Method and system for stimulating nozzle by Beijing porcelain nozzle printing software

Abstract

The invention relates to the technical field of stimulation signal control, in particular to a method and a system for stimulating a nozzle by Beijing porcelain nozzle printing software. According to the method and the system, through structural decomposition of a printing task parameter group, one-to-one indexing between a nozzle number and a pattern block position is completed, and a matching system with a sequence attribute is established; the controllability of parameter management and the accuracy of scheduling are enhanced, the time sequence identification of frequency coverage and use frequency is carried out through the long and short-term memory network, the screening mechanism of excitation numbers is optimized, the excitation intensity of unnecessary nozzles is reduced, the problem of thermal load accumulation caused by frequent excitation is relieved, and the service life of the nozzles is prolonged. A genetic algorithm is introduced to realize a multi-round iterative substitution and screening mechanism of combination numbers, on the basis of ensuring combination matching logic integrity, diversity and distribution efficiency of electrical stimulation frequency use are improved, collaborative optimization of a nozzle excitation strategy under multivariable dimensions is realized, and stability of system response and signal excitation execution efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the field of stimulation signal control technology, and in particular to a method and system for stimulating nozzles for Kyocera print head printing software. Background Art

[0002] The field of stimulus signal control technology aims to implement periodic or condition-triggered electrical stimulation on target drive elements by controlling parameters such as the amplitude, frequency, waveform, and duration of electrical signals, so as to maintain, activate, adjust, or protect their working state. It focuses on the balance between signal generation mechanisms, timing control methods, power consumption, and thermal effects, and is widely used in precision control systems that require high responsiveness and continuous availability.

[0003] A method for stimulating nozzles in Kyocera printhead printing software is designed to periodically provide a certain form of voltage signal stimulation during the printing process, even if the nozzle is temporarily not involved in pattern printing, to maintain ink fluidity and prevent clogging. The purpose is to avoid nozzles that have not been working for a long time after pattern switching, such as intermittent ink droplets, weak ink droplets, and broken needles that affect print quality, thereby ensuring print output stability and improving the reliable operation time and service life of the printing equipment.

[0004] When executing electrical signal stimulation control, the existing technology relies on a fixed template-based waveform parameter application mechanism, and lacks a continuous stimulation mechanism for the nozzle in a non-printing state. When the nozzle is not stimulated for a long time, it is easy to cause problems such as ink retention and thermal imbalance, resulting in abnormal phenomena such as intermittent ink droplet ejection, blurred edges, and partial graphic incompleteness after pattern switching. In addition, the existing technology does not optimize the power distribution law, and there are phenomena such as high-load nozzle excitation overlap and low-load nozzle vacancy, which affects the synchronous response and thermal stability of the entire group of nozzles, and manifests problems such as response lag and poor printing consistency. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a method and system for stimulating nozzles for Kyocera print head printing software.

[0006] To achieve the above-mentioned object, the present invention adopts the following technical solution: a method for stimulating nozzles for Kyocera printhead printing software, comprising the following steps: S1: Based on the print task parameter group, match the nozzle number with the pattern block position index, encode the number and position sequence, merge the inkjet frequency and historical excitation record segments, divide the nozzle temperature range, and generate the nozzle operation parameter set; S2: Based on the nozzle operating parameter set, a long short-term memory network is used to screen nozzle numbers and pattern block positions, classify frequency excitation segments, combine numbers and frequency excitations, perform combined matching metrics, screen numbers and matching results, compare nozzle temperature ranges, screen numbers, and generate a preferred excitation number set; S3: Based on the preferred excitation number set, the pattern blocks are sorted, and the nozzle numbers are mapped to the power inputs, the nozzle numbers with power values ​​of zero are marked, and the numbers are extracted to generate a power sorting list; S4: Based on the power sorting list, the power and frequency excitation segments are numbered and a genetic algorithm is used to construct a numbering and combination matching set, the combination matching results are compared and screened, and non-compliant combination replacements are generated, and the process is iterated to generate a power frequency matching table; S5: Based on the power frequency matching table, read the numbered power and frequency configuration, load the printing control system, bind the numbered configuration instructions, output the signal amplitude and frequency, and generate a nozzle excitation execution instruction set.

[0007] As a further solution of the present invention, the specific steps of generating the nozzle operation parameter set are: Based on the print task parameter group, the nozzle number is matched with the pattern block position index, the nozzle number is positioned in the segment, and the number and position sequence are encoded, and the number and segment sequence are sorted to generate the number position mapping data; Based on the number position mapping data, the inkjet frequency and the historical excitation record time segment are merged, the interval coverage is performed through the frequency data and the time period, and the frequency interval is divided, and the numerical value is segmented through the frequency boundary condition to generate the frequency excitation segment data; Based on the frequency excitation segment data, the nozzle temperature value interval is divided, and the temperature data is compared with the set interval boundary, and the numbering is integrated with the frequency excitation segment and the temperature interval. The combination classification is performed through the numbering to generate a nozzle operation parameter set.

[0008] As a further solution of the present invention, the printing task parameter group is specifically the nozzle number, pattern block position, inkjet frequency, historical excitation record, and nozzle temperature input into the printing software interface.

[0009] As a further solution of the present invention, the specific steps of generating the preferred incentive number set are: Based on the nozzle operation parameter set, a long short-term memory network is used to screen the nozzle numbers and pattern block positions, match the number positions with the task segments, and perform number validity verification. The number data is compared with the available status to generate a valid number set. Based on the valid number set, frequency excitation segments are classified, number data and frequency segments are associated and combined, number and frequency excitation combinations are established, and number sequence and frequency combination are connected to generate number excitation combination data; Based on the number incentive combination data, a combination matching metric is performed, an association score is performed by combining the number and the frequency, and the number and the matching result are screened. The numbers are screened by comparing the matching degree with the temperature range to generate a preferred incentive number set.

[0010] As a further solution of the present invention, the long short-term memory network is according to the formula: ; in: Represents the nozzle number matching state vector at the current moment, Represents the activation function, used for nonlinear mapping, represents the weight matrix between the input vector and the hidden state, represents the nozzle operating parameter input vector, Indicates the weight coefficient corresponding to the printing delay factor, Indicates the current printing delay factor of the nozzle. Indicates the weight coefficient corresponding to the pattern block complexity factor, represents the pattern block complexity factor, Indicates the weight coefficient corresponding to the task section priority coefficient, Indicates the task section priority coefficient, represents the input layer bias term, represents the weight matrix of the hidden state recurrent connection, represents the hidden state vector at the previous moment, represents the hidden layer bias term.

[0011] As a further solution of the present invention, the specific steps of generating the power ranking list are: Based on the preferred excitation number set, pattern block positions are sorted, segment positioning is performed using number position data, and number sequence adjustment is performed, the numbers are aligned with the position sequence, and sequential recording is performed to generate a number sorting result; Based on the number sorting result, mapping the nozzle number and the power input is performed, data association is performed through the number sequence and the power input value, and the power input is marked as zero number, and the number is paired with the power state, and the number state is recorded to generate number power mark data; Based on the numbered power tag data, non-zero power numbers are extracted, screening is performed based on the numbers and power states, and the numbers and power values ​​are arranged in sequence. A power sorting list is generated by matching the power values ​​with the number sequence.

[0012] As a further solution of the present invention, the specific steps of generating the power frequency matching table are: Based on the power sorting list, the numbered power and frequency excitation segments are combined, the numbered power and the frequency excitation data are paired, and the combination sequence is adjusted, and the numbered power and the combination sequence are synchronized to generate the numbered power frequency combination data; Based on the number power frequency combination data, a genetic algorithm is used to compare the combination matching results, a matching degree analysis is performed through the number and the combination data, and a matching combination screening is performed, and the number screening is performed through the matching degree data and the screening conditions to generate matching combination data; Based on the compliant combination data, non-compliant combinations are replaced, the replacement is performed through numbering and combination data, and combination set iterative processing is performed, the numbering and updated combination data are synchronously updated to generate a power frequency matching table.

[0013] As a further solution of the present invention, the genetic algorithm is according to the formula: ;

[0014] Among them: Fitness(i) represents the matching score of the i-th number combination, Indicates the actual value of the i-th number combination on the j-th indicator, Represents the ideal setting value of the j-th indicator in the target combination data, represents the matching importance weight of the j-th indicator, represents the nozzle vibration stability fluctuation factor corresponding to the i-th number combination, represents the historical execution failure rate of the i-th number combination in the task path segment, Stability factor The global weight of Failure rate factor The global weight of .

[0015] As a further solution of the present invention, the specific steps of generating the nozzle excitation execution instruction set are: Based on the power-frequency matching table, read the numbered power and frequency configuration, call the power value and extract the frequency value by number, make position correspondence between the number and the configuration index sequence, associate the number with the power-frequency data, and generate numbered configuration data; Based on the numbering configuration data, the printing control system is loaded, data is imported through the numbering and control parameters, and the numbering and control parameters are mapped, binding is performed through the numbering sequence and the system control bit, and the numbering and binding parameters are synchronously updated to generate numbering control binding data; Based on the numbering control binding data, the signal amplitude and frequency are output, the output instruction is generated through the numbering and binding parameters, and the signal output sequence is determined. The control relationship is established through the numbering and output sequence to generate the nozzle excitation execution instruction set.

[0016] A system for stimulating nozzles using Kyocera printhead printing software, the system for stimulating nozzles using Kyocera printhead printing software being used to execute the above-mentioned method for stimulating nozzles using Kyocera printhead printing software, the system comprising: Nozzle matching module: Based on the print task parameter group, the module calls the pattern block position coordinate value, nozzle number index value, inkjet frequency value, historical excitation record segment number, and nozzle temperature range value to complete the item-by-item correspondence between the nozzle number and the pattern block position coordinate value. The pattern block position number is rearranged and encoded according to the nozzle number sequence. The frequency number pairs with the same inkjet frequency value and the start and end time value in the historical excitation record segment number are selected to establish a frequency merging sequence. The nozzle temperature range value is numbered and marked in intervals of 10 degrees to generate a nozzle operation number set. Excitation screening module: Based on the nozzle operation number set, a long short-term memory network is used to extract pairings of pattern block position numbers and nozzle numbers. The frequency segment number is recorded for each pair of numbers. The number of occurrences of the frequency segment number corresponding to each nozzle number is counted. The frequency segment numbers are accumulated in sequence to form a nozzle number frequency coverage total value sequence. The relative frequency of occurrence of the nozzle number in the pattern block position is obtained, and the result is divided by the median of the frequency coverage total value sequence to output a ratio. Nozzle numbers with a ratio greater than 1.0 are screened, and a table comparison is performed on the temperature range in which they are located. Number items falling in different temperature ranges are removed to establish a preferred excitation number set; Power mapping module: Based on the preferred excitation number set, the coordinate order value of each nozzle number in the pattern block position list is extracted, and the coordinate order value is matched with the nozzle power input value to establish a power comparison table. The nozzle numbers with a power of 0 are recorded in the comparison table as a non-excitation number set. After removing them, the remaining numbers are rearranged from high to low according to the power value to form a number power comparison structure sequence, and a nozzle power sorting sequence is established; Combination optimization module: Based on the nozzle power sorting sequence, a genetic algorithm is used to extract the frequency segment number corresponding to each number, construct a number frequency control group, perform three rounds of cross-grouping operations on all number pairs in the number frequency control group, record the absolute value of the difference between the frequency segment number and the power value in each round of grouping, calculate the sum of the difference values ​​in all number combinations, sort by the minimum sum of the difference values ​​and retain the top three number combinations, detect repeated numbers in these combinations and classify them into a stable number set, and search the frequency segment number list for candidate numbers with an absolute difference value of less than 10 for numbers that do not appear, so as to form a frequency power combination list; Instruction output module: Based on the frequency-power combination list, the numbered power value and the frequency number are paired to generate a printing control instruction index table, and the power value of each number in the index table is called as the signal amplitude parameter, and the frequency number is used as the frequency value parameter to construct a number and signal instruction binding sequence, and establish a nozzle excitation signal set after permutation and combination.

[0017] Compared with the prior art, the advantages and positive effects of the present invention are: 1. This invention achieves a one-to-one indexing between nozzle numbers and pattern block positions through structured decomposition of print task parameter groups, establishing a matching system with sequential attributes. This enhances the controllability of parameter management and the accuracy of scheduling. 2. This invention uses a long-short-term memory network to identify time series of frequency coverage and usage frequency, optimizing the activation number screening mechanism, reducing the activation intensity of unnecessary nozzles, and alleviating the problem of heat load accumulation caused by frequent activation. 3. This invention utilizes a genetic algorithm to implement multiple rounds of iterative replacement and screening of combination numbers. While ensuring the logical integrity of combination matching, it optimizes the frequency band allocation density and coverage balance, enhances the diversity and distribution efficiency of electrical stimulation frequencies, and achieves multi-variable collaborative optimization of nozzle excitation strategies, improving the stability of system response and the efficiency of signal excitation execution. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of the workflow of the present invention; Figure 2 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0021] See also Figure 1The present invention provides a technical solution: a method for stimulating nozzles using Kyocera printhead printing software, comprising the following steps: S1: Based on the print task parameter group, match the nozzle number with the pattern block position index, encode the number and position sequence, merge the inkjet frequency and historical excitation record segments, divide the nozzle temperature range, and generate the nozzle operation parameter set; S2: Based on the nozzle operating parameter set, a long short-term memory network is used to screen nozzle numbers and pattern block positions, classify frequency excitation segments, combine numbers and frequency excitations, perform combined matching metrics, screen numbers and matching results, compare nozzle temperature ranges, screen numbers, and generate an optimal excitation number set; S3: Based on the preferred excitation number set, the pattern block positions are sorted, and the nozzle numbers are mapped to the power inputs. The nozzle numbers with a power value of zero are marked and extracted to generate a power sorting list. S4: Based on the power sorting list, the power and frequency excitation segments are numbered and a genetic algorithm is used to construct a number and combination matching set. The combination matching results are compared and screened, and non-compliant combination replacements are generated. The iterative process is performed to generate a power-frequency matching table. S5: Based on the power-frequency matching table, read the numbered power and frequency configuration, load the printing control system, bind the numbered configuration instructions, output the signal amplitude and frequency, and generate the nozzle excitation execution instruction set.

[0022] The specific steps to generate the nozzle operation parameter set are: Based on the print task parameter group, the nozzle number is matched with the pattern block position index, the nozzle number is positioned in the segment, and the number and position sequence are encoded, and the number and segment sequence are sorted to generate the number position mapping data; Based on the number position mapping data, the inkjet frequency and the historical excitation record time segment are merged, the interval coverage is performed through the frequency data and the time period, and the frequency interval is divided. The numerical segmentation is performed through the frequency boundary conditions to generate the frequency excitation segment data; Based on the frequency excitation segment data, the nozzle temperature value interval is divided, and the temperature data is compared with the set interval boundary. The frequency excitation segments and temperature intervals are numbered and integrated. The combination is classified by numbering to generate the nozzle operation parameter set; Based on the printing task parameter group, the nozzle number is matched with the pattern block position index. By performing position segment positioning on the nozzle number, a sequential position mapping method is used to compare each nozzle number with the coordinate index of the pattern block in the two-dimensional printing area one by one, and the horizontal and vertical coordinate values ​​of the pattern block are extracted as a two-dimensional position point set. The horizontal values ​​in the coordinate point set are used to sort the positions from small to large, and an initial mapping list of the number and the pattern block position is constructed. The sorted numbers are cumulatively coded for adjacent position differences, and the coding result of each number is combined with its index position to generate a number position sequence vector. The number and position sequence are coded, and the sequence coding result is used to construct a number and pattern block position segment index table. The number and segment sequence is sorted to generate number position mapping data. Based on the number position mapping data, the inkjet frequency and the historical excitation record time segment are merged. The segment frequency fusion method is used to jointly judge each frequency value in the frequency data and the start and end time of the excitation time period of the corresponding nozzle number. The judgment condition is whether the frequency value appears more than 3 times in a row within the start and end time interval. The data that meets the conditions are grouped by number and the time period is merged to generate a preliminary merge result. Each frequency in the merged result is projected and divided according to the time axis. Every 30 seconds is a time segment. The number of occurrences of each frequency segment is counted. The interval coverage is performed by frequency data and time period. The frequency with a statistical number of less than 3 is eliminated and the frequency interval is divided. The fixed boundary division method is used to set the boundary conditions of the frequency value to 100Hz, 200Hz, 300Hz and 500Hz. The numerical segmentation is performed, the frequency is classified into the corresponding interval segment, and the frequency excitation segment data is generated. Based on the frequency excitation segment data, the nozzle temperature value interval is divided, and the temperature boundary comparison method is used to compare the temperature data value corresponding to each nozzle number with the set interval boundary. The set interval boundaries are 30℃, 50℃, 70℃ and 90℃ respectively. The temperature value is compared with the boundary value in turn, and the interval segment number corresponding to the nozzle number is marked. The number is integrated with the frequency excitation segment data and the temperature interval. The integrated data is classified by a combined classification method. The number, frequency segment number and temperature segment number are formed into triplets. All triplets are arranged in ascending order according to the joint number value of the frequency number and the temperature number, and are reorganized and classified according to the number dimension to generate a nozzle operation parameter set.

[0023] The printing task parameter group includes the nozzle number, pattern block position, inkjet frequency, historical excitation record, and nozzle temperature input in the printing software interface.

[0024] The specific steps for generating the preferred incentive number set are: Based on the nozzle operation parameter set, a long short-term memory network is used to screen the nozzle number and pattern block position. The number position is matched with the task segment, and the number validity is checked. The number data is compared with the available status to generate a valid number set. Based on the valid number set, the frequency excitation segments are classified, the number data is associated with the frequency segments, and the number and frequency excitation combination is established. The number sequence is connected with the frequency combination to generate the number excitation combination data; Based on the number incentive combination data, perform combination matching measurement, perform association scoring by number and frequency combination, screen the number and matching results, screen the numbers by comparing the matching degree with the temperature range, and generate the preferred incentive number set; Based on the nozzle operation parameter set, a long short-term memory network is used to screen the nozzle number and pattern block position. The network structure used includes an input layer, three hidden layers and an output layer. The input of each timing unit in the input layer is a three-dimensional tensor matrix consisting of a number sequence, a pattern block position index, and a frequency excitation identifier. The tensor dimension is M×T×3, where M is the total number of nozzle numbers and T is the task time step. The three hidden layers use a gated structure with a forget gate and an input gate respectively. A weighted operation is performed on the input state and the previous state in each time step. The activation function uses a joint mapping of tanh and sigmoid. The output layer structure is a softmax normalized structure. Matching is performed between the number position and the task segment. The relative distance between the number position and the horizontal and vertical coordinates of the pattern block is converted and normalized, and mapped to a floating-point number range of 0 to 1. For each number, the number data is compared with the Boolean value of the available state to determine whether it is in a logically true state. Numbers with false states are eliminated to generate a valid number set. Based on the valid number set, the frequency excitation segments are classified, and the number combination classification method is adopted for the frequency excitation data. A one-to-one combination is constructed through the number data and the frequency segment number. The number of each combination pair is used as the primary key and the corresponding frequency segment number is used as the value item. A hash mapping structure is established, and the frequency segment numbers in the mapping structure are combined and merged in the order of the numbers. The items with adjacent frequency number differences not exceeding 50Hz are merged to form a number frequency comparison pair. The number and frequency excitation combination establishment is then executed, and the corresponding number position sequence of the combined frequency segment number sequence is written into a two-way linked list structure. The binding relationship between the number pointer and the frequency sequence node is updated in sequence on the two-way linked list. The number sequence is connected with the frequency combination to generate the number excitation combination data. Based on the numbered excitation combination data, a combination matching metric is performed, and an associated matching scoring method is adopted. The associated scoring is performed through the combination of number and frequency. The frequency segment number, number index value, and pattern block coordinate value are used as three dimensions to construct a three-dimensional comparison matrix with the dimension of number number × number of frequency segments × number of coordinates. The squared Euclidean distance between the frequency number and the longitudinal coordinate of the pattern block is used as the scoring basis. The value is calculated for each number-frequency combination and stored in the scoring table. The numbers and matching results are screened, and the numbers with scoring values ​​higher than the set threshold of 0.7 in the scoring table are screened out. The remaining numbers are compared bidirectionally with the temperature range numbers in the nozzle operation parameter set. If the temperature range number is consistent, the number is retained to generate the preferred excitation number set.

[0025] Long short-term memory network, according to the formula: ; in: Represents the nozzle number matching state vector at the current moment, Represents the activation function, used for nonlinear mapping, represents the weight matrix between the input vector and the hidden state, represents the nozzle operating parameter input vector, Indicates the weight coefficient corresponding to the printing delay factor, Indicates the current printing delay factor of the nozzle. Indicates the weight coefficient corresponding to the pattern block complexity factor, represents the pattern block complexity factor, Indicates the weight coefficient corresponding to the task section priority coefficient, Indicates the task section priority coefficient, represents the input layer bias term, represents the weight matrix of the hidden state recurrent connection, represents the hidden state vector at the previous moment, represents the hidden layer bias term; Execution process: First construct the input vector , including the nozzle number, the position coordinates of the pattern block in the two-dimensional printing reference grid, the task segment index and the current nozzle printing rate, which are extracted and standardized in real time by the nozzle control module, and then the nozzle printing delay factor is obtained. , calculated based on the nozzle drive response time and the host computer control signal transmission interval, and normalized to enhance the network convergence stability, while extracting the pattern block complexity factor By calculating the composite index formed by the combination of the curvature change value of the pattern block boundary and the number of color jumps per unit area, the microscopic precision requirements of the printing task pattern are reflected, and then the priority coefficient of the task segment is obtained. , calculated based on the time sensitivity of the segment and resource competition output by the scheduler, three innovative parameters Multiply by the corresponding weights , obtained by using a cross-validation and grid search joint training method to minimize the loss function on the historical nozzle activation data, and then mapping all the above results to the input vector weight , the previous hidden state and its weight mapping and two biases Add, and finally pass the activation function Perform nonlinear transformation and output hidden state , which is used to determine whether the current nozzle number is applicable to the target pattern block and task segment, thereby generating a valid number set and driving the nozzle to work accurately.

[0026] The specific steps to generate the power sort list are: Based on the preferred excitation number set, the pattern blocks are sorted, the segment positioning is performed through the number position data, and the number sequence is adjusted, the number is aligned with the position sequence, and sequential recording is performed to generate the number sorting result; Based on the number sorting result, the nozzle number and the power input are mapped, the data is associated with the number sequence and the power input value, and the power input is marked as zero. The number is paired with the power state, and the number state is recorded to generate the number power mark data; Based on the numbered power tag data, non-zero power numbers are extracted, and screening is performed based on the numbers and power status. The numbers and power values ​​are arranged in sequence, and a power sorting list is generated by matching the power values ​​with the number sequence. Based on the preferred excitation number set, the pattern block positions are sorted, and the coordinate segment alignment method is adopted to perform segment positioning through the number position data. The horizontal and vertical values ​​in the pattern block position coordinates corresponding to each nozzle number are encoded and combined. The encoding method is to multiply the horizontal coordinate by 100 and add the vertical coordinate to form a unique positioning value. All numbers and positioning values ​​are formed into a key-value pair structure. The key-value pairs are arranged in ascending order according to the positioning value to obtain a sorted number list, and the number sequence is adjusted. The adjusted number list is indexed and compared with the original pattern block sequence. A mapping relationship is formed between all numbers and corresponding position coordinates. The numbers in the mapping relationship are realigned according to the increasing method of the pattern block horizontal coordinates, and are recorded in sequence to generate a number sorting result. Based on the number sorting results, the nozzle number and power input are mapped. The number power matching method is used to perform data association through the number sequence and the power input value. The number sequence is used as the primary key and the power value is used as the numerical field. A typical mapping structure is constructed, and all numbers and power values ​​are filled into the structure one by one. The status marking operation is performed on the numbers with zero power input values. The marking method is to set the Boolean field is_zero_power to True. The status check of this field is performed on each number in the mapping structure, and the marking result is written into the corresponding record of the number. By pairing the number with the power status, all numbers and their power identification status are extracted, and the number status record is performed to generate the number power marking data; 3. Based on the numbered power tag data, non-zero power numbers are extracted, and the power value filtering and sorting method is used. Screening is performed by number and power status. The screening condition is the record with the is_zero_power field value of False. The numbers that meet the condition are used as the valid power number set. The power value corresponding to each number in the set is extracted, and the numbers and power values ​​are arranged in sequence. The power value field is used as the main sorting basis and the number is used as the secondary sorting key. All records are sorted in ascending order to form a number list sorted by power value. By matching the power value with the number sequence, a parallel structure of the power value field and the number field is generated, an increasing power value sequence is generated, a number-corresponding power sorting queue is constructed, and a power sorting list is generated.

[0027] The specific steps to generate the power frequency matching table are: Based on the power sorting list, the numbered power and frequency excitation segments are combined, the numbered power and power input and frequency excitation data are paired, and the combination sequence is adjusted, and the numbered power frequency combination data is generated by synchronizing the numbering sequence with the combination sequence; Based on the number power frequency combination data, a genetic algorithm is used to compare the combination matching results, a matching degree analysis is performed through the number and combination data, and a matching combination screening is performed, and the number screening is performed through the matching degree data and the screening conditions to generate matching combination data; Based on the matching combination data, replace the matching combination, perform replacement by numbering and combination data, and perform iterative processing of the combination set, synchronously update the numbering and update combination data to generate a power frequency matching table; Based on the power sorting list, the numbered power and frequency excitation segments are combined. The numbered frequency power pairing method is adopted. The number is paired with the power input and frequency excitation data. The number field is used as the index key. The corresponding power input value field and the frequency excitation segment number field are combined into a key-value pair. A numbered drive item mapping table is constructed. The mapping table is sorted in ascending order according to the power value field. At the same time, the frequency excitation segment numbers are synchronously arranged in the numbering sequence. The number field is compared with the numbering sequence in the power sorting list to confirm the consistency of the numbering sequence. If it is inconsistent, the combination order is rearranged according to the power sorting order. All numbered power-frequency triplets are synchronized in sequence to generate numbered power-frequency combination data. Based on the number power frequency combination data, a genetic algorithm is used to compare the combination matching results. The genetic algorithm used sets the initial population size to 100, and the individual structure is an array consisting of a triplet of number, power value, and frequency number. The fitness function is the weighted average of the difference between the number and the frequency number. The difference threshold is 50Hz or less, which is set as excellent, and the value exceeding is set as poor. The single-point crossover operation is performed in the power value field, and the crossover probability is set to 0.8. The mutation operation is located in the frequency field, and the mutation range is the exchange between the numbers of two adjacent frequency segments. The mutation probability is set to 0.1. The matching degree analysis is performed by the number and combination data to generate a fitness score table for all individuals in each generation. Individuals with a score higher than 0.85 are marked as retained. Based on this, they are used as combinations that meet the retention conditions. The number screening is performed by matching the matching degree data and the screening conditions to generate matching combination data. Based on the conforming combination data, non-conforming combinations are replaced. The number combination replacement iterative method is adopted to perform replacement through the number and combination data. The number records with a matching score lower than 0.6 are used as the elimination sequence. The spare combination with the smallest frequency number difference and a power value deviation of no more than ±10% is extracted from the original number power-frequency combination data. The replacement operation is performed one by one on the elimination sequence numbers. After each round of replacement, the current combination integrity and number consistency identifier are recorded. If there is a vacancy or number duplication, the current round of replacement is terminated and rolled back to the last valid round, and the combination set is iterated. The upper limit of the iteration number is set to 50 times. After each round, all numbers are synchronized with the updated combination data to generate a functional equivalent number frequency comparison set and a power frequency matching table.

[0028] Genetic algorithm, according to the formula: ; Among them: Fitness(i) represents the matching score of the i-th number combination, Indicates the actual value of the i-th number combination on the j-th indicator, Represents the ideal setting value of the j-th indicator in the target combination data, represents the matching importance weight of the j-th indicator, represents the nozzle vibration stability fluctuation factor corresponding to the i-th number combination, represents the historical execution failure rate of the i-th number combination in the task path segment, Stability factor The global weight of Failure rate factor The global weight of Execution process: First, extract the power value, frequency parameter, printing temperature coefficient and synchronization delay time difference corresponding to each candidate nozzle number combination to form a vector , and obtain the ideal setting value in the target combination data , calculate the absolute value of the difference between the actual value and the target value for each dimension, and then multiply it by the preset matching weight These weights are determined by experimental tuning and matching accuracy sensitivity analysis, and then the power fluctuation data of the nozzle in multiple excitation cycles are collected and the disturbance mean square error is calculated as the vibration stability factor. Then, the number of failures of the combination of numbers in the nozzle control record in the recent task segment scheduling process is normalized to obtain the historical failure rate , multiply the two innovative error factors by the corresponding global weights and The two weight coefficients are determined by optimizing the training set using a five-fold cross-validation combined with a grid search method. Finally, the sum of all weighted error terms and constant terms is placed in the denominator structure, and the fitness score Fitness(i) of the current combination is calculated by the inverse function, which is used to screen and generate the optimal nozzle numbering scheme that meets the combination data.

[0029] The specific steps to generate the nozzle excitation execution instruction set are: Based on the power-frequency matching table, the numbered power and frequency configuration is read, the power value is called and the frequency value is extracted by the number, the position is matched by the number and the configuration index sequence, and the number is associated with the power-frequency data to generate the numbered configuration data; Based on the number configuration data, the printing control system is loaded, data is imported through the number and control parameters, and the number and control parameters are mapped. Binding is performed through the number sequence and the system control bit, and the number and binding parameters are updated synchronously to generate the number control binding data; Based on the number control binding data, the signal amplitude and frequency are output, the output instruction is generated through the number and binding parameters, and the signal output sequence is determined. The control relationship is established through the number and output sequence to generate the nozzle excitation execution instruction set; Based on the power-frequency matching table, the numbered power and frequency configuration is read. The configuration data index reading method is used to execute the power value call and extract the frequency value through the number. The number field is set as the primary index, the corresponding power field is set to value1, and the frequency field is set to value2. A three-field mapping structure is constructed, and the number value is used to locate the index position in the mapping structure. Value1 and value2 are extracted to complete the synchronous reading of the number and power-frequency value. The position is matched by the number and the configuration index sequence. The number order in the index sequence is compared one by one with the extracted number. If the order is inconsistent, the extracted value sequence is rearranged according to the index sequence order, and the number is associated with the power-frequency data. The rearranged number, power value and frequency value are combined into a key-value pair dictionary structure to generate the numbered configuration data. Based on the number configuration data, the printing control system is loaded, and the number control import method is adopted. Data import is performed through the number and control parameters to construct a control parameter structure. The structure fields include number id, control bit bit_index, amplitude, and frequency. For each number, its power value is bound to the amplitude field, the frequency value is bound to the frequency field, and the number value is bound to the id field. The bit_index list in the system controller is written in the order of numbers to form a complete control parameter table, and the number and control parameter mapping is performed. The hash mapping structure uses the number as the key and the control parameter structure as the value to achieve fast retrieval mapping. Binding is performed with the system control bit through the number sequence. For each bit_index in the control bit array, the corresponding structure address is filled in the order of numbers, and the number and binding parameters are updated synchronously. The old pointer in the system buffer is replaced with the latest control parameter structure address to generate number control binding data. Based on the number control binding data, the signal amplitude and frequency are output. The output instruction generation method is adopted, and the output instruction generation is executed through the number and binding parameters. The amplitude and frequency fields are extracted from the binding data, and the amplitude value is converted into a 16-bit DAC digital signal amplitude. The frequency value is converted into a PWM control period. The control instruction frame is combined one by one through the number index sequence. The format is [number id, high-order amplitude byte, low-order amplitude byte, frequency period byte]. The signal output sequence is determined, and a control relationship is established through the number and output sequence. The instruction frame is written into the output buffer and the frame is output in sequence according to the number sequence to generate a nozzle excitation execution instruction set.

[0030] See also Figure 2 A system for stimulating nozzles using Kyocera printhead printing software is provided. The system is used to perform the above-mentioned method for stimulating nozzles using Kyocera printhead printing software. The system comprises: Nozzle matching module: Based on the print task parameter group, the module calls the pattern block position coordinate value, nozzle number index value, inkjet frequency value, historical excitation record segment number, and nozzle temperature range value to complete the item-by-item correspondence between the nozzle number and the pattern block position coordinate value. The pattern block position number is rearranged and encoded according to the nozzle number sequence. The frequency number pairs with the same inkjet frequency value and the start and end time value in the historical excitation record segment number are selected to establish a frequency merging sequence. The nozzle temperature range value is numbered and marked in intervals of 10 degrees to generate a nozzle operation number set. Excitation screening module: Based on the nozzle operation number set, a long short-term memory network is used to extract the pairing items of pattern block position number and nozzle number. The frequency segment number is recorded for each pair of numbers. The number of occurrences of the frequency segment number corresponding to each nozzle number is counted. The frequency segment numbers are accumulated in sequence to form a total frequency coverage value sequence of the nozzle number. The relative frequency of occurrence of the nozzle number in the pattern block position is obtained and divided by the median of the total frequency coverage value sequence. The ratio is output and the nozzle numbers with a ratio greater than 1.0 are screened. The temperature range in which they are located is compared with the table, and the number items falling in different temperature ranges are removed to establish the optimal excitation number set; Power mapping module: Based on the preferred excitation number set, the coordinate order value of each nozzle number in the pattern block position list is extracted and matched with the nozzle power input value to establish a power comparison table. The comparison table records the nozzle numbers with a power of 0 as the non-excitation number set. After removing them, the remaining numbers are rearranged from high to low according to the power value to form a number power comparison structure sequence and establish a nozzle power sorting sequence; Combination optimization module: Based on the nozzle power sorting sequence, a genetic algorithm is used to extract the frequency segment number corresponding to each number, construct a number frequency control group, and perform three rounds of cross-grouping operations on all number pairs in the number frequency control group. The absolute value of the difference between the frequency segment number and the power value is recorded in each round of grouping. The sum of the difference values ​​in all number combinations is counted, and the number combinations with the smallest sum of the difference values ​​are sorted and the top three groups are retained. Repeated numbers in these combinations are detected and classified into a stable number set. For numbers that do not appear, candidate numbers with an absolute difference value of less than 10 are searched from the frequency segment number list to form a frequency power combination list; Instruction output module: Based on the frequency-power combination list, the numbered power values ​​are paired with the frequency numbers to generate a printing control instruction index table. The power value of each number in the index table is called as the signal amplitude parameter, and the frequency number is used as the frequency value parameter to construct a number and signal instruction binding sequence. After permutation and combination, the nozzle excitation signal set is established.

[0031] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A method for stimulating nozzles for Kyocera print head printing software, characterized in that: The following steps are involved: S1: Based on the print task parameter group, match the nozzle number with the pattern block position index, encode the number and position sequence, merge the inkjet frequency and historical excitation record segments, divide the nozzle temperature range, and generate the nozzle operation parameter set; S2: Based on the nozzle operating parameter set, a long short-term memory network is used to screen nozzle numbers and pattern block positions, classify frequency excitation segments, combine numbers and frequency excitations, perform combined matching metrics, screen numbers and matching results, compare nozzle temperature ranges, screen numbers, and generate a preferred excitation number set; S3: Based on the preferred excitation number set, the pattern blocks are sorted, and the nozzle numbers are mapped to the power inputs, the nozzle numbers with power values ​​of zero are marked, and the numbers are extracted to generate a power sorting list; S4: Based on the power sorting list, the power and frequency excitation segments are numbered and a genetic algorithm is used to construct a numbering and combination matching set, the combination matching results are compared and screened, and non-compliant combination replacements are generated, and the process is iterated to generate a power frequency matching table; S5: Based on the power frequency matching table, read the numbered power and frequency configuration, load the printing control system, bind the numbered configuration instructions, output the signal amplitude and frequency, and generate a nozzle excitation execution instruction set.

2. The method for stimulating nozzles for Kyocera print head printing software according to claim 1, characterized in that: The specific steps of generating the nozzle operation parameter set are: Based on the print task parameter group, the nozzle number is matched with the pattern block position index, the nozzle number is positioned in the segment, and the number and position sequence are encoded, and the number and segment sequence are sorted to generate the number position mapping data; Based on the number position mapping data, the inkjet frequency and the historical excitation record time segment are merged, the interval coverage is performed through the frequency data and the time period, and the frequency interval is divided, and the numerical value is segmented through the frequency boundary condition to generate the frequency excitation segment data; Based on the frequency excitation segment data, the nozzle temperature value interval is divided, and the temperature data is compared with the set interval boundary, and the numbering is integrated with the frequency excitation segment and the temperature interval. The combination classification is performed through the numbering to generate a nozzle operation parameter set.

3. The method for stimulating nozzles for Kyocera print head printing software according to claim 1, characterized in that: The printing task parameter group specifically includes the nozzle number, pattern block position, inkjet frequency, historical excitation record, and nozzle temperature input into the printing software interface.

4. The method for stimulating nozzles for Kyocera print head printing software according to claim 1, characterized in that: The specific steps of generating the preferred incentive number set are: Based on the nozzle operation parameter set, a long short-term memory network is used to screen the nozzle numbers and pattern block positions, match the number positions with the task segments, and perform number validity verification. The number data is compared with the available status to generate a valid number set. Based on the valid number set, frequency excitation segments are classified, number data and frequency segments are associated and combined, number and frequency excitation combinations are established, and number sequence and frequency combination are connected to generate number excitation combination data; Based on the number incentive combination data, a combination matching metric is performed, an association score is performed by combining the number and the frequency, and the number and the matching result are screened. The numbers are screened by comparing the matching degree with the temperature range to generate a preferred incentive number set.

5. The method for stimulating nozzles for Kyocera print head printing software according to claim 4, characterized in that: The long short-term memory network is based on the formula: ； in: Represents the nozzle number matching state vector at the current moment, Represents the activation function, used for nonlinear mapping, represents the weight matrix between the input vector and the hidden state, represents the nozzle operating parameter input vector, Indicates the weight coefficient corresponding to the printing delay factor, Indicates the current printing delay factor of the nozzle. Indicates the weight coefficient corresponding to the pattern block complexity factor, represents the pattern block complexity factor, Indicates the weight coefficient corresponding to the task section priority coefficient, Indicates the task section priority coefficient, represents the input layer bias term, represents the weight matrix of the hidden state recurrent connection, represents the hidden state vector at the previous moment, represents the hidden layer bias term.

6. The method for stimulating nozzles for Kyocera print head printing software according to claim 1, characterized in that: The specific steps of generating the power ranking list are: Based on the preferred excitation number set, pattern block positions are sorted, segment positioning is performed using number position data, and number sequence adjustment is performed, the numbers are aligned with the position sequence, and sequential recording is performed to generate a number sorting result; Based on the number sorting result, mapping the nozzle number and the power input is performed, data association is performed through the number sequence and the power input value, and the power input is marked as zero number, and the number is paired with the power state, and the number state is recorded to generate number power mark data; Based on the numbered power tag data, non-zero power numbers are extracted, screening is performed based on the numbers and power states, and the numbers and power values ​​are arranged in sequence. A power sorting list is generated by matching the power values ​​with the number sequence.

7. The method for stimulating nozzles for Kyocera print head printing software according to claim 1, characterized in that: The specific steps of generating the power frequency matching table are: Based on the power sorting list, the numbered power and frequency excitation segments are combined, the numbered power and the frequency excitation data are paired, and the combination sequence is adjusted, and the numbered power and the combination sequence are synchronized to generate the numbered power frequency combination data; Based on the number power frequency combination data, a genetic algorithm is used to compare the combination matching results, a matching degree analysis is performed through the number and the combination data, and a matching combination screening is performed, and the number screening is performed through the matching degree data and the screening conditions to generate matching combination data; Based on the compliant combination data, non-compliant combinations are replaced, the replacement is performed through numbering and combination data, and combination set iterative processing is performed, the numbering and updated combination data are synchronously updated to generate a power frequency matching table.

8. The method for stimulating nozzles using Kyocera printhead printing software according to claim 7, characterized in that: The genetic algorithm is based on the formula: ； Among them: Fitness(i) represents the matching score of the i-th number combination, Indicates the actual value of the i-th number combination on the j-th indicator, Represents the ideal setting value of the j-th indicator in the target combination data, represents the matching importance weight of the j-th indicator, represents the nozzle vibration stability fluctuation factor corresponding to the i-th number combination, represents the historical execution failure rate of the i-th number combination in the task path segment, Stability factor The global weight of Failure rate factor The global weight of .

9. The method for stimulating nozzles using Kyocera printhead printing software according to claim 1, wherein: The specific steps of generating the nozzle excitation execution instruction set are: Based on the power-frequency matching table, read the numbered power and frequency configuration, call the power value and extract the frequency value by number, make position correspondence between the number and the configuration index sequence, associate the number with the power-frequency data, and generate numbered configuration data; Based on the numbering configuration data, the printing control system is loaded, data is imported through the numbering and control parameters, and the numbering and control parameters are mapped, binding is performed through the numbering sequence and the system control bit, and the numbering and binding parameters are synchronously updated to generate numbering control binding data; Based on the numbering control binding data, the signal amplitude and frequency are output, the output instruction is generated through the numbering and binding parameters, and the signal output sequence is determined. The control relationship is established through the numbering and output sequence to generate the nozzle excitation execution instruction set.

10. A system for stimulating nozzles using Kyocera printhead software, characterized in that: The method for stimulating nozzles using Kyocera printhead printing software according to any one of claims 1 to 9, wherein the system comprises: Nozzle matching module: Based on the print task parameter group, the module calls the pattern block position coordinate value, nozzle number index value, inkjet frequency value, historical excitation record segment number, and nozzle temperature range value to complete the item-by-item correspondence between the nozzle number and the pattern block position coordinate value. The pattern block position number is rearranged and encoded according to the nozzle number sequence. The frequency number pairs with the same inkjet frequency value and the start and end time value in the historical excitation record segment number are selected to establish a frequency merging sequence. The nozzle temperature range value is numbered and marked in intervals of 10 degrees to generate a nozzle operation number set. Excitation screening module: Based on the nozzle operation number set, a long short-term memory network is used to extract pairings of pattern block position numbers and nozzle numbers. The frequency segment number is recorded for each pair of numbers. The number of occurrences of the frequency segment number corresponding to each nozzle number is counted. The frequency segment numbers are accumulated in sequence to form a nozzle number frequency coverage total value sequence. The relative frequency of occurrence of the nozzle number in the pattern block position is obtained, and the result is divided by the median of the frequency coverage total value sequence to output a ratio. Nozzle numbers with a ratio greater than 1.0 are screened, and a table comparison is performed on the temperature range in which they are located. Number items falling in different temperature ranges are removed to establish a preferred excitation number set; Power mapping module: Based on the preferred excitation number set, the coordinate order value of each nozzle number in the pattern block position list is extracted, and the coordinate order value is matched with the nozzle power input value to establish a power comparison table. The nozzle numbers with a power of 0 are recorded in the comparison table as a non-excitation number set. After removing them, the remaining numbers are rearranged from high to low according to the power value to form a number power comparison structure sequence, and a nozzle power sorting sequence is established; Combination optimization module: Based on the nozzle power sorting sequence, a genetic algorithm is used to extract the frequency segment number corresponding to each number, construct a number frequency control group, perform three rounds of cross-grouping operations on all number pairs in the number frequency control group, record the absolute value of the difference between the frequency segment number and the power value in each round of grouping, calculate the sum of the difference values ​​in all number combinations, sort by the minimum sum of the difference values ​​and retain the top three number combinations, detect repeated numbers in these combinations and classify them into a stable number set, and search the frequency segment number list for candidate numbers with an absolute difference value of less than 10 for numbers that do not appear, so as to form a frequency power combination list; Instruction output module: Based on the frequency-power combination list, the numbered power value and the frequency number are paired to generate a printing control instruction index table, and the power value of each number in the index table is called as the signal amplitude parameter, and the frequency number is used as the frequency value parameter to construct a number and signal instruction binding sequence, and establish a nozzle excitation signal set after permutation and combination.

Images