Intelligent maintenance method and equipment for industrial printer and storage medium

Abstract

The invention discloses an intelligent maintenance method and device for an industrial printer and a storage medium. The method comprises the following steps that S1, circuits of printer components are modularly integrated; S2, sensors of the printer components are modularly integrated; S3, feedback circuits of a nozzle circuit module, a paper feeding circuit module and an ink box circuit module are arranged; s4, setting sensor feedback signal channels of the nozzle group, the paper feeding group and the ink box group; s5, establishing a real-time monitoring and analyzing system for a feedback circuit and a feedback signal; S6, establishing an analysis model; the problem that an existing printer detection system cannot predict printer faults and fault prognosis is solved, the purpose of intelligent printer maintenance integrating maintenance and detection is achieved, and through circuit and sensor modular integration and feedback setting, the state of each component can be accurately captured in real time, and abnormity can be found in time.

Description

Technical Field

[0001] The present invention relates to the field of printers, and in particular to an intelligent maintenance method, equipment and storage medium for industrial printers. Background Art

[0002] In terms of the maintenance of industrial printers, there are currently a variety of common technical solutions and maintenance methods. However, these existing technical solutions have exposed certain limitations in actual application.

[0003] Many industrial printers rely mainly on manual inspection and maintenance operations during their daily maintenance. Professional technicians need to regularly check each component of the printer one by one, such as whether the nozzle is clogged, whether the ink cartridge or toner cartridge is in normal condition, the wear of mechanical parts, and the operating status of the electrical system. However, this manual maintenance method has many disadvantages. On the one hand, its efficiency is extremely low, and manual inspection requires a lot of time and manpower costs, especially for multiple industrial printers in a large-scale production environment. Manual inspection one by one will seriously affect production efficiency. On the other hand, manual operation is easily affected by human errors, and omissions or misjudgments may occur during the inspection process, which makes it impossible to accurately detect problems with the printer, resulting in potential faults that cannot be handled in a timely manner.

[0004] Some industrial printers only use a single maintenance tool or software during maintenance. These tools or software often have relatively simple functions and can only monitor or analyze certain specific aspects of the printer. For example, some tools can only detect the remaining ink volume of the ink cartridge, but cannot fully monitor other key factors such as the temperature of the print head, the printing speed, and the operating status of mechanical parts. This single-tool maintenance method makes it difficult to fully understand the overall status of the printer and cannot promptly discover some hidden faults in other unmonitored aspects, thus affecting the normal operation and service life of the printer.

[0005] Most of the existing technical solutions fail to achieve real-time monitoring of the status of industrial printers. Usually, the printer is inspected and analyzed only at specified intervals or when it has obvious signs of failure. During the normal operation of the printer, there is a lack of continuous and real-time status monitoring mechanism. This leads to the inability to detect abnormal conditions in time when some sudden and progressive failures occur inside the printer, and thus it is impossible to take effective solutions in the early stage of the failure. The failure is often not discovered until it has seriously affected the normal operation of the printer, or even caused production interruption, which seriously affects the continuity and efficiency of production.

[0006] At present, the maintenance of industrial printers mostly adopts a passive maintenance method, that is, the corresponding repair and processing work is mainly carried out after the printer has failed. This passive maintenance mode has obvious defects. First, when the printer fails, it needs to be shut down for maintenance, which will lead to a long downtime, directly affecting the production progress and causing economic losses to the enterprise. Secondly, since the failure has already occurred, the maintenance process may be more complicated, requiring more manpower, material resources and time costs, resulting in high overall maintenance costs.

[0007] Due to the lack of real-time monitoring mechanism, existing technologies often fail to respond quickly when printer problems occur. Once a printer fails, such as a nozzle blockage or motor failure, effective countermeasures cannot be taken immediately because the abnormality cannot be detected in time, resulting in production interruption, which seriously affects production efficiency and the economic benefits of the enterprise.

[0008] Maintenance methods that rely on manual labor and a single tool make maintenance costs high. Manual maintenance requires the labor costs of professional technicians, and manual operation is inefficient and time-consuming. A single tool cannot fully monitor the status of the printer, resulting in the need to call different tools or professionals for further inspection multiple times, adding additional expenses. Especially when professional technicians are required to perform on-site services, such as dealing with some complex electrical system failures or mechanical component damage, the cost will be higher.

[0009] Most existing technical solutions fail to establish an effective fault prediction mechanism and cannot accurately predict the potential faults of the printer. During the operation of the printer, the status of its various components is constantly changing, and some potential fault hazards may gradually form. However, due to the lack of effective analysis and prediction capabilities of operating data, it is impossible to know in advance which components may fail, and thus it is impossible to take measures in advance to prevent the failure from occurring. It can only be handled passively after the failure occurs.

[0010] Although a certain amount of operating data is collected during the operation of the printer, such as printing speed, nozzle temperature, and ink cartridge status, existing technical solutions often do not fully utilize these data for in-depth analysis and fault prediction. These data contain rich information about the operating status and potential faults of the printer, but due to the lack of appropriate data analysis methods and tools, or the lack of an effective data processing process, these data are simply recorded and fail to play an important role in fault prediction and optimization of maintenance strategies.

[0011] Therefore, there is an urgent need for an intelligent maintenance method for industrial printers that can detect and predict printer failures in real time. Summary of the invention

[0012] In order to solve the above-mentioned technical problems, the present invention provides an intelligent maintenance method, device and storage medium for an industrial printer.

[0013] The technical solution of the present invention is achieved in this way:

[0014] Provided is an intelligent maintenance method for an industrial printer, comprising the following steps:

[0015] S1, modularizing and integrating the circuits of the printer components, wherein the integrated printer circuits include a nozzle circuit module, a paper feed circuit module, and an ink cartridge circuit module;

[0016] S2, modularly integrating the sensors of the printer components, wherein the integrated printer component sensors include a nozzle group, a paper feed group, and an ink cartridge group;

[0017] S3, setting feedback circuits of the nozzle circuit module, paper feeding circuit module and ink cartridge circuit module;

[0018] S4, setting sensor feedback signal channels of the printhead group, paper feed group and ink cartridge group;

[0019] S5, establish a real-time monitoring and analysis system for feedback circuits and feedback signals, through which the information fed back by each module and corresponding sensor can be accurately captured, so as to timely discover abnormal conditions that may occur during the operation of the printer, and establish a printer log to record the parameters detected by the printer in real time;

[0020] S6, establishing an analysis model, and setting a preset threshold value for the detection parameters of the printer according to the type of the printer, normalizing the information of the feedback circuit and the feedback signal, extracting abnormal parameters, and analyzing the abnormal parameters;

[0021] S7, automatically adjusting the printer according to the analysis result and predicting the adjustment result and time, and sending a manual adjustment notice when the predicted value is lower than the preset value, and sending the manual adjustment steps to the maintenance personnel;

[0022] S8, according to S5 and S6, after the printer is turned on, if no abnormal parameters are received, record it in the printer log;

[0023] S9, perform daily maintenance on the printer and reset and adjust all parameters of the printer.

[0024] The nozzle group centrally integrates the nozzle temperature sensor and the nozzle pressure sensor, and the sensor data of the integrated nozzle group are uploaded at the same time, and the sensors are reasonably arranged so that they can accurately sense the working status of the nozzle;

[0025] The paper feeding group integrates the paper feed sensor and paper discharge sensor in the paper feeding channel and the paper jam sensor in the paper feeding channel, and the integrated sensor data are uploaded at the same time. These sensors work together to fully monitor all aspects of the paper feeding process;

[0026] The ink cartridge group integrates the sensors on the ink cartridge, including the ink cartridge liquid level sensor and the ink cartridge chip status sensor, and the integrated sensor data are uploaded at the same time; by reasonably arranging these sensors, the status of the ink cartridge can be fully understood.

[0027] Preferably, in step S3, a nozzle feedback circuit is provided, and a current detection resistor is connected in series to the main circuit of the nozzle circuit to convert the current signal into a voltage signal and then feed it back to the control system.

[0028] Preferably, in step S6, various information from the feedback circuit and feedback signal are normalized, and parameters of different magnitudes and types are unified into a comparable scale range. For different parameters such as the nozzle drive voltage, the paper feed motor speed, and the remaining ink amount in the ink cartridge, the value range is calculated to be between 0 and 1 or other set standard ranges. The normalization method used can be Z-score normalization or Min-Max normalization method.

[0029] Preferably, in step S7, based on the analysis of the current operating state of the printer and the automatic adjustment measures that have been taken, a mathematical model or an empirical formula is used to predict the result and required time after adjustment; by analyzing the relationship between the remaining amount of ink and the printing mode and resolution, it is predicted how long the ink cartridge can support the printing task after switching the printing mode or lowering the resolution, and the predicted result and time are recorded in the printer log for subsequent review and evaluation of the adjustment effect.

[0030] A printing device comprises at least one processor, at least one memory and computer program instructions stored in the memory. When the computer program instructions are executed by the processor, the above-mentioned method is implemented.

[0031] A storage medium stores computer program instructions, which implement the above-mentioned method when the computer program instructions are executed by a processor.

[0032] The present invention solves the problem that the current printer detection system cannot predict printer failures and failure prognosis, and achieves the purpose of intelligent maintenance of printers that combines maintenance and detection. In the present invention, through modular integration of circuits and sensors and feedback settings, the status of each component can be accurately captured in real time, and anomalies can be discovered in time. By establishing an analysis model to normalize information, etc., abnormal parameters can be effectively extracted and analyzed to help accurately judge faults, and the printer can be automatically controlled according to the analysis results. Adjustment situations can also be predicted, manual intervention can be reduced, and efficiency can be improved. Various parameters are recorded through printer logs to provide detailed basis for maintenance, performance evaluation, etc., covering daily maintenance and parameter reset adjustment, which is conducive to keeping the printer in good condition, extending its service life, and improving the overall stability and reliability of printer operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a schematic diagram of an intelligent maintenance method for an industrial printer according to the present invention. DETAILED DESCRIPTION

[0034] In order to further explain the technical means and effects taken by the present invention to achieve the predetermined invention purpose, exemplary embodiments will be described in detail here, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are only examples of methods and systems consistent with some aspects of the present application as detailed in the attached claims.

[0035] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0036] The following is a detailed description of the specific implementation methods, features and functions of the present invention in conjunction with the accompanying drawings and preferred embodiments.

[0037] Example 1

[0038] like Figure 1 As shown, the present invention provides an intelligent maintenance method for an industrial printer, comprising the following steps:

[0039] S1, modularizing and integrating the circuits of the printer components, wherein the integrated printer circuits include a nozzle circuit module, a paper feed circuit module, and an ink cartridge circuit module;

[0040] S2, modularly integrating the sensors of the printer components, wherein the integrated printer component sensors include a nozzle group, a paper feed group, and an ink cartridge group;

[0041] S3, setting feedback circuits of the nozzle circuit module, paper feeding circuit module and ink cartridge circuit module;

[0042] S4, setting sensor feedback signal channels of the printhead group, paper feed group and ink cartridge group;

[0043] S5, establish a real-time monitoring and analysis system for feedback circuits and feedback signals, through which the information fed back by each module and corresponding sensor can be accurately captured, so as to timely discover abnormal conditions that may occur during the operation of the printer, and establish a printer log to record the parameters detected by the printer in real time;

[0044] S6, establishing an analysis model, and setting a preset threshold value for the detection parameters of the printer according to the type of the printer, normalizing the information of the feedback circuit and the feedback signal, extracting abnormal parameters, and analyzing the abnormal parameters;

[0045] S7, automatically adjusting the printer according to the analysis result and predicting the adjustment result and time, and sending a manual adjustment notification and manual adjustment steps when the predicted value is lower than the preset value;

[0046] S8, according to S5 and S6, after the printer is turned on, if no abnormal parameters are received, record it in the printer log;

[0047] S9, perform daily maintenance on the printer and reset and adjust all parameters of the printer.

[0048] Preferably, in step S1,

[0049] The nozzle circuit module includes a nozzle current detection module, a nozzle voltage detection module and a nozzle control module. The nozzle control module receives feedback information from the current, voltage, resistance detection module and the printer control system, and performs real-time control on the operation state of the nozzle circuit according to a preset parameter range and printing task requirements;

[0050] The paper feeding circuit module includes a paper feeding current detection module, a paper feeding voltage detection module and a paper feeding control module. The paper feeding control module receives feedback information from the current, voltage, speed detection module and the printer control system, and controls the operation state of the paper feeding circuit in real time according to the preset parameter range and printing task requirements;

[0051] The ink cartridge circuit module includes an ink cartridge current detection module, an ink cartridge voltage detection module and an ink cartridge control module. The ink cartridge control module receives feedback information from the current, voltage, and speed detection modules and the printer control system, and performs real-time control on the operating state of the ink cartridge circuit according to a preset parameter range and printing task requirements.

[0052] In the nozzle control module, paper feeding control module and ink cartridge control module, when it is detected that the current or voltage signal exceeds the preset limit threshold, the parameter is automatically adjusted with the highest priority adjustment level;

[0053] Preferably, in step S2,

[0054] The nozzle group integrates various sensors related to the nozzle, nozzle temperature sensor, nozzle pressure sensor, into a nozzle group, and arranges the sensors reasonably so that they can accurately sense the working status of the nozzle;

[0055] The paper feeding group integrates various sensors in the paper feeding channel, such as the paper feed sensor, the paper discharge sensor, and the paper jam sensor in the paper feeding channel, into a paper feeding group. These sensors work together to comprehensively monitor all aspects of the paper feeding process.

[0056] The ink cartridge group integrates the relevant sensors on the ink cartridge, the ink cartridge liquid level sensor and the ink cartridge chip status sensor to form the ink cartridge group. By properly arranging these sensors, the status of the ink cartridge can be fully understood.

[0057] Preferably, in step S3, a nozzle feedback circuit is set up, and a current detection resistor is connected in series to the main circuit of the nozzle circuit to convert the current signal into a voltage signal and then feed it back to the control system. When the nozzle current changes abnormally, such as overcurrent or undercurrent, the control system can take corresponding measures based on the feedback signal, such as adjusting the output parameters of the nozzle driver chip or issuing an alarm signal; in addition, a paper feed channel state feedback circuit is set up, and the voltage and current instantaneous variables of the paper feed motor are used as feedback signals to timely feed back to the control system whether there is a paper jam in the paper feed channel and whether the paper passes smoothly. When an abnormal paper jam occurs, the control system can immediately take measures to stop paper feeding and issue an alarm signal; for the ink cartridge circuit module, an ink pump feedback circuit is mainly set up to check the current and voltage usage of the ink pump. When the positive ink pump is working normally and the parameters of the ink cartridge liquid level sensor are normal, but ink cannot be discharged, it proves from the side that there is an abnormality in the ink supply pipeline or nozzle, and it can be checked.

[0058] Preferably, in step S5, the system is responsible for receiving, analyzing and processing signals from the feedback circuit and the sensor feedback signal channel, including the following sub-steps:

[0059] S51, identifying feedback signals from different sources, by setting different signal identifiers or address codes, distinguishing whether the feedback signal is from the nozzle circuit module, the paper feed circuit module or the ink cartridge circuit module, and whether the sensor feedback signal is from the nozzle group, the paper feed group or the ink cartridge group;

[0060] S52, establish a real-time data cache mechanism to temporarily store the received signals in the cache area for subsequent batch analysis and processing; S53, implement a multi-task processing function to ensure that while monitoring and analyzing the feedback signals, it does not affect the execution of other normal operating functions of the printer, such as the processing of print tasks and the response to user instructions;

[0061] S54, establish a printer log function to record the various parameters detected by the printer through real-time monitoring of the feedback circuit and feedback signal. The log content includes but is not limited to the timestamp, the voltage and current parameters of the nozzle circuit module, the temperature and pressure parameters of the nozzle group sensor, the speed and channel status parameters of the paper feed circuit module, the paper feed, paper output and paper jam conditions of the paper feed group sensor, the ink remaining amount of the ink cartridge circuit module and the parameters of the ink cartridge chip status;

[0062] S55, selected according to the specific configuration and needs of the printer, ensures that the log data can be stored for a long time and can be easily queried, analyzed and exported when needed, providing a strong basis for printer maintenance, troubleshooting, etc.

[0063] Preferably, in step S6, various information from the feedback circuit and feedback signal are normalized, and parameters of different magnitudes and types are unified into a comparable scale range. For different parameters such as the nozzle drive voltage, the paper feed motor speed, and the remaining ink amount in the ink cartridge, the value range is calculated to be between 0 and 1 or other set standard ranges. The normalization method used can be Z-score normalization or Min-Max normalization method.

[0064] Preferably, in step S6, an abnormal threshold is preset according to the printer model, and is determined based on factors such as the performance parameters and historical data of the printer. For example, for the nozzle temperature, according to the design specifications of the printer and previous operating experience, it is set that when the temperature exceeds 40°C, it is considered an abnormal situation; for the paper feed motor speed, when the speed is lower than 80% of the set target speed or higher than 120%, it is considered an abnormal situation;

[0065] Parameters exceeding the abnormal threshold are extracted as abnormal parameters, and these abnormal parameters are deeply analyzed. The analysis content includes but is not limited to the frequency, duration, and correlation with other parameters of the abnormal parameters. By analyzing these contents, the operating status of the printer can be further understood and the possible hidden faults or problems can be determined. For example, if the temperature of the printhead is abnormal and lasts for a long time, and there is a certain correlation with the fluctuation of the printhead drive voltage, it may indicate that there is a problem with the heating system or drive circuit of the printhead.

[0066] Preferably, in step S7, based on the analysis of the current operating state of the printer and the automatic adjustment measures that have been taken, a mathematical model or an empirical formula is used to predict the result and required time after adjustment; by analyzing the relationship between the remaining amount of ink and the printing mode and resolution, it is predicted how long the ink cartridge can support the printing task after switching the printing mode or lowering the resolution, and the predicted result and time are recorded in the printer log for subsequent review and evaluation of the adjustment effect.

[0067] Preferably, the daily maintenance includes regular cleaning of the printer, including cleaning the nozzle, paper path, ink cartridge installation location, etc. For inkjet printers, use a special cleaning fluid to clean the nozzle to remove dried ink, dust and other impurities on the nozzle surface to keep the nozzle in good working condition; clean paper scraps, dust and other debris in the paper path to ensure that paper can pass smoothly; check whether the ink cartridge is firmly installed and whether there is ink leakage on the surface of the ink cartridge.

[0068] Preferably, in step S2, a visual recognition sensor is also included to check the wear of the mechanical parts of the printer, the paper feed motor, the belt, and the roller. If obvious signs of wear are found in the parts, such as wear on the shaft of the paper feed motor, loose belts, thinning of the rubber layer on the roller surface, etc., the corresponding parts are replaced or repaired in time to ensure the normal operation of the printer.

[0069] A printing device comprises at least one processor, at least one memory and computer program instructions stored in the memory, and when the computer program instructions are executed by the processor, the method described above is implemented.

[0070] A storage medium stores computer program instructions, which implement the above-mentioned method when the computer program instructions are executed by a processor.

[0071] Example 2

[0072] An intelligent maintenance method for an industrial printer comprises the following steps:

[0073] S1 collects data from various components and operating links of the printer, including nozzle temperature, ink remaining in the ink cartridge, paper feed speed, motor current, voltage and other parameters. At the same time, it records the operating status of the printer, such as normal operation, paper jam, nozzle blockage and other corresponding data.

[0074] For example, by installing a temperature sensor near the nozzle to continuously obtain temperature data, setting a liquid level sensor on the ink cartridge to obtain information on the remaining ink amount, etc.

[0075] S2, clean the collected data to remove abnormal values ​​and erroneous data. For example, if the nozzle temperature sensor has an obviously unreasonable extremely high value due to a fault at a certain moment, it will be removed.

[0076] Data standardization is performed to make parameters of different magnitudes comparable. Methods such as Z-score standardization can be used to convert data into a standard form with a mean of 0 and a standard deviation of 1 so that the neural network can better process it.

[0077] S3, establish a neural network, and use a multi-layer perceptron (MLP) as the basic architecture. Set the input layer, hidden layer, and output layer according to the printer detection and maintenance requirements. The number of input layer nodes corresponds to the number of different parameters collected. For example, if there are 10 different printer parameters collected, the input layer will have 10 nodes.

[0078] The number of hidden layers and nodes can be determined through experiments and optimization. Generally, a small number of layers and nodes are set first, such as 2-3 hidden layers, with 10-50 nodes per layer, and then adjusted according to the training effect. The number of nodes in the output layer is determined according to the result to be predicted. For example, to predict whether the printer is faulty (normal / faulty), the output layer is set with 1 node, and the activation function outputs 0 or 1 to represent different states.

[0079] S4, select appropriate activation functions in the hidden layer and output layer. For example, the ReLU (Rectified Linear Unit) activation function in the hidden layer can improve the model training efficiency. The output layer can select the sigmoid function (for binary classification problems, such as determining whether there is a fault) or the softmax function (for multi-classification problems, such as determining different types of faults) according to the type of prediction results.

[0080] Select an optimization algorithm to adjust network parameters, such as stochastic gradient descent (SGD) and its variants Adagrad, Adadelta, etc. Select a suitable optimization algorithm according to the characteristics of the training data and the convergence of the model to improve the training effect.

[0081] S5, build a data model, divide the preprocessed data set into a training set and a test set according to a certain ratio, and the common ratio is 80% training set and 20% test set. For example, 1000 sets of printer data are collected, 800 sets are used for training, and 200 sets are used for testing model performance.

[0082] S6, model training, inputs the input data in the training set (i.e., various printer parameters) into the neural network, and the corresponding target output (such as whether the printer is faulty and the type of fault) is used as supervision information.

[0083] The neural network performs forward propagation to calculate the prediction result based on the input data and target output, and then calculates the error between the prediction result and the actual result through the back propagation algorithm, and adjusts the parameters such as weights and biases in the network according to the error. This process is repeated until the model converges, that is, the error reaches an acceptable small value or meets the set number of training rounds.

[0084] S7, model application, during the normal operation of the printer, the parameter data of each component is collected in real time and preprocessed, and then input into the trained neural network model.

[0085] The model outputs the prediction results of the current status of the printer based on the input data, such as whether there is a risk of failure, the possible type of failure, etc. For example, if the model output shows that the nozzle has a high probability of clogging, corresponding measures can be taken in time.

[0086] S8, customized maintenance strategy, formulates maintenance strategy based on the prediction results of the neural network model. If an impending failure is predicted, arrange maintenance personnel to inspect and repair in advance, prepare the required replacement parts, etc.

[0087] For some parameters that can be adjusted automatically, such as paper feed speed and nozzle drive voltage, they can be automatically adjusted based on the model prediction results and preset rules to try to eliminate potential fault hazards and extend the normal operation time of the printer.

[0088] Preferably, in step S2, the method for identifying outliers is:

[0089] S21, based on statistical rules: In addition to simply checking for obviously unreasonable extremely high or extremely low values, the range of outliers can also be determined by calculating the mean, standard deviation and other statistical quantities of the data. For example, for a certain parameter data, the value that exceeds the mean ±3 times the standard deviation is set as an outlier. Taking the nozzle temperature data as an example, if the mean of the normal working temperature range of the nozzle is 30°C and the standard deviation is 5°C, then data points with temperatures below 15°C or above 45°C may be considered outliers.

[0090] S22, based on box plot: draw box plots of each parameter data, and the values ​​outside the upper and lower whiskers in the box plot are usually considered to be outliers. The upper and lower limits of the box plot are determined by the first quartile (Q1) minus 1.5 times the interquartile range (IQR = Q3-Q1) and the third quartile (Q3) plus 1.5 times the interquartile range. In this way, outliers in the data can be more comprehensively identified, especially for data with a certain distribution pattern.

[0091] Preferably, in step S2, the abnormal data needs to be processed, and the abnormal value processing method is:

[0092] S23, when a data point is determined to be an abnormal value and it is judged to be caused by one-time factors such as sensor failure, momentary interference, etc., it can be directly deleted from the data set. For example, if the motor current value suddenly reaches a maximum value at a certain moment, far exceeding the normal working current range, and it is found through inspection that it is caused by the sensor being subjected to temporary electromagnetic interference, then the abnormal value data point can be deleted.

[0093] S24, if the correct value of the abnormal value can be reasonably inferred based on the context of the data or other relevant information, it can be corrected and replaced. For example, the ink remaining amount data of the ink cartridge has an obviously unreasonable low value, but combined with the previous consumption trend and the printer's working hours and other information, a reasonable remaining amount value can be estimated. In this case, the abnormal value can be replaced with the estimated value;

[0094] Preferably, in step S6, when the training set is divided into the data set according to the ratio of 80% training set and 20% test set, in order to make the division more reasonable, a stratified sampling method may be considered. That is, the data is stratified according to the different operating states of the printer (such as normal operation, paper jam, nozzle clogging, etc.), and then the training set and test set are extracted in proportion in each layer to ensure that the training set and the test set can better cover the samples of various printer operating states, avoid the situation where the data distribution deviation of the training set and the test set is too large, thereby improving the generalization ability of the model. For example, if stratified sampling is not performed, most of the data in the training set may be in normal working state, while there are more data in abnormal states such as paper jam or nozzle clogging in the test set, which will cause the model to perform poorly when facing various situations in actual applications.

[0095] Preferably, when the input data (i.e., various printer parameters) in the training set are input into the neural network, the data first enters the input layer. Each node in the input layer corresponds to a printer parameter, and these nodes pass the data to the hidden layer. After being processed by the hidden layer, the data continues to be passed to the output layer. The output layer also calculates based on the data received from the hidden layer and its own weights and biases, and finally outputs the prediction results. For example, for the binary classification problem of judging whether the printer is faulty, if the output layer uses a sigmoid function, the output value is between 0 and 1, and the printer can be judged whether it is faulty according to the set threshold (such as 0.5). A value greater than 0.5 indicates a fault, and a value less than 0.5 indicates normal. After the prediction result is calculated by forward propagation, the prediction result needs to be compared with the corresponding target output (such as whether the printer is faulty and the type of fault, etc.) to calculate the error. For the binary classification problem, the commonly used error calculation method is the cross entropy loss function. After the error is calculated, the error is back-propagated from the output layer to the input layer through the back propagation algorithm to adjust the parameters such as weights and biases in the network. During the back-propagation process, according to the chain rule, the contribution of each node to the error is calculated in turn to determine how each weight and bias should be adjusted. Usually, the model can be judged whether it has converged by observing the changes in the loss function value (such as the cross entropy loss function value) during the training process. If the loss function value no longer decreases significantly after several consecutive rounds of training, or drops below a pre-set threshold (such as 0.01), the model can be considered to have converged. In addition, the changes in performance indicators (such as accuracy, recall, etc.) on the validation set (i.e., test set) can also be observed at the same time. When these indicators also reach a good level on the validation set and no longer change significantly, it can also be judged that the model has converged. Once the model converges, training can be stopped and the trained model can be used for subsequent printer detection and maintenance applications.

[0096] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. An intelligent maintenance method for an industrial printer, characterized in that: The steps include: S1, modularizing and integrating the circuits of the printer components, wherein the integrated printer circuits include a nozzle circuit module, a paper feed circuit module, and an ink cartridge circuit module; S2, modularly integrating the sensors of the printer components, wherein the integrated printer component sensors include a nozzle group, a paper feed group, and an ink cartridge group; S3, setting feedback circuits of the nozzle circuit module, paper feeding circuit module and ink cartridge circuit module; S4, setting sensor feedback signal channels of the printhead group, paper feed group and ink cartridge group; S5, establish a real-time monitoring and analysis system for feedback circuits and feedback signals, through which the information fed back by each module and corresponding sensor can be accurately captured, so as to timely discover abnormal conditions that may occur during the operation of the printer, and establish a printer log to record the parameters detected by the printer in real time; S6, establishing an analysis model, and setting a preset threshold value for the detection parameters of the printer according to the type of the printer, normalizing the information of the feedback circuit and the feedback signal, extracting abnormal parameters, and analyzing the abnormal parameters; S7, automatically adjusting the printer according to the analysis result and predicting the adjustment result and time, and sending a manual adjustment notification and manual adjustment steps when the predicted value is lower than the preset value.

2. The intelligent maintenance method for an industrial printer according to claim 1, characterized in that: It also includes the following steps, S8, based on steps S5 and S6, after the printer is turned on, if no abnormal parameters are received, record in the printer log; S9, perform daily maintenance on the printer and reset and adjust all parameters of the printer.

3. The intelligent maintenance method for an industrial printer according to claim 1, characterized in that: In step S2, The nozzle group centrally integrates the nozzle temperature sensor and the nozzle pressure sensor, and the sensor data of the integrated nozzle group are uploaded at the same time, and the sensors are reasonably arranged so that they can accurately sense the working status of the nozzle; The paper feeding group integrates the paper feed sensor and paper discharge sensor in the paper feeding channel and the paper jam sensor in the paper feeding channel, and the integrated sensor data are uploaded at the same time. These sensors work together to fully monitor all aspects of the paper feeding process; The ink cartridge group integrates the sensors on the ink cartridge, including the ink cartridge liquid level sensor and the ink cartridge chip status sensor, and the integrated sensor data are uploaded at the same time; by reasonably arranging these sensors, the status of the ink cartridge can be fully understood.

4. The intelligent maintenance method for an industrial printer according to claim 1, characterized in that: In step S3, a nozzle feedback circuit is set up, and a current detection resistor is connected in series to the main circuit of the nozzle circuit to convert the current signal into a voltage signal and then feed it back to the control system.

5. The intelligent maintenance method for an industrial printer according to claim 1, characterized in that: In step S6, various information from the feedback circuit and feedback signal are normalized, and parameters of different magnitudes and types are unified into a comparable scale range. Different parameters such as the nozzle drive voltage, the paper feed motor speed, and the remaining ink amount in the ink cartridge are calculated so that their value range is between 0 and 1 or other set standard ranges. The normalization method used can be Z-score normalization or Min-Max normalization method.

6. The intelligent maintenance method for an industrial printer according to claim 1, characterized in that: In step S7, based on the analysis of the current operating state of the printer and the automatic adjustment measures that have been taken, a mathematical model or an empirical formula is used to predict the result and required time after the adjustment; By analyzing the relationship between the remaining ink level and the printing mode and resolution, it is predicted how long the ink cartridge can support printing tasks after switching the printing mode or lowering the resolution. The predicted results and time are recorded in the printer log for subsequent review and evaluation of the effect of the adjustment.

7. A printing device, comprising at least one processor, at least one memory and computer program instructions stored in the memory, wherein when the computer program instructions are executed by the processor, the method according to any one of claims 1 to 6 is implemented.

8. A storage medium having computer program instructions stored thereon, which implement the method according to any one of claims 1 to 6 when the computer program instructions are executed by a processor.

Images