Printing parameter dynamic adjustment system and method for printing machine
By monitoring the printing process in real time in a digital direct injection printer and dynamically adjusting parameters using the printing deviation analysis model, the problem of preset parameters is solved, the printing quality and production efficiency are improved, and the system's adaptability is enhanced.
Patent Information
- Application Number
- CN202411260371.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-10
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2044-09-10
AI Technical Summary
During the printing process, due to environmental factors and changes in equipment status, the preset printing parameters may not be applicable, resulting in color deviation, inaccurate pattern alignment or uneven printing, affecting product quality and production efficiency.
By setting up a sensor module in a digital direct injection printer, using the collaborative work of the server and the control system, dynamically adjusting the printing parameters based on the pre-trained printing deviation analysis model, and optimizing the printing process in real time.
It significantly improves the accuracy and overall quality of the printing pattern, reduces material waste, improves production efficiency, and enhances the system's adaptability to meet diverse printing needs.
Smart Images

Figure CN118789944B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of digital direct-injection printing machine control, and in particular to a system and method for dynamically adjusting printing parameters of a printing machine. Background Art
[0002] In modern textile manufacturing, digital direct printing technology is widely used in personalized and small batch production due to its high efficiency and flexibility. The printing process involves printing the design pattern directly onto the fabric, which requires extremely high printing accuracy and color consistency to meet the market demand for high-quality products.
[0003] In the related technology, digital direct-injection printers currently generally operate with preset printing parameters. These parameters are set before the printing task begins and remain unchanged throughout the printing process. During the actual printing process, due to changes in environmental factors such as fabric material, humidity, temperature, and changes in the physical state of ink and print head over time, the preset printing parameters may not be applicable, resulting in color deviation, inaccurate pattern alignment, or uneven printing. At this time, even if the user discovers the problem in time, it will take a lot of time to stop the machine and adjust the parameters. If the problem cannot be discovered in time, it will affect the product quality, thereby reducing production efficiency and product quality. Summary of the invention
[0004] The present application provides a system for dynamically adjusting printing parameters of a printing machine. In order to have a basic understanding of some aspects of the disclosed embodiments, a simple summary is given below. This summary is not a general review, nor is it intended to identify key / important components or describe the scope of protection of these embodiments. Its only purpose is to present some concepts in a simple form as a preface to the detailed description that follows.
[0005] In a first aspect, an embodiment of the present application provides a system for dynamically adjusting printing parameters of a printing machine, the system comprising:
[0006] The server and the digital direct-injection printing machine; wherein the digital direct-injection printing machine comprises a control system, a guide belt module, a nozzle module and a sensor module arranged on the digital direct-injection printing machine; wherein,
[0007] The server is connected to the control system in communication;
[0008] The control system is used to control the guide belt module and the nozzle module to print the cloth to be printed according to the initial printing parameters when receiving the printing instruction, and to receive the sensing data fed back by the sensor module in real time, and to send the sensing data and the initial printing parameters to the server;
[0009] A server, for receiving the sensing data and initial printing parameters from the control system, preprocessing the sensing data, obtaining real-time printing information, inputting the real-time printing information into a pre-trained printing deviation analysis model, outputting the deviation analysis result corresponding to the real-time printing information, and dynamically adjusting the initial printing parameters online when the deviation analysis result indicates that there is a problem with the printed pattern, and sending the adjusted latest printing parameters to the control system, wherein the pre-trained printing deviation analysis model is obtained by machine learning based on historical printing data in the control system and the sensor module;
[0010] The control system is also used to receive the latest printing parameters from the server, and based on the latest printing parameters, continue to execute the steps of controlling the guide belt module and the nozzle module to print the fabric to be printed.
[0011] Optionally, generate a pre-trained print deviation analysis model by following these steps, including:
[0012] Collect historical printing data from the control system and sensor module of the digital direct printing machine, the historical printing data including historical printing parameters and historical sensing data;
[0013] For historical printing parameters, a type tag used to characterize whether there is a printing problem is marked, and historical printing parameters carrying the type tag are obtained;
[0014] Generate a training set for model training based on historical printing parameters and historical perception data carrying type labels;
[0015] A neural network is used to construct a printing deviation analysis model;
[0016] Based on the training set, machine learning is performed on the printing deviation analysis model to generate a pre-trained printing deviation analysis model.
[0017] Optionally, for historical printing parameters, a type label is marked to indicate whether there is a printing problem, including:
[0018] Obtain historical patterns from historical perception data;
[0019] Obtaining color deviation parameters, alignment parameters, and printing uniformity parameters of historical patterns;
[0020] According to the color deviation parameter, alignment parameter and printing uniformity parameter, determine whether there is a printing problem in the historical pattern;
[0021] In the case where there is a printing problem in the historical pattern, a first type of label used to indicate the presence of the printing problem is marked on the historical printing parameters; or,
[0022] When there is no printing problem with the historical pattern, a second type of label for indicating that there is no printing problem is marked for the historical printing parameters.
[0023] Optionally, a training set for model training is generated based on historical printing parameters and historical perception data carrying type labels, including:
[0024] Constructing the covariance matrix of historical perception data to measure the linear relationship between different variables in the historical perception data;
[0025] Calculate the eigenvalues of the covariance matrix;
[0026] According to the size of the eigenvalue, the eigenvectors corresponding to the eigenvalues whose number is greater than the preset threshold are selected as the principal components;
[0027] Mapping different variables in the historical perception data to the principal components, generating scores of different variables in the historical perception data on the principal components;
[0028] Based on the scores of different variables in the historical perception data on the principal components, the redundancy of different variables in the historical perception data is calculated;
[0029] Variables with redundancy greater than a preset threshold are considered key features;
[0030] Bind the key features to the historical printing parameters carrying the type labels to obtain the training set for model training.
[0031] Optionally, different variables in the historical sensing data include printing speed, ink flow, nozzle status, and operating environment parameters;
[0032] Based on the scores of different variables in the historical perception data on the principal components, the redundancy of different variables in the historical perception data is calculated, including:
[0033] The total variance is generated based on the scores of printing speed, ink flow, nozzle status, and working environment parameters on the principal components;
[0034] Based on the total variance, calculate the variance contribution rate of printing speed, ink flow, nozzle status and working environment parameters;
[0035] The redundancy of the printing speed, ink flow, nozzle status and operating environment parameters is calculated according to the variance contribution rates of the printing speed, ink flow, nozzle status and operating environment parameters.
[0036] Optionally, the variance contribution rate of each variable is calculated as:
[0037]
[0038] in, It is The variance contribution of the variables, It is variables, namely, printing speed, ink flow, nozzle status and operating environment parameters. is the preset number of principal components, For the The characteristic values of the spare sensor data, is the total variance;
[0039] The redundancy calculation formula for each variable is:
[0040]
[0041] in, For the The redundancy of variables, is the preset number of principal components, For the The variance contribution of the variables, For the The square of the scores of the variables on the principal components.
[0042] Optionally, the training set includes binding data of key features and historical printing parameters carrying type labels, the key features are obtained based on the analysis of historical perception data, and the type labels are used to characterize whether there are printing problems; the printing deviation analysis model includes a feature extraction module and a parameter constraint module, and the feature extraction module includes multiple convolutional layers and a fully connected layer;
[0043] According to the training set, the printing deviation analysis model is machine-learned to generate a pre-trained printing deviation analysis model, including:
[0044] Input the key features and the binding data of the historical printing parameters carrying the type label into multiple convolutional layers and fully connected layers in sequence, and output the sample feature vector of each training data in the training set;
[0045] Input the sample feature vector of each training data into the parameter constraint module and output the constraint value of the model;
[0046] determining the model constraint value as the model loss value;
[0047] When the model loss value reaches the minimum, a pre-trained printing deviation analysis model is generated; alternatively, when the model loss value does not reach the minimum, the constraint value is back-propagated to update the model parameters of the printing deviation analysis model, and the step of inputting the sample feature vector of each training data into the parameter constraint module is continued.
[0048] Optionally, the parameter constraint module includes a pre-established loss constraint function; wherein the loss constraint function is:
[0049]
[0050] in, is the constraint value, is the weight used to balance quality and efficiency, is the print quality loss, which quantifies the difference between the printed pattern and the expected pattern. is the model parameter The regularization term of ; where
[0051] The loss function of print quality loss is:
[0052]
[0053] in, is the number of samples in the training set, It is The feature vector of the key features of the samples, It is The feature vector of the historical printing parameters carrying the type label of the samples, The tag value for the type tag.
[0054] Optionally, pre-process the sensing data to obtain real-time printing information, including:
[0055] Eliminate outliers or noise in the perception data through a box plot to obtain first perception data;
[0056] Using linear interpolation to supplement missing values of the first perception data to obtain second perception data;
[0057] The second perception data is downsampled or aggregated to reduce the amount of calculation and extract key information to obtain real-time printing information.
[0058] In a second aspect, a method for dynamically adjusting printing parameters of a printing machine is provided, the method comprising:
[0059] When receiving the printing instruction, the control system controls the guide belt module and the nozzle module to print the fabric to be printed according to the initial printing parameters, receives the sensing data fed back by the sensor module in real time, and sends the sensing data and the initial printing parameters to the server;
[0060] The server receives the sensing data and initial printing parameters from the control system, pre-processes the sensing data, obtains real-time printing information, inputs the real-time printing information into a pre-trained printing deviation analysis model, outputs the deviation analysis results corresponding to the real-time printing information, and dynamically adjusts the initial printing parameters online when the deviation analysis results indicate that there are problems with the printed pattern, and sends the adjusted latest printing parameters to the control system. The pre-trained printing deviation analysis model is obtained by machine learning based on historical printing data in the control system and the sensor module.
[0061] The control system receives the latest printing parameters from the server, and based on the latest printing parameters, continues to execute the steps of controlling the guide belt module and the nozzle module to print the fabric to be printed.
[0062] In an embodiment of the present application, real-time monitoring and intelligent optimization of the digital direct-injection printing machine during the printing process are achieved through the collaborative working mechanism of the control system and the server. By receiving the perception data fed back by the sensor module in real time, the control system can promptly capture any deviations in the printing process and send these data together with the initial printing parameters to the server. The server uses these data to conduct in-depth analysis through a pre-trained printing deviation analysis model to quickly identify and predict printing quality problems. Once a problem is found, the server can dynamically adjust the printing parameters online and feed back the latest parameters to the control system in real time to guide it to continue to execute the printing task. This method of online cyclic dynamic adjustment of printing parameters not only significantly improves the accuracy and overall quality of the printed pattern, but also effectively reduces material waste, improves production efficiency, and enhances the system's adaptability to meet diverse printing needs.
[0063] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0064] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.
[0065] Figure 1 1 is a schematic diagram of the system structure of a printing parameter dynamic adjustment system for a printing machine provided in an embodiment of the present application;
[0066] Figure 2 This is a data representation of real-time printing information provided by an embodiment of the present application;
[0067] Figure 3 This is a data representation scheme after dynamic parameter adjustment provided in an embodiment of the present application;
[0068] Figure 4It is a method flow diagram of a method for dynamically adjusting printing parameters of a printing machine provided in an embodiment of the present application;
[0069] Figure 5 It is a schematic diagram of a method flow of another method for dynamically adjusting printing parameters of a printing machine provided in an embodiment of the present application;
[0070] Figure 6 It is a structural schematic diagram of a printing parameter dynamic adjustment device for a printing machine provided in an embodiment of the present application;
[0071] Figure 7 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0072] The following description and the accompanying drawings sufficiently illustrate specific embodiments of the present application to enable those skilled in the art to practice them.
[0073] It should be clear that the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.
[0074] When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Instead, they are only examples of systems and methods consistent with some aspects of the present application as detailed in the attached claims.
[0075] In the description of the present application, it should be understood that the terms "first", "second", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances. In addition, in the description of the present application, unless otherwise specified, "multiple" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects associated before and after are in an "or" relationship.
[0076] In an embodiment of the present application, real-time monitoring and intelligent optimization of the digital direct-injection printing machine during the printing process are achieved through the collaborative working mechanism of the control system and the server. By receiving the perception data fed back by the sensor module in real time, the control system can promptly capture any deviations in the printing process and send these data together with the initial printing parameters to the server. The server uses these data to conduct in-depth analysis through a pre-trained printing deviation analysis model to quickly identify and predict printing quality problems. Once a problem is found, the server can dynamically adjust the printing parameters online and feed back the latest parameters to the control system in real time to guide it to continue to perform the printing task. This method of online cyclic dynamic adjustment of printing parameters not only significantly improves the accuracy and overall quality of the printed pattern, but also effectively reduces material waste and improves production efficiency. At the same time, it enhances the system's adaptability to meet diverse printing needs. The following exemplary embodiments are used for detailed description.
[0077] See also Figure 1 , Figure 1 This is a system structure diagram of a system for dynamically adjusting printing parameters of a printing machine provided in an embodiment of the present application, the system comprising: a server and a digital direct-injection printing machine; wherein the digital direct-injection printing machine comprises a control system, a guide belt module, a nozzle module and a sensor module arranged on the digital direct-injection printing machine; wherein the server is communicatively connected with the control system.
[0078] In an embodiment of the present application, a control system is used to control the guide tape module and the nozzle module to print on the fabric to be printed according to the initial printing parameters when a printing instruction is received, and to receive the perception data fed back by the sensor module in real time, and to send the perception data and the initial printing parameters to the server; the server is used to receive the perception data and the initial printing parameters from the control system, pre-process the perception data, obtain real-time printing information, input the real-time printing information into a pre-trained printing deviation analysis model, output the deviation analysis result corresponding to the real-time printing information, and dynamically adjust the initial printing parameters online when the deviation analysis result indicates that there is a problem with the printed pattern, and send the adjusted latest printing parameters to the control system, the pre-trained printing deviation analysis model is obtained by machine learning based on the historical printing data in the control system and the sensor module; the control system is also used to receive the latest printing parameters from the server, and based on the latest printing parameters, continue to execute the step of controlling the guide tape module and the nozzle module to print on the fabric to be printed.
[0079] Among them, the server refers to a remote computing system that provides various services, such as data processing, storage and analysis. In this application, the server is responsible for receiving data from the control system and performing further processing and analysis. A digital direct-injection printer is a machine that can print patterns directly on various materials, and is usually used in the production of clothing, textiles, etc. The control system is a core component in a digital direct-injection printer, which is responsible for receiving printing instructions and controlling the various modules of the machine to operate according to predetermined parameters. The guide belt module is responsible for transporting the fabric to the printing area in the printer to ensure smooth movement and correct positioning of the fabric. The nozzle module is the part of the printer that performs the actual printing operation, and it forms a pattern on the fabric by spraying ink. The sensor module is a collection of sensors installed on the printer, which is used to collect various data during the printing process, such as printing speed, ink flow, etc. The printing deviation analysis model is a model trained using machine learning technology, which can analyze data during the printing process, identify printing deviations, and predict possible problems. The initial printing parameters are parameters set before the start of the printing task, such as printing speed, ink volume, etc., which determine the basic behavior of printing. Perception data is the real-time data collected by the sensor module about the printing process, which reflects the actual situation during the printing process. Real-time printing information is the perception data received and pre-processed by the server, which reflects the status of the current printing task.
[0080] For example, a clothing manufacturer wants to use this digital direct printing machine to print a batch of T-shirts. First, the manufacturer will input the initial printing parameters, such as printing speed and ink volume, through the control system. When the control system receives the printing instruction, the guide belt module starts to transport the T-shirt to the printing position, and the nozzle module starts to print the pattern according to the instruction of the control system. The sensor module monitors the printing process in real time, collects data such as ink flow and printing speed, and feeds this data back to the control system. The control system sends these sensed data and initial printing parameters to the server. After receiving the data, the server performs preprocessing, generates real-time printing information, and inputs it into the printing deviation analysis model. If the model analysis results show that there is a deviation in the pattern, the server will dynamically adjust the printing parameters, such as increasing the ink injection amount or adjusting the printing speed, and send the adjusted parameters back to the control system.
[0081] Specifically, the process of preprocessing the perception data to obtain real-time printing information includes: eliminating outliers or noise in the perception data through a box plot to obtain first perception data; using linear interpolation to supplement missing values in the first perception data to obtain second perception data; downsampling or aggregating the second perception data to reduce the amount of calculation and extract key information to obtain real-time printing information.
[0082] For example, removing noise and outliers from the data to ensure data accuracy. This includes removing data points that are outside the normal range or filling in missing data. Converting raw data into a format suitable for analysis.
[0083] Furthermore, it is also possible to: convert analog signals into digital signals, or unify data from different sensors to the same timestamp. Adjust the scale of the data to make it within a unified range. Extract useful features from the raw data, which can represent key information in the printing process, such as the uniformity of ink distribution, changes in printing speed, etc. After processing, real-time printing information can be obtained. Real-time printing information, for example Figure 2 shown.
[0084] Specifically, the specific steps of online dynamic adjustment of initial printing parameters include: determining the parameters to be adjusted based on the deviation analysis results, and adjusting the parameter values of the parameters to be adjusted within a preset range. For example, if the color is deviated, adjust the ink ratio or the nozzle injection pressure. If the alignment is inaccurate, adjust the guide belt speed or the nozzle movement speed. The printing parameters after dynamic adjustment are, for example, Figure 3 shown.
[0085] In an embodiment of the present application, real-time monitoring and intelligent optimization of the digital direct-injection printing machine during the printing process are achieved through the collaborative working mechanism of the control system and the server. By receiving the perception data fed back by the sensor module in real time, the control system can promptly capture any deviations in the printing process and send these data together with the initial printing parameters to the server. The server uses these data to conduct in-depth analysis through a pre-trained printing deviation analysis model to quickly identify and predict printing quality problems. Once a problem is found, the server can dynamically adjust the printing parameters online and feed back the latest parameters to the control system in real time to guide it to continue to execute the printing task. This method of online cyclic dynamic adjustment of printing parameters not only significantly improves the accuracy and overall quality of the printed pattern, but also effectively reduces material waste, improves production efficiency, and enhances the system's adaptability to meet diverse printing needs.
[0086] In some embodiments of the present application, a pre-trained print deviation analysis model is generated according to the following steps, including: collecting historical print data from the control system and sensor module of the digital direct-injection printer, the historical print data including historical printing parameters and historical perception data; marking the historical printing parameters with type labels used to characterize whether there are printing problems, and obtaining the historical printing parameters carrying the type labels; generating a training set for model training based on the historical printing parameters and historical perception data carrying the type labels; constructing a print deviation analysis model using a neural network; and performing machine learning on the print deviation analysis model based on the training set to generate a pre-trained print deviation analysis model.
[0087] Among them, historical printing data is data collected during past printing processes, including printing parameters and perception data. Historical printing parameters are printing settings used in historical printing processes, such as printing speed, ink flow, etc. Historical perception data is data collected by sensors during historical printing processes, such as actual measurements of printing speed and ink flow. Type labels are labels used to mark whether there are printing problems with historical printing parameters, such as "normal" or "deviation". The training set is a data set used for model training, including historical printing parameters with type labels and corresponding perception data. A neural network is a machine learning model that mimics the way the human brain processes information and is used to build a printing deviation analysis model.
[0088] In the embodiment of the present application, by collecting historical printing data in the control system and sensor module of the digital direct-injection printer, including printing parameters and perception data, data with printing problems can be marked and a training set with type labels can be created. Using these data, a printing deviation analysis model is constructed and trained using neural network technology, so that it can identify and predict possible deviations in the printing process through machine learning. This pre-trained model can significantly improve printing quality, reduce material waste, improve production efficiency, and enable the printer to adapt to various printing tasks, realizing an intelligent and automated production process.
[0089] In some embodiments of the present application, for historical printing parameters, a specific process of marking a type label for characterizing whether there is a printing problem includes: obtaining a historical pattern from historical perception data; obtaining color deviation parameters, alignment parameters, and printing uniformity parameters of the historical pattern; determining whether there is a printing problem with the historical pattern based on the color deviation parameters, alignment parameters, and printing uniformity parameters; if there is a printing problem with the historical pattern, marking the historical printing parameters with a first type label for characterizing the existence of the printing problem; or, if there is no printing problem with the historical pattern, marking the historical printing parameters with a second type label for characterizing the absence of the printing problem.
[0090] The historical pattern refers to a specific pattern printed in a previous printing task. The color deviation parameter is a parameter used to measure the difference between the color in the printed pattern and the expected color. The printing uniformity parameter is a parameter used to measure whether the ink distribution in the printed pattern is uniform. Printing problems are problems that may occur during the printing process, such as color deviation, alignment errors, or uneven printing. The first type of label is used to mark the label of the pattern with printing problems. The second type of label is used to mark the label of the pattern without printing problems.
[0091] In an embodiment of the present application, by analyzing historical perception data, historical patterns can be extracted, and the color deviation, alignment accuracy, and printing uniformity of these patterns can be evaluated. Based on these evaluation parameters, it can be determined whether each historical pattern has printing problems. For those patterns with problems, the first type of label will be marked on the corresponding historical printing parameters so that the problem pattern can be identified when training the printing deviation analysis model; for those patterns without problems, the second type of label will be marked to indicate that these parameters meet the quality standards. This method helps to build an accurate printing deviation analysis model that can learn to identify the conditions that cause printing problems and predict problems that may occur in future printing tasks, thereby achieving early prevention and adjustment, improving printing quality, reducing scrap rates, and optimizing production processes.
[0092] In some embodiments of the present application, the specific process of generating a training set for model training based on historical printing parameters and historical perception data carrying type labels includes: constructing a covariance matrix of historical perception data to measure the linear relationship between different variables in the historical perception data; calculating the eigenvalues of the covariance matrix; according to the size of the eigenvalues, selecting eigenvectors corresponding to eigenvalues with a preset number greater than a preset threshold as principal components; mapping different variables in the historical perception data to principal components, and generating scores of different variables in the historical perception data on principal components; calculating the redundancy of different variables in the historical perception data based on the scores of different variables in the historical perception data on principal components; taking variables with redundancy greater than a preset threshold as key features; binding the key features to the historical printing parameters carrying type labels to obtain a training set for model training.
[0093] Among them, the covariance matrix is a matrix used to measure the linear relationship between a set of variables. In historical perception data, the covariance matrix can show the correlation between different variables. Eigenvalue is a concept in linear algebra, which is used to describe certain characteristics of a matrix. In the covariance matrix, the size of the eigenvalue can reflect the variability of the data. Eigenvectors are vectors corresponding to eigenvalues, which are invariant vectors under matrix transformation. In principal component analysis, eigenvectors are used to determine the main direction of change in the data. The principal component is the most important direction of change in the data extracted by principal component analysis (PCA) and is used for data dimensionality reduction. Score In principal component analysis, the values obtained by mapping the original variables to the principal components represent the projection of the variables in the direction of the principal components. Redundancy is a measure of how much information a variable contains about other variables. In data analysis, variables with high redundancy may not be key features. Key features are variables that have the greatest impact on the results in data analysis, and they are the focus of model training. Historical printing parameters carrying type labels are historical printing parameters with problem type labels, which are used to train models to identify printing problems.
[0094] In an embodiment of the present application, by constructing a covariance matrix of historical perception data, we can quantify the linear relationship between different variables and calculate eigenvalues to measure the variability of the data. Eigenvectors with eigenvalues greater than a preset threshold are selected as principal components, and then the original variables are mapped to these principal components to obtain scores. Based on these scores, we can calculate the redundancy of each variable and identify key features with redundancy below the preset threshold. Combining these key features with historical printing parameters with type labels, we obtain a high-quality training set that can provide accurate and representative data for the printing deviation analysis model. This method not only improves the training efficiency of the model, but also enhances the model's ability to identify and predict printing problems, thereby achieving higher quality control and optimization in the production process.
[0095] Among them, different variables in the historical perception data include printing speed, ink flow, nozzle status and working environment parameters.
[0096] In some embodiments of the present application, based on the scores of different variables in the historical perception data on the principal components, the specific process of calculating the redundancy of different variables in the historical perception data includes: generating a total variance according to the scores of the printing speed, ink flow, nozzle status and operating environment parameters on the principal components; calculating the variance contribution rate of the printing speed, ink flow, nozzle status and operating environment parameters according to the total variance; calculating the redundancy of the printing speed, ink flow, nozzle status and operating environment parameters according to the variance contribution rate of the printing speed, ink flow, nozzle status and operating environment parameters.
[0097] Specifically, the variance contribution rate calculation formula of each variable is:
[0098]
[0099] in, It is The variance contribution of the variables, It is variables, namely, printing speed, ink flow, nozzle status and operating environment parameters. is the preset number of principal components, For the The characteristic values of the spare sensor data, is the total variance;
[0100] The redundancy calculation formula for each variable is:
[0101]
[0102] in, For the The redundancy of variables, is the preset number of principal components, For the The variance contribution of the variables, For the The square of the scores of the variables on the principal components.
[0103] Among them, the training set includes binding data of key features and historical printing parameters carrying type labels. The key features are obtained based on the analysis of historical perception data, and the type labels are used to characterize whether there are printing problems; the printing deviation analysis model includes a feature extraction module and a parameter constraint module. The feature extraction module includes multiple convolutional layers and fully connected layers.
[0104] In some embodiments of the present application, a print deviation analysis model is subjected to machine learning based on a training set, and the specific process of generating a pre-trained print deviation analysis model includes: inputting the binding data of key features and historical printing parameters carrying type labels into multiple convolutional layers and a fully connected layer in sequence, and outputting a sample feature vector of each training data in the training set; inputting the sample feature vector of each training data into a parameter constraint module, and outputting the constraint value of the model; determining the model constraint value as the model loss value; generating a pre-trained print deviation analysis model when the model loss value reaches a minimum; or, when the model loss value does not reach a minimum, backpropagating the constraint value to update the model parameters of the print deviation analysis model, and continuing to execute the step of inputting the sample feature vector of each training data into the parameter constraint module.
[0105] Specifically, the parameter constraint module includes a pre-established loss constraint function; wherein the loss constraint function is:
[0106]
[0107] in, is the constraint value, is the weight used to balance quality and efficiency, is the print quality loss, which quantifies the difference between the printed pattern and the expected pattern. is the model parameter The regularization term of ; where
[0108] The loss function of print quality loss is:
[0109]
[0110] in, is the number of samples in the training set, It is The feature vector of the key features of the samples, It is The feature vector of the historical printing parameters carrying the type label of the samples, The tag value for the type tag.
[0111] Among them, the convolution layer is a layer used to extract the spatial features of input data in a neural network, especially in a convolutional neural network (CNN). The fully connected layer is a layer in a neural network, in which each input neuron is connected to each output neuron, and is usually used to integrate information from the previous layer. The sample feature vector is a vector obtained after the input data is processed by the convolution layer and the fully connected layer, which represents the key features of the data. The parameter constraint module is a module used to process the sample feature vector and output the model constraint values, which are used to guide the training of the model. Backpropagation is a method used to calculate the gradient of the loss function in a neural network, in which the weights and biases of the network can be updated.
[0112] In an embodiment of the present application, a rich training set is constructed by combining key features with historical printing parameters with type labels. These data are input into the convolutional layer and the fully connected layer to extract the feature vector of the sample. These feature vectors are then fed into the parameter constraint module, which outputs the constraint values of the model, which are used as inputs to the model loss function. During the training process, the model parameters are continuously adjusted to minimize the model loss value through the back propagation algorithm. When the model loss value reaches the minimum, an optimized printing deviation analysis model is obtained. If the loss value has not been minimized, back propagation will continue to be performed to update the model parameters until the model reaches the expected accuracy. This process not only improves the prediction accuracy of the model, but also reduces human intervention by automatically adjusting parameters, improves the automation and intelligence level of the printing process, and thus achieves higher quality control and optimization in production.
[0113] In an embodiment of the present application, real-time monitoring and intelligent optimization of the digital direct-injection printing machine during the printing process are achieved through the collaborative working mechanism of the control system and the server. By receiving the perception data fed back by the sensor module in real time, the control system can promptly capture any deviations in the printing process and send these data together with the initial printing parameters to the server. The server uses these data to conduct in-depth analysis through a pre-trained printing deviation analysis model to quickly identify and predict printing quality problems. Once a problem is found, the server can dynamically adjust the printing parameters online and feed back the latest parameters to the control system in real time to guide it to continue to execute the printing task. This method of online cyclic dynamic adjustment of printing parameters not only significantly improves the accuracy and overall quality of the printed pattern, but also effectively reduces material waste, improves production efficiency, and enhances the system's adaptability to meet diverse printing needs.
[0114] See also Figure 4, which is a flow chart of a method for dynamically adjusting printing parameters of a printing machine provided in an embodiment of the present application. Figure 4 As shown, the detection method of the embodiment of the present application may include the following steps:
[0115] S101, when receiving a printing instruction, the control system controls the guide belt module and the nozzle module to print the fabric to be printed according to the initial printing parameters, receives the sensing data fed back by the sensor module in real time, and sends the sensing data and the initial printing parameters to the server;
[0116] S102, the server receives the sensing data and initial printing parameters from the control system, pre-processes the sensing data, obtains real-time printing information, inputs the real-time printing information into a pre-trained printing deviation analysis model, outputs the deviation analysis result corresponding to the real-time printing information, and dynamically adjusts the initial printing parameters online when the deviation analysis result indicates that there is a problem with the printed pattern, and sends the adjusted latest printing parameters to the control system, wherein the pre-trained printing deviation analysis model is obtained by machine learning based on historical printing data in the control system and the sensor module;
[0117] S103, the control system receives the latest printing parameters from the server, and based on the latest printing parameters, continues to execute the step of controlling the guide belt module and the nozzle module to print on the cloth to be printed.
[0118] In an embodiment of the present application, real-time monitoring and intelligent optimization of the digital direct-injection printing machine during the printing process are achieved through the collaborative working mechanism of the control system and the server. By receiving the perception data fed back by the sensor module in real time, the control system can promptly capture any deviations in the printing process and send these data together with the initial printing parameters to the server. The server uses these data to conduct in-depth analysis through a pre-trained printing deviation analysis model to quickly identify and predict printing quality problems. Once a problem is found, the server can dynamically adjust the printing parameters online and feed back the latest parameters to the control system in real time to guide it to continue to execute the printing task. This method of online cyclic dynamic adjustment of printing parameters not only significantly improves the accuracy and overall quality of the printed pattern, but also effectively reduces material waste, improves production efficiency, and enhances the system's adaptability to meet diverse printing needs.
[0119] See also Figure 5 , a flow chart of a method for dynamically adjusting printing parameters of a printing machine is provided for an embodiment of the present application, and is applied to a server. Figure 5 As shown, the detection method of the embodiment of the present application may include the following steps:
[0120] S201, receiving sensing data and initial printing parameters from a control system;
[0121] S202, pre-processing the sensing data to obtain real-time printing information;
[0122] S203, inputting the real-time printing information into a pre-trained printing deviation analysis model, and outputting a deviation analysis result corresponding to the real-time printing information;
[0123] S204, when the deviation analysis result indicates that there is a problem with the printed pattern, dynamically adjusting the initial printing parameters online;
[0124] S205, sending the latest adjusted printing parameters to the control system, and the pre-trained printing deviation analysis model is obtained by machine learning based on historical printing data in the control system and the sensor module.
[0125] In an embodiment of the present application, real-time monitoring and intelligent optimization of the digital direct-injection printing machine during the printing process are achieved through the collaborative working mechanism of the control system and the server. By receiving the perception data fed back by the sensor module in real time, the control system can promptly capture any deviations in the printing process and send these data together with the initial printing parameters to the server. The server uses these data to conduct in-depth analysis through a pre-trained printing deviation analysis model to quickly identify and predict printing quality problems. Once a problem is found, the server can dynamically adjust the printing parameters online and feed back the latest parameters to the control system in real time to guide it to continue to execute the printing task. This method of online cyclic dynamic adjustment of printing parameters not only significantly improves the accuracy and overall quality of the printed pattern, but also effectively reduces material waste, improves production efficiency, and enhances the system's adaptability to meet diverse printing needs.
[0126] The following is an embodiment of the device of the present application, which can be used to execute the embodiment of the method of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the method of the present application.
[0127] See also Figure 6 , which shows a schematic diagram of the structure of a printing parameter dynamic adjustment device for a printing machine provided by an exemplary embodiment of the present application. The printing parameter dynamic adjustment device for a printing machine can be implemented as all or part of an electronic device through software, hardware or a combination of both. The device 1 includes a receiving module 10, a preprocessing module 20, an output module 30, an adjustment module 40, and a sending module 50.
[0128] A receiving module 10, for receiving sensing data and initial printing parameters from a control system;
[0129] A preprocessing module 20, used for preprocessing the sensing data to obtain real-time printing information;
[0130] An output module 30, used to input the real-time printing information into a pre-trained printing deviation analysis model, and output a deviation analysis result corresponding to the real-time printing information;
[0131] An adjustment module 40, for dynamically adjusting initial printing parameters online when the deviation analysis result indicates that there is a problem with the printed pattern;
[0132] The sending module 50 is used to send the latest adjusted printing parameters to the control system. The pre-trained printing deviation analysis model is obtained by machine learning based on the historical printing data in the control system and the sensor module.
[0133] It should be noted that the apparatus for dynamically adjusting printing parameters for a printing machine provided in the above embodiment only uses the division of the above functional modules as an example when executing the method for dynamically adjusting printing parameters for a printing machine. In actual applications, the above functional distribution can be completed by different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the apparatus for dynamically adjusting printing parameters for a printing machine provided in the above embodiment and the method for dynamically adjusting printing parameters for a printing machine belong to the same concept, and the implementation process thereof is detailed in the method embodiment, which will not be repeated here.
[0134] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.
[0135] In an embodiment of the present application, real-time monitoring and intelligent optimization of the digital direct-injection printing machine during the printing process are achieved through the collaborative working mechanism of the control system and the server. By receiving the perception data fed back by the sensor module in real time, the control system can promptly capture any deviations in the printing process and send these data together with the initial printing parameters to the server. The server uses these data to conduct in-depth analysis through a pre-trained printing deviation analysis model to quickly identify and predict printing quality problems. Once a problem is found, the server can dynamically adjust the printing parameters online and feed back the latest parameters to the control system in real time to guide it to continue to execute the printing task. This method of online cyclic dynamic adjustment of printing parameters not only significantly improves the accuracy and overall quality of the printed pattern, but also effectively reduces material waste, improves production efficiency, and enhances the system's adaptability to meet diverse printing needs.
[0136] The present application also provides a computer-readable medium having program instructions stored thereon, and when the program instructions are executed by a processor, the method for dynamically adjusting printing parameters for a printing machine provided by the above-mentioned various method embodiments is implemented.
[0137] The present application also provides a computer program product including instructions, which, when executed on a computer, enables the computer to execute the method for dynamically adjusting printing parameters of a printing machine according to the above-mentioned various method embodiments.
[0138] See also Figure 7 , is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 7 As shown, the electronic device 1000 may include: at least one processor 1001 , at least one network interface 1004 , a user interface 1003 , a memory 1005 , and at least one communication bus 1002 .
[0139] The communication bus 1002 is used to realize the connection and communication between these components.
[0140] The user interface 1003 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 1003 may also include a standard wired interface and a wireless interface.
[0141] The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).
[0142] Among them, the processor 1001 may include one or more processing cores. The processor 1001 uses various interfaces and lines to connect various parts of the entire electronic device 1000, and executes various functions and processes data of the electronic device 1000 by running or executing instructions, programs, code sets or instruction sets stored in the memory 1005, and calling data stored in the memory 1005. Optionally, the processor 1001 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 1001 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 1001, and it can be implemented separately through a chip.
[0143] Among them, the memory 1005 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read-Only Memory). Optionally, the memory 1005 includes a non-transitory computer-readable storage medium. The memory 1005 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 1005 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 1005 may also be optionally at least one storage system located away from the aforementioned processor 1001. As Figure 7 As shown, the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module, and an application program for dynamically adjusting printing parameters of a printing machine.
[0144] exist Figure 7 In the electronic device 1000 shown, the user interface 1003 is mainly used to provide an input interface for the user and obtain data input by the user; and the processor 1001 can be used to call the printing parameter dynamic adjustment application for the printing machine stored in the memory 1005, and specifically perform the following operations:
[0145] Receive sensing data and initial printing parameters from the control system;
[0146] Pre-process the sensing data to obtain real-time printing information;
[0147] Input the real-time printing information into a pre-trained printing deviation analysis model, and output the deviation analysis result corresponding to the real-time printing information;
[0148] In the case where the deviation analysis results indicate that there are problems with the printed pattern, the initial printing parameters are dynamically adjusted online;
[0149] The latest adjusted printing parameters are sent to the control system, and the pre-trained printing deviation analysis model is obtained by machine learning based on the historical printing data in the control system and sensor module.
[0150] In an embodiment of the present application, real-time monitoring and intelligent optimization of the digital direct-injection printing machine during the printing process are achieved through the collaborative working mechanism of the control system and the server. By receiving the perception data fed back by the sensor module in real time, the control system can promptly capture any deviations in the printing process and send these data together with the initial printing parameters to the server. The server uses these data to conduct in-depth analysis through a pre-trained printing deviation analysis model to quickly identify and predict printing quality problems. Once a problem is found, the server can dynamically adjust the printing parameters online and feed back the latest parameters to the control system in real time to guide it to continue to execute the printing task. This method of online cyclic dynamic adjustment of printing parameters not only significantly improves the accuracy and overall quality of the printed pattern, but also effectively reduces material waste, improves production efficiency, and enhances the system's adaptability to meet diverse printing needs.
[0151] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program for dynamic adjustment of printing parameters of a printing machine can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the storage medium of the program for dynamic adjustment of printing parameters of a printing machine can be a disk, an optical disk, a read-only storage memory, or a random access memory, etc.
[0152] The above disclosure is only the preferred embodiment of the present application, which certainly cannot be used to limit the scope of rights of the present application. Therefore, equivalent changes made according to the claims of the present application are still within the scope covered by the present application.
[0153] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program for dynamic adjustment of the printing parameters of the printing machine can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the storage medium for dynamic adjustment of the printing parameters of the printing machine can be a disk, an optical disk, a read-only storage memory, or a random access memory, etc.
[0154] The above disclosure is only the preferred embodiment of the present application, which certainly cannot be used to limit the scope of rights of the present application. Therefore, equivalent changes made according to the claims of the present application are still within the scope covered by the present application.
Claims
1. A printing parameter dynamic adjustment system for a printing machine, characterized in that: The system comprises: A service end and a digital direct-injection printing machine; wherein the digital direct-injection printing machine comprises a control system, a guide belt module, a nozzle module and a sensor module arranged on the digital direct-injection printing machine; wherein, The server is connected to the control system in communication; wherein, The control system is used to control the guide belt module and the nozzle module to print on the cloth to be printed according to the initial printing parameters when receiving the printing instruction, and to receive the sensing data fed back by the sensor module in real time, and to send the sensing data and the initial printing parameters to the server; The server is used to receive the sensing data and initial printing parameters from the control system, pre-process the sensing data, obtain real-time printing information, input the real-time printing information into a pre-trained printing deviation analysis model, output the deviation analysis result corresponding to the real-time printing information, and dynamically adjust the initial printing parameters online when the deviation analysis result indicates that there is a problem with the printed pattern, and send the adjusted latest printing parameters to the control system, wherein the pre-trained printing deviation analysis model is obtained by machine learning based on the historical printing data in the control system and the sensor module; The control system is also used to receive the latest printing parameters from the server, and based on the latest printing parameters, continue to execute the step of controlling the guide belt module and the nozzle module to print the cloth to be printed.
2. The system according to claim 1, characterized in that Follow these steps to generate a pre-trained stamp deviation analysis model, including: Collecting historical printing data from the control system of the digital direct-injection printing machine and the sensor module, wherein the historical printing data includes historical printing parameters and historical sensing data; For the historical printing parameters, a type tag used to characterize whether there is a printing problem is marked, and the historical printing parameters carrying the type tag are obtained; Generate a training set for model training based on the historical printing parameters carrying the type label and the historical perception data; A neural network is used to construct a printing deviation analysis model; The printing deviation analysis model is subjected to machine learning according to the training set to generate a pre-trained printing deviation analysis model.
3. The system according to claim 2, characterized in that The type label for marking whether there is a printing problem for the historical printing parameters includes: Acquiring a historical pattern from the historical perception data; Obtaining color deviation parameters, alignment parameters, and printing uniformity parameters of the historical pattern; Determining whether there is a printing problem with the historical pattern according to the color deviation parameter, the alignment parameter, and the printing uniformity parameter; In the case where the historical pattern has a printing problem, marking the historical printing parameters with a first type of label used to indicate the existence of the printing problem; or, In the case that there is no printing problem with the historical pattern, a second type of label for indicating that there is no printing problem is marked for the historical printing parameters.
4. The system according to claim 2, characterized in that The generating a training set for model training based on the historical printing parameters of the carrying type label and the historical perception data includes: Constructing a covariance matrix of the historical perception data to measure the linear relationship between different variables in the historical perception data; Calculating eigenvalues of the covariance matrix; According to the magnitude of the eigenvalues, eigenvectors corresponding to eigenvalues whose number is greater than a preset threshold are selected as principal components; Mapping different variables in the historical perception data to the principal components to generate scores of different variables in the historical perception data on the principal components; Calculating the redundancy of different variables in the historical perception data based on the scores of different variables in the historical perception data on the principal components; The variable whose redundancy is greater than a preset threshold is taken as a key feature; The key features are bound to the historical printing parameters of the carrying type label to obtain a training set for model training.
5. The system according to claim 4, characterized in that The different variables in the historical sensing data include printing speed, ink flow, nozzle status and operating environment parameters; The calculating the redundancy of different variables in the historical perception data based on the scores of different variables in the historical perception data on the principal components includes: generating a total variance according to the scores of the printing speed, ink flow, nozzle status and working environment parameters on the principal components; Calculating variance contribution rates of the printing speed, ink flow, nozzle status, and operating environment parameters according to the total variance; The redundancy of the printing speed, ink flow, nozzle state and operating environment parameters is calculated according to the variance contribution rates of the printing speed, ink flow, nozzle state and operating environment parameters.
6. The system according to claim 5, characterized in that The variance contribution rate of each variable is calculated as follows: in, It is The variance contribution of the variables, It is variables, namely, printing speed, ink flow, nozzle status and operating environment parameters. is the preset number of principal components, For the The characteristic values of the spare sensor data, is the total variance; where, The redundancy calculation formula for each variable is: in, For the The redundancy of variables, is the preset number of principal components, For the The variance contribution of the variables, For the The square of the score of the variable on the principal component.
7. The system according to claim 2, characterized in that The training set includes binding data of key features and historical printing parameters carrying type tags, wherein the key features are obtained based on the analysis of the historical perception data, and the type tags are used to characterize whether there are printing problems; the printing deviation analysis model includes a feature extraction module and a parameter constraint module, and the feature extraction module includes multiple convolutional layers and a fully connected layer; The step of performing machine learning on the printing deviation analysis model according to the training set to generate a pre-trained printing deviation analysis model includes: Inputting the key features and the binding data of the historical printing parameters carrying the type label into the multiple convolutional layers and the fully connected layer in sequence, and outputting the sample feature vector of each training data in the training set; Inputting the sample feature vector of each training data into the parameter constraint module, and outputting the constraint value of the model; Determining the model constraint value as a model loss value; When the model loss value reaches the minimum, a pre-trained printing deviation analysis model is generated; or, when the model loss value does not reach the minimum, the constraint value is back-propagated to update the model parameters of the printing deviation analysis model, and the step of inputting the sample feature vector of each training data into the parameter constraint module is continued.
8. The system according to claim 7, characterized in that The parameter constraint module includes a pre-established loss constraint function; wherein the loss constraint function is: in, is the constraint value, is the weight used to balance quality and efficiency, is the print quality loss, which quantifies the difference between the printed pattern and the expected pattern. is the model parameter The regularization term of ; where The loss function of the print quality loss is: in, is the number of samples in the training set, It is The feature vector of the key features of the samples, It is The feature vector of the historical printing parameters carrying the type label of the samples, The tag value for the type tag.
9. The system according to claim 1, characterized in that The preprocessing of the sensing data to obtain real-time printing information includes: Eliminate abnormal values or noise in the perception data through a box plot to obtain first perception data; Using linear interpolation to supplement missing values of the first perception data to obtain second perception data; The second perception data is downsampled or aggregated to reduce the amount of calculation and extract key information to obtain real-time printing information.
10. A method for dynamically adjusting printing parameters of a printing machine implemented by the system according to any one of claims 1 to 9, characterized in that: The method comprises: When receiving the printing instruction, the control system controls the guide belt module and the nozzle module to print on the cloth to be printed according to the initial printing parameters, receives the sensing data fed back by the sensor module in real time, and sends the sensing data and the initial printing parameters to the server; The server receives the sensing data and initial printing parameters from the control system, pre-processes the sensing data, obtains real-time printing information, inputs the real-time printing information into a pre-trained printing deviation analysis model, outputs a deviation analysis result corresponding to the real-time printing information, and dynamically adjusts the initial printing parameters online when the deviation analysis result indicates that there is a problem with the printed pattern, and sends the adjusted latest printing parameters to the control system, wherein the pre-trained printing deviation analysis model is obtained by machine learning based on the historical printing data in the control system and the sensor module; The control system receives the latest printing parameters from the service end, and based on the latest printing parameters, continues to execute the step of controlling the guide belt module and the nozzle module to print the cloth to be printed.
Citation Information
Patent Citations
High-precision concrete 3D printer control system
CN117400385A
Printing quality improving method and device of depth vision technology and storage medium
CN117774537A
