nozzle range, effective print width, and allowable offset range in the sub-scanning direction are then output. It should be noted that since blockage nozzles may only appear within the two spare intervals without affecting the actual image printing quality, indiscriminately activating random offset compensation for all cases would cause unnecessary print width loss and efficiency reduction. Therefore, the technical solution of this application further traverses the blockage nozzle numbers in the blockage location identification diagram, verifying whether they fall within the effective working nozzle range to confirm the necessity of random offset compensation, and outputs the effective working nozzle range, effective print width, and allowable offset range in the sub-scanning direction. Through the above processing, the compensation process is triggered only when the blockage actually affects the molding quality, avoiding unnecessary resource consumption.

[0026] More specifically, in a specific example of this application, the plugging location identification diagram records three plugging nozzles, with physical serial numbers 30, 800, and 965, respectively, and the effective working nozzle range is [51, 950]. The physical serial number of each plugging nozzle in the plugging location identification diagram is read one by one, and its numerical range is determined by comparing it with the start and end boundaries of the effective working nozzle range. The serial number of nozzle 30 is less than the lower boundary of the range (51), belonging to the lower reserve range, and will not affect the image printing area; the serial number of nozzle 800 is between 51 and 950, falling within the effective working nozzle range, and its plugging will directly cause material deposition gaps at the corresponding positions in the printed image; the serial number of nozzle 965 is greater than the upper boundary of the range (950), belonging to the upper reserve range, and similarly will not affect the image content. Since at least one blocked nozzle number falls within the valid working nozzle range, it is confirmed that random offset compensation is necessary. The valid working nozzle range [51, 950], the valid print width of 38070 micrometers, and the allowable offset range in the sub-scan direction [-50, +50] are then output as parameters for subsequent steps. If all blocked nozzles are located in the spare range and not within the valid working range, there is no need to initiate the random offset compensation process; regular printing can be performed directly according to the nominal nozzle window.

[0027] Specifically, in step S3, based on the current printing layer number, uniformly distributed random integer sampling is performed within the allowable offset range of the sub-scanning direction to obtain the number of offset holes in the current layer. It should be noted that, given that the physical position of the plugging holes in the nozzle array remains fixed, if each layer of printing uses the same nozzle window position, the plugging hole defects will accumulate along the same sub-scanning coordinate in the layer-by-layer stacking, forming continuous cracks. However, by introducing randomized offsets layer by layer, the same plugging hole can correspond to different substrate coordinates in different layers, dispersing the defects over a larger area. Based on this, the technical solution of this application further performs uniformly distributed random integer sampling within the allowable offset range of the sub-scanning direction based on the current printing layer number to obtain the number of offset holes in the current layer. Through the above processing, an independent and unpredictable offset can be generated for each printing layer, making the location of the plugging hole defects randomly distributed rather than fixedly repeated in the layer-by-layer printing, providing a differential offset basis for the formation of the subsequent mesh-like interweaving dispersion effect.

[0028] Figure 4 This is a flowchart illustrating a random offset compensation method for inkjet printer nozzle clogging according to an embodiment of this application. The method involves uniformly distributing random integer sampling within the allowable offset range in the sub-scanning direction based on the current printing layer number to obtain the number of offset holes in the current layer. (See flowchart for example.) Figure 4As shown, step S3 includes: S31, performing a hash obfuscation operation on the current printing layer number based on a preset perturbation salt value to generate a random number generation seed, and initializing the pseudo-random number generator state accordingly; S32, calling the pseudo-random number generator to perform single-step iteration on the pseudo-random number generator state to generate a normalized random value; S33, performing linear scaling and floor-rounding quantization on the normalized random value based on the number of spare nozzles in the allowed offset range of the sub-scanning direction to obtain the number of offset nozzles in the current layer.

[0029] In step S31, based on a preset perturbation salt value, a hash confusion operation is performed on the current printed layer number to generate a random number generation seed, and the pseudo-random number generator state is initialized accordingly. It should be noted that since the printed layer number itself is a monotonically increasing continuous integer, if it is directly used as the seed input for the pseudo-random number generator, the random values ​​generated between adjacent layers will exhibit a strong correlation, leading to regular repetition or gradual changes in the offsets of consecutive layers, thus weakening the randomness of defect dispersion. Based on this, the technical solution of this application further performs a hash confusion operation on the current printed layer number based on a preset perturbation salt value to generate a random number generation seed, and initializes the pseudo-random number generator state accordingly. Through the above processing, ordered layer numbers can be transformed into discretely distributed seed values, allowing the pseudo-random number generator to operate independently in different layers with sufficiently differentiated initial states.

[0030] More specifically, in a concrete example of this application, the current print layer number L is 127, the preset perturbation salt value is a fixed multiplication constant, and the preset modulo large prime number is a selected prime number. The specific process of performing hash obfuscation operation on the current print layer number is as follows: perform integer multiplication operation between the layer number L and the preset perturbation salt value, and then perform modulo operation on the product result using the preset large prime number to obtain the random number generation seed. Its calculation expression is:

[0031] in, Generate a seed for the generated random numbers. This is the current print layer number. To preset the perturbation salt value, mod is the modulo operation. A large prime number is preset for modulo operation. This multiplication-modulo operation ensures that even if adjacent layer indices differ by only 1, the seed value obtained after modulo operation with a large prime number exhibits a jump distribution in the numerical space, eliminating the linear correlation of input indices. Subsequently, the calculated random number generation seed is loaded into the internal register of the pseudo-random number generator to complete state initialization, enabling the generator to enter a deterministic working state uniquely corresponding to the current layer. This deterministic feature guarantees that completely consistent offset results can be reproduced under the same layer indices and the same salt values, meeting the consistency requirements when resuming printing after a print job interruption.

[0032] In step S32, a pseudo-random number generator is invoked to perform a single-step iteration on its state to generate a normalized random value. It should be noted that since the pseudo-random number generator's internal registers only store deterministic initial state information after state initialization and have not yet generated actual random values ​​usable for offset calculation, a state iteration operation is required to output a statistically uniform numerical result. Therefore, the technical solution of this application further invokes the pseudo-random number generator to perform a single-step iteration on its state to generate a normalized random value. Through the above processing, a floating-point value between 0 and 1 that satisfies uniform distribution characteristics can be obtained, providing a continuous-domain random source for subsequent mapping to the discrete offset aperture number.

[0033] More specifically, in a concrete example of this application, the pseudo-random number generator employs a linear congruential algorithm to implement state iteration. Based on the initialized pseudo-random number generator state, a single-step iteration is performed: the current state value is multiplied by a preset multiplication coefficient, then a preset increment constant is added, and the remainder is taken with respect to a preset modulus to obtain the updated state value. Subsequently, the updated integer state value is normalized by dividing by the modulus, mapping it to the half-open interval [0,1) to obtain a normalized random value. This normalized random value statistically follows a uniform distribution on the interval [0,1), with each value generated appearing with equal probability throughout the entire interval, without local clustering or periodic repetition. This normalized random value is passed as an intermediate result to the next step for further scaling and quantization to the number of discrete integer offset apertures in the sub-scanning direction.

[0034] In step S33, based on the number of spare nozzles within the allowable offset range of the sub-scanning direction, the normalized random value is linearly scaled and rounded down to obtain the number of offset nozzles in the current layer. It should be noted that since the normalized random value is a continuous floating-point number between 0 and 1, and the physical offset of the nozzle window must be executed in integer nozzle spacing units, the continuous random value needs to be converted into a discrete integer value within the allowable offset range of the sub-scanning direction to drive the nozzle window to perform precise translation. Based on this, the technical solution of this application further uses the number of spare nozzles within the allowable offset range of the sub-scanning direction to linearly scale and round down the normalized random value to obtain the number of offset nozzles in the current layer. Through the above processing, uniformly distributed continuous random values ​​can be mapped with equal probability to a discrete integer within the offset range, serving as the specific instruction for translating the nozzle window in the current layer.

[0035] More specifically, in a particular example of this application, the number of spare nozzles R is 50, and the allowable offset range of the sub-scanning direction is [-50, +50], which contains 101 valid integer values. Assuming the normalized random value u generated in the previous step is 0.637, the calculation process for linear scaling and floor-rounding quantization is as follows:

[0036] in, This represents the number of offset holes in the current layer. For normalized random values, This represents the number of spare nozzles. The formula first multiplies the normalized random value by the total number of integers within the offset range, 2R+1, to perform linear scaling from the [0,1) interval to the [0,101) interval. Then, it subtracts the number of spare nozzles R to shift the interval to [-50,51). Finally, it performs a floor function to quantize the real value into the largest integer not greater than this value. Substituting the specific values, the current layer's offset nozzle count is 14. This result indicates that the effective working nozzle window of the current printing layer is offset positively by 14 nozzle spacings relative to the nominal position in the sub-scanning direction. Since the normalized random value is uniformly distributed in the [0,1) interval, after the above linear scaling and floor function, each integer value in the [-50,+50] interval has an equal probability of being selected, ensuring that the offset unbiasedly covers the entire allowable offset range between layers.

[0037] Specifically, in step S4, the effective working nozzle range is shifted as a whole using the current layer offset hole count. The pixel-to-nozzle mapping of the current layer slice image data and the shifted nozzle range is recalculated. Furthermore, the excitation state of the pixels corresponding to blocked nozzles is set to skipped based on the blockage location marker map, generating the current layer offset nozzle excitation sequence. It should be noted that, given that the current layer offset hole count is already determined, this abstract integer offset needs to be converted into specific drive instructions executable by the nozzle hardware. This means clearly defining which physical nozzle is responsible for spraying each pixel column after the offset, while simultaneously shielding known blocked nozzle channels to avoid invalid drive. Based on this, the technical solution of this application further utilizes the current layer offset hole count to shift the effective working nozzle range as a whole, recalculates the pixel-to-nozzle mapping of the current layer slice image data and the shifted nozzle range, and sets the excitation state of the pixels corresponding to blocked nozzles to skip based on the blockage location marker map, generating the current layer offset nozzle excitation sequence. Through the above processing, the random offset can be converted into a complete excitation sequence containing the driving parameters of each physical channel, so that the blocked channel is precisely shielded while the normal channel is driven independently according to the image gray value, providing pulse control data that can be directly sent to the printhead to perform the current layer scanning printing.

[0038] Figure 5This document describes a method for random offset compensation of inkjet printer nozzle clogging according to an embodiment of this application. It involves shifting the effective working nozzle range as a whole using the current layer's offset nozzle count, recalculating the pixel-to-nozzle mapping between the current layer slice image data and the shifted nozzle range, and setting the excitation state of the corresponding pixel of the clogged nozzle to skip, based on a clogging location marker map, thus generating a flowchart of the current layer's offset nozzle excitation sequence. Figure 5 As shown, step S4 includes: S41, adding the start and end numbers of the effective working nozzle interval to the current layer offset nozzle number and performing overall translation and reconstruction to obtain the actual excitation nozzle interval of the current layer; S42, based on the effective printing width, aligning each pixel column of the current layer slice image data sequentially to each physical nozzle channel in the actual excitation nozzle interval of the current layer to establish a pixel-nozzle matching mapping table; S43, based on the pixel-nozzle matching mapping table, generating the current layer offset nozzle excitation sequence.

[0039] In step S41, the start and end numbers of the effective working nozzle interval are added to the current layer's offset hole number, and then the entire sequence is translated and reconstructed to obtain the actual excitation nozzle interval of the current layer. It should be noted that since the current layer's offset hole number represents a logical offset relative to the nominal position, it needs to be superimposed on the start and end numbers of the effective working nozzle interval to determine the actual physical channel range of the current layer participating in inkjet printing. Based on this, the technical solution of this application further adds the current layer's offset hole number to the start and end numbers of the effective working nozzle interval and then performs an overall translation and reconstruction to obtain the actual excitation nozzle interval of the current layer. Through the above processing, the abstract offset can be transformed into a clear physical channel number boundary, providing a definite spatial reference for subsequent pixel-to-nozzle mapping alignment.

[0040] More specifically, in a specific example of this application, the effective working nozzle range is [51, 950], and the current layer offset nozzle number is 14. Adding the start and end numbers of the effective working nozzle range to the current layer offset nozzle number yields the actual activated nozzle range of the current layer as [65, 964]. This range still contains 900 consecutive physical nozzle channels, consistent with the nominal number of effective working nozzles. The entire range is shifted forward in the sub-scanning direction by a physical distance of 14 nozzle spacings. Since the number of spare nozzles is 50, the shifted end number 964 does not exceed the physical boundary of the printhead array; that is, the shifted range still falls entirely within the available channels of the printhead, and there is no risk of exceeding the boundary. The actual activated nozzle range of the current layer determines that the consecutive physical channels from nozzle number 65 to nozzle number 964 will be activated and used during printing of this layer, while the remaining channels remain silent.

[0041] In step S42, based on the effective print width, each pixel column of the current layer slice image data is sequentially aligned to each physical nozzle channel in the actual excitation nozzle range of the current layer to establish a pixel-nozzle matching mapping table. It should be noted that since the physical channel number changes after the nozzle window is offset, and the pixel column index of the slice image data always starts from column 1, a one-to-one correspondence between pixel columns and the offset physical nozzles needs to be established to determine which specific physical channel the grayscale driving information of each pixel should be sent to. Based on this, the technical solution of this application further aligns each pixel column of the current layer slice image data sequentially to each physical nozzle channel in the actual excitation nozzle range of the current layer based on the effective print width to establish a pixel-nozzle matching mapping table. Through the above processing, a precise correspondence between pixel grayscale values ​​and physical nozzle channels can be provided for subsequent excitation sequence generation.

[0042] More specifically, in a specific example of this application, the actual excitation nozzle range of the current layer is [65, 964], the effective print width corresponds to 900 pixel columns, and the slice image data of the current layer is also discrete into 900 pixel columns in the sub-scan direction. The first pixel column of the slice image data is aligned to the starting channel of the actual excitation nozzle range of the current layer, i.e., nozzle number 65; the second pixel column is aligned to nozzle number 66, and so on, with the 900th pixel column aligned to the ending channel of the range, i.e., nozzle number 964. For any pixel column index x, its corresponding physical nozzle channel number is determined according to the following relationship:

[0043] in, Let x be the physical nozzle number mapped to the x-th pixel column. This is the starting number of the effective working nozzle range under nominal conditions. This represents the number of offset holes in the current layer. The pixel column index is used, with values ​​ranging from 1 to the number of valid working nozzles M. All 900 pairs of pixel column indices and their corresponding physical nozzle numbers are recorded sequentially to form a pixel-nozzle matching mapping table. Each record in this table contains a pixel column index and its unique corresponding physical nozzle channel number. Subsequent steps use this table to determine which physical channel the grayscale driving information for each pixel should be sent to for spraying.

[0044] In step S43, an excitation sequence of the nozzles after the current layer offset is generated based on the pixel-nozzle matching mapping table. It should be noted that since the pixel-nozzle matching mapping table has established a one-to-one correspondence between image pixels and physical nozzles, it is necessary to further combine the nozzle blockage location identification map to make a nozzle-by-nozzle decision on the excitation state of each channel, excluding blocked channels from the driving sequence, and assigning pulse driving parameters matching the image grayscale values ​​to normal channels. Based on this, the technical solution of this application further generates an excitation sequence of the nozzles after the current layer offset based on the pixel-nozzle matching mapping table. Through the above processing, a complete excitation command containing the driving states of all physical channels can be formed, which can be directly sent to the printhead driving circuit to execute the current layer scanning and printing.

[0045] More specifically, in a specific example of this application, the pixel-nozzle matching mapping table contains 900 records, each corresponding to a pixel column index and its mapped physical nozzle number. Each record in the mapping table is traversed sequentially, and the physical nozzle number corresponding to the current record is read and searched for in the nozzle blockage location identification map. If the physical nozzle number exists in the blockage record of the nozzle blockage location identification map, the excitation state of that channel is forcibly set to skip, and its driving parameter is set to zero, i.e., no voltage pulse is applied to that piezoelectric channel to avoid applying a driving signal to the blocked nozzle and causing abnormal back pressure. If the physical nozzle number is not in the blockage record, the grayscale value of the corresponding position of the pixel column is read from the current layer slice image data, and the grayscale value is converted into a driving pulse width parameter according to a linear quantization relationship. The higher the grayscale value, the larger the pulse width and the larger the corresponding ink droplet volume. The excitation logic of each channel is determined according to the following relationship:

[0046] in, Physical channels in the nozzle excitation sequence after current layer offset The driving parameter values, This is a diagram showing the location of the plug. This represents the grayscale driving value of the x-th pixel column in the current layer slice image data. The driving states of all 900 channels are sequentially encapsulated according to the physical nozzle number to generate the nozzle excitation sequence after the current layer offset.

[0047] The above embodiments achieve hole blocking and normal channel driving through independent decision-making for each hole, and realize a mesh-like interwoven distribution of defects by combining interlayer random offset. In application scenarios with higher requirements for the uniformity of the density of the molded parts, the following physical fact can be further considered: In the inkjet additive manufacturing process containing particulate ink, after the ink droplets land on the powder bed or substrate layer, they will spread laterally within a certain range due to the combined effect of capillary force and surface tension. The normal operating nozzles adjacent to the hole blocking sites have the objective ability to increase the droplet volume by appropriately increasing the driving pulse width, so that the droplets spread towards the direction of the hole blocking site to partially cover the material gap area. Based on this, in another embodiment, in order to further improve the material coverage of the hole blocking area in a single-layer plane, a neighborhood compensation weight allocation mechanism based on the inverse distance decay can be introduced on the basis of the above, so as to recover the grayscale information at the hole blocking site to the adjacent normal channel according to the decay weight, thereby realizing the active repair of the local material density of the hole blocking area in a single-layer plane.

[0048] Figure 6 This is a flowchart illustrating a random offset compensation method for inkjet printer nozzle clogging according to an embodiment of this application. It describes a gain reconstruction of the pixel-nozzle matching map based on neighborhood attenuation compensation to generate the nozzle excitation sequence after the current layer offset. Figure 6 As shown, step S43 includes: S431, based on a preset neighborhood compensation radius, performing a neighborhood normal nozzle search on both sides and distance inverse attenuation weight normalization on each blocked channel in the pixel-nozzle matching map that hits the blocked hole position marker to obtain a neighborhood compensation weight distribution table; S432, based on the neighborhood compensation weight distribution table and a preset compensation gain coefficient, performing weighted neighborhood amortization and amplitude limiting processing on the gray values ​​of each blocked pixel position in the current layer slice image data to obtain a gray value parameter sequence after neighborhood gain compensation; S433, traversing the pixel-nozzle matching map, setting the channel excitation parameter of the hit blocked hole position marker to zero, taking the corresponding value in the gray value parameter sequence after neighborhood gain compensation for the driving parameter of the non-blocked channel, and generating the current layer offset nozzle excitation sequence.

[0049] In step S431, based on a preset neighborhood compensation radius, the neighboring normal nozzles on both sides of each blocked channel in the pixel-nozzle matching map that hits the blocked hole location marker are searched and normalized with the inverse distance decay weight to obtain a neighborhood compensation weight distribution table. It should be noted that, given that in the first embodiment, each normal nozzle generates driving parameters independently based only on the original grayscale value of its own pixel position, and does not respond to the material deficiency at the adjacent blocked hole position, the pixel grayscale value corresponding to the blocked hole is completely discarded and not received by any physical channel. However, in the actual physical scenario, the lateral spreading ability of ink droplets on the substrate surface decreases with increasing physical distance. The closer the normal nozzle is to the blocked hole, the higher its ink droplet coverage efficiency in the blocked hole area. Therefore, it is necessary to establish a quantitative spatial compensation weight relationship. Based on this, the technical solution of this application further performs a neighboring normal nozzle search on both sides of each blocked channel in the pixel-nozzle matching map that hits the blocked hole location marker and normalizes with the inverse distance decay weight to obtain a neighborhood compensation weight distribution table. Through the above processing, the physical spreading distance attenuation characteristics of ink droplets can be transformed into a calculable discrete weight distribution, so that the compensation direction and compensation intensity match the actual physical coverage of ink droplets on the substrate surface.

[0050] More specifically, in a specific example of this application, for each blocked channel in the pixel-nozzle matching mapping table that matches the blockage location marker, in the sub-scanning direction, with the blockage as the center and a preset neighborhood compensation radius as the search boundary, normally functioning nozzles located within this radius and not recorded in the blockage location marker are searched on both sides. Based on the number of physical channel intervals between each neighboring normal nozzle and the central blockage, the original compensation weight value of each neighboring nozzle is calculated using the inverse distance decay criterion. Then, the original weight values ​​of all participating neighboring nozzles are normalized so that the sum of all neighborhood compensation weights corresponding to a single blockage is strictly equal to 1, thus obtaining a neighborhood compensation weight distribution table. Specifically, the preset neighborhood compensation radius K, measured in units of the number of nozzles, represents the range of searching K channels to the left and right from the center of the blockage. For each blocked channel, the state of each adjacent channel within the K channel range on both sides is checked one by one, excluding neighboring channels that are also blocked, and only normally functioning nozzles are retained as compensation candidates. Subsequently, the absolute channel interval between each candidate normal nozzle and the central plug is calculated, and the reciprocal of this interval is taken as the original compensation weight value. The smaller the interval, the larger the weight, reflecting the attenuation characteristic that the closer the physical distance, the stronger the spreading coverage. Finally, the original weight values ​​of all candidate normal nozzles of the plug are summed as the normalization denominator. The original weight value of each candidate nozzle is divided by this denominator to obtain the normalized compensation weight. For the j-th plug, the normalized compensation weight of the k-th normal nozzle in its neighborhood is determined as follows:

[0051] in, The normalized compensation weight value assigned to the j-th plugged hole and its k-th neighboring normal nozzle. This represents the offset of the neighboring channel relative to the plug, with positive values ​​indicating a rightward offset and negative values ​​indicating a leftward offset, and the values ​​satisfying the following conditions: and , The preset neighborhood compensation radius is measured in units of the number of nozzles. This represents the absolute channel spacing between neighboring nozzles and the central plug. Relative to the position of the hole-blocking pixel The physical nozzle number obtained from the pixel-nozzle matching map after offsetting by m pixels. This is a diagram showing the location of the plug. This is an indicator function that takes the value 1 when the condition in parentheses is true, and 0 otherwise. It is used to exclude neighboring channels that are also blocked in the normalized denominator.

[0052] The denominator of this weighting formula only sums and accumulates the actual working normal nozzles within the neighborhood. When there are multiple consecutive blocked channels within the neighborhood of a blocked nozzle, the compensation responsibility will be automatically concentrated on the available nozzles further away, and the weight allocation will be dynamically adjusted accordingly, ensuring the robustness of the compensation scheme to multi-hole continuous blockage scenarios.

[0053] Furthermore, in a scenario-based example, assuming that physical nozzle number 800 in the pixel-nozzle matching map is a blocked channel, the preset neighborhood compensation radius K is 3, meaning that 3 channel positions are searched on each side. Channels 799, 798, and 797 on the left, and 801, 802, and 803 on the right, a total of 6 channels, are included in the search range. After comparison with the blocked hole location identification map, nozzle number 799 is also blocked and is excluded, while the remaining 5 channels are in normal working condition. The reciprocal distance weights are calculated for each of the 5 candidate normal nozzles: channel number 798 has an interval of 2, with an original weight of 1 / 2; channel number 797 has an interval of 3, with an original weight of 1 / 3; channel number 801 has an interval of 1, with an original weight of 1 / 1; channel number 802 has an interval of 2, with an original weight of 1 / 2; and channel number 803 has an interval of 3, with an original weight of 1 / 3. The normalized denominator is the sum of the five original weights mentioned above. The final normalized compensation weight for each candidate channel is its original weight value divided by this denominator. It can be seen that channel 801, closest to the blocked orifice, received the largest compensation weight. On the left, since channel 799 was also blocked and was excluded, the compensation responsibility was automatically concentrated on channels 798 and 797, demonstrating the mechanism's adaptive adjustment capability in multi-orifice continuous blockage scenarios. Summarizing the neighborhood compensation weights corresponding to all blocked channels constitutes a complete neighborhood compensation weight distribution table.

[0054] In step S432, based on the neighborhood compensation weight distribution table and the preset compensation gain coefficient, the gray values ​​of each blocked pixel position in the current layer slice image data are weighted and superimposed with neighborhood apportionment and limited to obtain a gray value parameter sequence after neighborhood gain compensation. It should be noted that, given that the neighborhood compensation weight distribution table has determined the compensation ratio that each normal nozzle in the neighborhood should bear, but has not yet clarified the actual gray value increment to be compensated, nor has it superimposed the increment to the driving parameters of each nozzle, it is necessary to quantitatively recover the material deposition requirements that should have been performed at the blocked position but could not be implemented due to the nozzle blockage to the neighborhood channel according to the weight ratio. Based on this, the technical solution of this application further performs weighted neighborhood apportionment and superimposition with neighborhood apportionment and limited to obtain a gray value parameter sequence after neighborhood gain compensation based on the neighborhood compensation weight distribution table and the preset compensation gain coefficient. Through the above processing, the pixel grayscale information that was originally completely discarded at the hole-blocking point can be recovered to the adjacent normal channel with a physically achievable attenuation ratio, thereby increasing the material coverage of the hole-blocking area in a single layer from 0% in the first embodiment to the target ratio controlled by the compensation gain coefficient.

[0055] More specifically, in a concrete example of this application, for each blocked channel in the pixel-nozzle matching mapping table that matches the blockage location identifier, the original grayscale value corresponding to the blocked pixel location is extracted from the current layer slice image data. This grayscale value essentially represents the droplet volume drive requirement that the blocked hole should eject during normal operation. This grayscale value is multiplied by a preset neighborhood compensation gain coefficient to obtain the grayscale increment to be allocated. This coefficient is used to control the aggressiveness of the compensation force, and its value ranges between 0 and 1 to avoid overcompensation leading to overload of the neighborhood nozzle drive. Subsequently, according to the pairing records of each normal nozzle in the neighborhood corresponding to the blocked hole in the neighborhood compensation weight distribution table, the grayscale increment to be allocated is distributed to the corresponding neighborhood positions one by one according to the weight ratio. Finally, all non-blocked nozzle positions in the mapping table are traversed, and the original grayscale drive value of that position in the current layer slice image data is summed and superimposed with the total gain accumulated from all the blocked holes around it and allocated to that position. The summation is then limited and truncated by the physical upper limit value of the nozzle drive level pulse to ensure hardware safety. The final grayscale driving parameters for each non-clogging nozzle position after compensation are determined as follows:

[0056] in, The final grayscale driving parameter value is the pixel position x corresponding to the non-clogging nozzle, after neighborhood compensation gain superposition. This is the original grayscale driving value of pixel position x in the current layer slice image data, representing the inkjet requirements that this nozzle itself should undertake. The preset neighborhood compensation gain coefficient controls the proportion of grayscale information transferred from the hole-blocking area to the neighborhood. This is the set of all hole-blocking indices that benefit from pixel position x in their neighborhood; that is, which hole-blocking radii fall within position x. Let j be the pixel column position corresponding to the j-th plug in the mapping table. Location of the plug in the current layer slice image data The original grayscale value at that location represents the total amount of ink droplet drive demand that should have been executed but could not be ejected due to blockage. Let x be the normalized compensation weight assigned to the j-th plug in the neighborhood compensation weight distribution table at position x, where The relative pixel displacement between the two. The physical upper limit of the nozzle drive level pulse parameter is determined by the safe operating range of the nozzle piezoelectric ceramic or heating element.

[0057] When a normal nozzle is located in the overlapping area of ​​the compensation neighborhood of multiple plugs, the compensation increments from different plugs will have a cumulative superposition effect at that location. The introduction of the limiting operator ensures that even in extreme scenarios with multiple plugs densely distributed, the driving parameters will not exceed the hardware safety boundary, thus avoiding polarization degradation or thermal failure of the piezoelectric element due to overload pulses.

[0058] Further, in a scenario-based example, assuming the original grayscale value of the pixel corresponding to the 800th blocked nozzle is 200, and the preset neighborhood compensation gain coefficient is 0.8, the grayscale increment to be distributed is 160. According to the neighborhood compensation weight distribution table, the normalized compensation weight obtained by the 801st normal nozzle is 0.375, so the grayscale gain allocated to the 801st position is 60. Assuming the original grayscale value of the 801st nozzle's own pixel position is 180, and this position only falls within the neighborhood of the 800th blocked nozzle, its final grayscale driving parameter is 240. This value does not exceed the physical upper limit of the driving level pulse of 255, so 240 is directly taken as the final driving parameter and written into the grayscale parameter sequence after neighborhood gain compensation. If a normal nozzle is located in the neighborhood overlap area of ​​two blocked nozzles, and the gain from the two blocked nozzles exceeds the physical upper limit of 255 after accumulation at this position, then 255 is truncated to protect the safe working state of the piezoelectric element. After traversing all non-blocking channels to complete the above superposition and limiting processing, a complete grayscale parameter sequence after neighborhood gain compensation is obtained.

[0059] In step S433, the pixel-nozzle matching mapping table is traversed, the channel excitation parameters of the matched hole-blocking location markers are set to zero, and the driving parameters of the non-blocking channels are taken from the corresponding values ​​in the neighborhood gain-compensated grayscale parameter sequence to generate the current layer offset nozzle excitation sequence. It should be noted that, given that the neighborhood gain-compensated grayscale parameter sequence already contains the compensated and enhanced driving values ​​of each non-blocking channel, but has not yet been merged with the skip instruction for the blocked channel into a complete excitation sequence that can be directly sent to the nozzle driving circuit, the two need to be uniformly encapsulated to form a hardware-executable final driving instruction set. Based on this, the technical solution of this application further traverses the pixel-nozzle matching mapping table, sets the channel excitation parameters of the matched hole-blocking location markers to zero, and takes the corresponding values ​​in the neighborhood gain-compensated grayscale parameter sequence for the driving parameters of the non-blocking channels to generate the current layer offset nozzle excitation sequence. Through the above processing, a complete excitation sequence with both physical shielding and neighborhood gain compensation characteristics can be formed, which can actively repair the local material density in the hole-blocking area within a single-layer plane while maintaining the random offset interlayer dispersion mechanism.

[0060] More specifically, in a concrete example of this application, all physical nozzle channels in the pixel-nozzle matching map are traversed. For blocked channels that match the blockage location marker, their pulse drive parameters are forcibly set to zero to maintain the physical skip state, avoiding applying voltage to the blocked piezoelectric channels and causing abnormal back pressure or diaphragm damage. For all non-blocked channels, the value at the corresponding position in the grayscale parameter sequence after neighborhood gain compensation is directly used as its final pulse width drive parameter. The drive states of all channels are encapsulated in order according to the physical nozzle sequence number to generate the nozzle excitation sequence after the current layer offset. The drive parameters of each channel in the final excitation sequence are determined according to the following logic:

[0061] in, Physical channels in the nozzle excitation sequence after current layer offset The final driving parameter values, This refers to the physical nozzle number corresponding to pixel column x in the pixel-nozzle matching mapping table. This is a diagram showing the location of the plug. The compensated driving value at position x in the grayscale parameter sequence after neighborhood gain compensation already contains the dual components of the original grayscale and the superposition of neighborhood gain.

[0062] The formula is consistent in form with the excitation sequence encapsulation structure of the first embodiment, but the driving value of the non-blocking channel has been upgraded from a single value that only carries the grayscale of its own pixel in the first embodiment to a composite driving value that carries the gain superposition of its own grayscale and the grayscale information of the neighboring hole-blocking channels according to the attenuation weight. The hole-blocking channel is still physically skipped to avoid hardware damage, while the ink droplet volume of the neighboring normal nozzles is moderately increased by increasing the driving pulse width, so that the spreading and covering distance of the ink droplets on the substrate surface in the direction of hole blockage is extended accordingly. Thus, while maintaining the random offset interlayer dispersion mechanism, the active repair of the local material density in the hole-blocking area is achieved in a single-layer plane.

[0063] Furthermore, in a scenario-based example, assume that the actual excitation nozzle range after the current layer offset contains 900 physical channels, with channel 800 being blocked. When traversing to channel 800, it is confirmed to be blocked by comparing with the blockage location identification map. The driving parameter of this channel is set to zero, and no pulse signal is applied to this piezoelectric element. When traversing to channel 801, it is confirmed that it is not in the blockage record. The compensated driving value 240 at the corresponding position is read from the neighborhood gain-compensated grayscale parameter sequence as the final pulse width parameter of this channel. This value contains the superposition result of the original grayscale value 180 of channel 801 itself and the weighted gain amount 60 from the blockage of channel 800. After completing the encapsulation by traversing all channels according to this logic, the 800th hole-blocking channel remains at zero drive to protect hardware safety, while the drive pulse width of the normal channels on both sides is increased compared to the first embodiment. As the corresponding ink droplet volume increases, the lateral spreading range on the substrate surface extends towards the 800th channel, partially covering the material gap in the hole-blocking area.

[0064] Specifically, in step S5, the printhead is driven to perform scanning printing of the current layer according to the nozzle excitation sequence after the current layer offset, and steps S3 to S4 are repeated for each subsequent layer, so that the defect positions of the same plugging hole are randomly dispersed to form a mesh-like interwoven distribution during layer-by-layer printing. It should be noted that since the previous steps have generated a complete excitation sequence including plugging hole shielding and neighborhood gain compensation, it is necessary to convert it into actual physical inkjet action and verify whether the defects have been sufficiently dispersed in the layer-by-layer iteration to confirm the effectiveness of the random offset compensation strategy. Based on this, the technical solution of this application further drives the printhead to perform scanning printing of the current layer according to the nozzle excitation sequence after the current layer offset, and repeats steps S3 to S4 for each subsequent layer, so that the defect positions of the same plugging hole are randomly dispersed to form a mesh-like interwoven distribution during layer-by-layer printing. Through the above processing, the spatial distribution of defects can be quantitatively evaluated while completing physical printing, confirming that macroscopic continuous cracks have been eliminated and the overall strength of the molded part has not been affected.

[0065] More specifically, in a particular example of this application, the implementation of step S5 includes the following three sub-steps.

[0066] In sub-step S51, the current layer offset nozzle excitation sequence is converted into pulse control signals for the printhead drive circuit. This drives the printhead to reciprocate in the main scanning direction while ejecting ink droplets from each nozzle according to the drive parameters of each channel. Clogged channels with zero drive parameters do not release ink droplets, resulting in dot-like defects at the corresponding offset positions on the substrate. After the current layer is scanned and printed, the physical serial numbers of all skipped nozzles in the current layer offset nozzle excitation sequence are extracted. The defect coordinates of these skipped channels are then calculated using the current layer offset nozzle count and nozzle spacing. Since the nozzle window is offset by several channels of the current layer offset nozzles, the actual defect position of the clogged nozzle in the substrate coordinate system requires subtracting the sum of the nominal starting serial number and the current layer offset nozzle count from the clogged nozzle physical serial number, and then multiplying by the nozzle spacing to obtain its absolute physical coordinates. The calculated coordinates of each clogged nozzle are then summarized to form the current layer defect physical coordinate set.

[0067] In sub-step S52, the coordinates of each defect point in the current layer's physical coordinate set are bound to the current printing layer number, forming a spatial coordinate data entry with a layer timestamp. This entry is then appended and merged into the layer-by-layer defect location distribution record. The layer-by-layer defect location distribution record is initialized as an empty set during the first layer printing and is continuously accumulated and updated as each layer printing progresses, ultimately recording the actual physical coordinates of all plug holes in each layer and their corresponding layer number throughout the entire printing task.

[0068] In sub-step S53, based on a preset mesh interlacing distribution threshold, the standard deviation and dispersion of the defect coordinates of the same plug in each layer of the layer-by-layer defect location distribution record are calculated. Specifically, the physical coordinate sequence of the defects formed by the same plug in the completed H-layer printing is extracted, the arithmetic mean of the sequence is calculated, and then the sample standard deviation is calculated as a quantitative indicator of the dispersion of the defect spatial distribution.

[0069] in, The standard deviation of the spatial distribution of defect coordinates in layer H for the same plug hole. This represents the total number of historical print layers that have been recorded. Let be the physical coordinates of the defect point caused by the plugging hole in the k-th layer on the substrate. This is the arithmetic mean of the defect coordinates of the plug in layer H. The calculated standard deviation is compared with a preset mesh-like interlacing distribution threshold. If the standard deviation is greater than or equal to the threshold, it is determined that the defect points of the plug have been sufficiently discretely distributed in the sub-scanning direction, presenting a mesh-like interlacing distribution pattern rather than forming continuous linear cracks along a fixed position. If there are still layers to be printed, return to step S3 to repeat the complete process of random offset generation and excitation sequence construction for the next layer until all layers are printed. Through the cumulative effect of layer-by-layer random offset, the distribution range of the defect position of the same plug in the sub-scanning direction covers the physical distance of the number of spare nozzles in both positive and negative directions. The continuous cracks originally concentrated in a single coordinate are distributed to multiple discrete points within this range, forming a mesh-like interlacing distribution. The overall strength of the molded part in this area is not weakened by the penetration of a single crack.

[0070] In summary, a random offset compensation method for inkjet printer nozzle clogging according to an embodiment of this application is explained, which constructs a two-dimensional collaborative compensation architecture of interlayer random offset dispersion and intralayer neighborhood gain compensation. Spare channels are reserved at both ends of the nozzle array to provide offset margin. During each layer printing, a random offset is generated based on the layer number to translate the working nozzle window as a whole, and the mapping relationship between pixels and nozzles is reconstructed, so that the defect positions of the same clogging hole are randomly dispersed layer by layer to form a mesh-like interwoven distribution, eliminating macroscopic continuous cracks. Simultaneously, a neighborhood compensation weight mechanism based on distance reciprocal decay is introduced, distributing the grayscale driving information of the clogging pixel to the normal nozzles on both sides with normalized weights. By increasing the neighborhood driving pulse, the ink droplets expand and spread towards the clogging direction, and a hardware upper limit is used for amplitude limiting protection. Active repair of the material density in the clogging area is achieved within a single layer, thereby reducing the severity of single-layer defects while dispersing the spatial position of defects, and improving the overall density and mechanical strength of the molded part.

[0071] Furthermore, a random offset compensation system for inkjet printhead clogging is also provided.

[0072] Figure 7 This is a block diagram of a random offset compensation system for inkjet printer nozzle clogging according to an embodiment of this application. Figure 7As shown, a random offset compensation system 100 for inkjet printer nozzle clogging according to an embodiment of this application includes: a nozzle state detection module 110, used to determine the nozzle droplet ejection state of each nozzle based on nozzle array specification parameters including the total number of nozzles and nozzle spacing, to obtain a clogging position identification map; an offset margin calculation module 120, used to define the effective printing interval and estimate the offset margin of the nozzle array corresponding to the clogging position identification map based on the total number of nozzles and the preset number of spare nozzles at the upper and lower ends, to obtain the effective working nozzle interval and the allowable offset range in the sub-scanning direction; and a random offset generation module 130, used to generate a random offset in the sub-scanning direction based on the current printing layer number. The number of offset holes in the current layer is obtained by uniformly distributing random integer sampling within the shift range; the mapping reconstruction and sequence generation module 140 is used to perform overall translation of the effective working nozzle range using the number of offset holes in the current layer, and to recalculate the mapping between pixels and nozzles using the current layer slice image data and the translated nozzle range, and to set the excitation state of the corresponding pixel of the blocked nozzle to skip in combination with the hole blockage location identification map, thereby generating the current layer offset nozzle excitation sequence; the layer-by-layer printing execution module 150 is used to drive the nozzle to perform current layer scanning printing according to the current layer offset nozzle excitation sequence, and to repeatedly execute the random offset generation module to the mapping reconstruction and sequence generation module for each subsequent layer, so that the defect position of the same blocked hole is randomly dispersed in the layer-by-layer printing to form a mesh-like interwoven distribution.

[0073] As described above, the random offset compensation system 100 for inkjet printer nozzle clogging according to embodiments of this application can be implemented in various types of computing devices or control units. For example, it can be deployed in an industrial control host of an inkjet additive manufacturing equipment, or integrated into an embedded processing unit within a printhead driver board. In one possible implementation, the random offset compensation system 100 for inkjet printer nozzle clogging according to embodiments of this application can be integrated into the computing device as a software module and / or a hardware module. For example, the random offset compensation system 100 for inkjet printhead clogging can be a resident data processing service in the operating system of the computing device. This software module is configured to perform per-hole state determination and clogging location identification of the ink droplet ejection detection signal, delineation of the effective printing interval and offset margin calculation based on the number of spare nozzles, generation of layer-by-layer random offsets based on layer number hash confusion, recalculation of the mapping between offset pixels and nozzles, reconstruction of excitation sequence gain based on neighborhood attenuation compensation, and verification of the mesh interlacing discreteness of layer-by-layer defect location distribution. Alternatively, it can be a dedicated inkjet printhead clogging compensation and defect dispersion control algorithm program developed for the computing device. Of course, the random offset compensation system 100 for inkjet printhead clogging can also be one of many hardware modules of the computing device or control unit, or it can be embedded in a field-programmable gate array circuit to accelerate the real-time calculation of per-hole state determination and neighborhood compensation weight distribution in parallel, or it can be an integrated circuit for printhead drive pulse generation and offset compensation control for a specific application.

[0074] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for compensating for random offset of inkjet printer nozzle clogging, characterized in that, include: Step S1: Based on the nozzle array specification parameters including the total number of nozzles and the nozzle spacing, determine the nozzle droplet ejection status of each nozzle, including the nozzle number and its corresponding ejection amplitude, to obtain a nozzle blockage location identification map. Step S2: Based on the total number of nozzles and the preset number of spare nozzles at the top and bottom ends, the effective printing range of the nozzle array corresponding to the nozzle blockage location marking diagram is defined and the offset margin is estimated to obtain the effective working nozzle range and the allowable offset range of the sub-scanning direction. Step S3: Based on the current printing layer number, perform uniformly distributed random integer sampling within the allowable offset range in the sub-scanning direction to obtain the number of offset holes in the current layer; Step S4: The effective working nozzle range is translated as a whole using the number of offset holes in the current layer, and the mapping between pixels and nozzles is recalculated between the current layer slice image data and the translated nozzle range. The excitation state of the corresponding pixel of the blocked nozzle is set to skipped in combination with the hole blockage location identification map, and the nozzle excitation sequence after the current layer offset is generated. Step S5: Drive the nozzle to scan and print the current layer according to the nozzle excitation sequence after the current layer offset, and repeat steps S3 to S4 for each subsequent layer, so that the defect positions of the same hole are randomly dispersed in the layer-by-layer printing to form a mesh-like interwoven distribution.

2. The random offset compensation method for inkjet printer nozzle clogging according to claim 1, characterized in that, Step S1 includes: Step S11: According to the total number of nozzles in the nozzle array specification parameters, perform multi-channel amplitude separation and sequence reconstruction on the nozzle ink droplet ejection detection signal to obtain the ejection amplitude feature vector of each nozzle. Step S12: Based on the preset normal injection judgment threshold, the injection amplitude feature vectors of each nozzle are compared with the injection amplitude threshold and classified into states to obtain the nozzle state classification vector. Step S13: Using the nozzle spacing in the nozzle array specification parameters, perform physical coordinate mapping and map construction on the nozzles marked as blocked in the nozzle state classification vector to obtain the blockage location identification map.

3. The random offset compensation method for inkjet printer nozzle clogging according to claim 1, characterized in that, Step S2 includes: Step S21: Set the number of spare nozzles reserved at both ends according to the total number of nozzles and the equipment boundary tolerance requirements, and construct the allowable offset range of the sub-scanning direction with the number of spare nozzles as the upper limit. Step S22: Based on the total number of nozzles and the number of spare nozzles at both ends, determine the number of effective working nozzles, and calculate the effective printing width and define the effective working nozzle range in combination with the nozzle spacing. Step S23: Traverse the sequence number of each blocked nozzle in the nozzle location identification diagram, check whether it falls within the effective working nozzle range to confirm the necessity of random offset compensation, and output the effective working nozzle range, effective print width, and allowable offset range of the sub-scanning direction.

4. The random offset compensation method for inkjet printer nozzle clogging according to claim 1, characterized in that, Step S3 includes: Step S31: Based on the preset perturbation salt value, perform a hash confusion operation on the current printing layer number to generate a random number generation seed, and initialize the pseudo-random number generator state accordingly. Step S32: Call the pseudo-random number generator to perform a single-step iteration on the pseudo-random number generator state to generate normalized random values. Step S33: Based on the number of spare nozzles within the allowable offset range of the sub-scanning direction, the normalized random value is linearly scaled and rounded down to obtain the number of offset nozzles in the current layer.

5. The random offset compensation method for inkjet printer nozzle clogging according to claim 1, characterized in that, Step S4 includes: Step S41: Add the start and end numbers of the effective working nozzle interval to the number of offset holes in the current layer and perform overall translation and reconstruction to obtain the actual excitation nozzle interval of the current layer. Step S42: Based on the effective printing width, align each pixel column of the current layer slice image data to each physical nozzle channel in the actual excitation nozzle range of the current layer in sequence to establish a pixel-nozzle matching mapping table. Step S43: Based on the pixel-nozzle matching mapping table, generate the nozzle excitation sequence after offset of the current layer.

6. The random offset compensation method for inkjet printer nozzle clogging according to claim 1, characterized in that, Step S5 includes: Step S51: Drive the nozzle to perform scanning printing according to the nozzle excitation sequence after the current layer offset, and use the current layer offset number of nozzles and nozzle spacing to back calculate the defect coordinates of the skipped channel to obtain the set of physical coordinates of the current layer defects. Step S52: After binding the physical coordinate set of the current layer defects with the current printing layer number through timestamp, append and merge it into the layer-by-layer defect location distribution record to complete the spatiotemporal dimension cumulative update of the defect distribution data; Step S53: Based on the preset mesh-like interwoven distribution threshold, calculate the standard deviation and evaluate the dispersion of the defect coordinates of the same plug in each layer in the layer-by-layer defect location distribution record to verify whether the defect points have formed a mesh-like interwoven distribution.

7. The random offset compensation method for inkjet printer nozzle clogging according to claim 5, characterized in that, Step S43 includes: Traverse the nozzle sequence number in the pixel-nozzle matching mapping table and search for a match in the hole blockage location identification map. In this process, the excitation state of the blocked channel is set to skip, and the normal channel is quantized into driving parameters according to the gray value of the current layer slice image data to generate the nozzle excitation sequence after the current layer offset.

8. The random offset compensation method for inkjet printer nozzle clogging according to claim 5, characterized in that, Step S43 includes: Step S431: Based on the preset neighborhood compensation radius, perform a search for normal nozzles in the neighborhood on both sides and normalize the distance inverse attenuation weight for each blocked channel in the pixel-nozzle matching mapping table that hits the blocked hole location identification map to obtain the neighborhood compensation weight distribution table. Step S432: Based on the neighborhood compensation weight distribution table and the preset compensation gain coefficient, perform weighted neighborhood amortization and amplitude limiting on the gray values ​​of each blocked pixel position in the current layer slice image data to obtain the gray parameter sequence after neighborhood gain compensation. Step S433: Traverse the pixel-nozzle matching mapping table, set the channel excitation parameter of the matched hole blockage location marker to zero, take the corresponding value in the grayscale parameter sequence after neighborhood gain compensation for the driving parameter of the non-blocked channel, and generate the current layer offset hole excitation sequence.

9. A random offset compensation system for inkjet printer nozzle clogging, characterized in that, include: The nozzle status detection module is used to determine the nozzle droplet ejection status of each nozzle based on the nozzle array specification parameters including the total number of nozzles and the nozzle spacing, and to obtain a nozzle blockage location identification map. The offset margin calculation module is used to define the effective printing range and estimate the offset margin of the nozzle array corresponding to the hole blockage location marking diagram based on the total number of nozzles and the preset number of spare nozzles at the upper and lower ends, so as to obtain the effective working nozzle range and the allowable offset range of the sub-scanning direction. The random offset generation module is used to perform uniformly distributed random integer sampling within the allowable offset range in the sub-scanning direction based on the current printing layer number to obtain the number of offset holes in the current layer. The mapping reconstruction and sequence generation module is used to perform an overall translation of the effective working nozzle range using the number of offset holes in the current layer, and to recalculate the mapping between pixels and nozzles using the current layer slice image data and the translated nozzle range. It also combines the hole blockage location identification map to set the excitation state of the corresponding pixel of the blocked nozzle to skip, and generates the nozzle excitation sequence after the current layer offset. The layer-by-layer printing execution module is used to drive the nozzle to scan and print the current layer according to the nozzle excitation sequence after the current layer offset. It also repeats the random offset generation module to the mapping reconstruction and sequence generation module for each subsequent layer, so that the defect positions of the same hole are randomly dispersed in the layer-by-layer printing to form a mesh-like interwoven distribution.

10. The random offset compensation system for inkjet printer nozzle clogging according to claim 9, characterized in that, The mapping reconstruction and sequence generation module includes: The interval translation and reconstruction unit is used to add the start and end numbers of the effective working nozzle interval to the number of offset holes in the current layer and perform overall translation and reconstruction to obtain the actual excitation nozzle interval of the current layer. The pixel-nozzle mapping unit is used to align each pixel column of the current layer slice image data to each physical nozzle channel in the actual excitation nozzle range of the current layer based on the effective print width in order to establish a pixel-nozzle matching mapping table. The excitation sequence generation unit is used to generate the excitation sequence of the nozzle after the current layer offset based on the pixel-nozzle matching mapping table.

Images