Printing communication defect detection method and system based on machine learning
Through the machine learning-based printed communication defect detection method, the problem of difficulty in identifying complex exceptions and quickly processing defects in the prior art is solved, and higher detection accuracy and response efficiency are achieved, and device stability and intelligent management are enhanced.
Patent Information
- Application Number
- CN202510145673.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-08
- Publication Date
- 2025-05-23
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing printing communication defect detection technology relies on rule-driven methods, making it difficult to identify complex and variable abnormal situations, and the ability to extract data features is limited, resulting in high false alarms and missed alarm rates, and the inability to quickly locate and process defects, delaying problem solving time.
Using machine learning-based printing communication defect detection method, we use the acquisition of historical printed communication execution data, determine defects and successful data, build a defect detection model, collect real-time data to generate defect reports, ensure the comprehensiveness and pertinence of data processing, and improve the reliability and adaptability of the model.
It improves the accuracy and comprehensiveness of defect classification, analysis and diagnosis, improves the efficiency of defect detection and response, enhances the operation stability of printing communication equipment, and achieves more accurate and intelligent equipment management.
Smart Images
Figure CN120030388A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of digital information transmission, and in particular to a printing communication defect detection method and system based on machine learning. Background Art
[0002] Existing printing communication defect detection technologies mostly rely on rule-driven methods, which detect anomalies in device communications by setting fixed parameters and thresholds. Such methods rely on manual experience and prior knowledge of the device's operating mode, and can only identify known, fixed-mode defects, making it difficult to cope with complex and changeable abnormal situations in the communication environment. In addition, traditional methods have limited ability to extract data features, making it difficult to capture potential abnormal patterns from multi-dimensional data, resulting in high false alarm and missed alarm rates. In terms of real-time performance, traditional methods usually require manual intervention or offline analysis, and are unable to quickly locate and handle defects, delaying problem resolution time, and thus affecting system stability and production efficiency.
[0003] Therefore, the present invention provides a printing communication defect detection method and system based on machine learning. Summary of the invention
[0004] The present invention provides a printing communication defect detection method and system based on machine learning. By determining defective printing communication execution data and successful printing communication execution data, determining standard feature data and defect feature data, building a defect detection model, collecting real-time printing communication execution data of printing communication equipment and generating a real-time printing communication defect report, the comprehensiveness and pertinence of data processing can be ensured, the reliability and adaptability of the model can be improved, the accuracy and comprehensiveness of defect classification, analysis and diagnosis can be improved, the defect detection, response efficiency and intelligence level can be improved, the operating stability of printing communication equipment can be enhanced, and more accurate and intelligent equipment management can be achieved.
[0005] In one aspect, the present invention provides a method for detecting printed communication defects based on machine learning, comprising: 101: Acquire historical printing communication execution data of a printing communication device, and determine defective printing communication execution data and successful printing communication execution data based on the historical printing communication execution data; 102: determining standard feature data based on the successful printing communication execution data, and determining defect feature data based on the defective printing communication execution data; 103: Building a defect detection model based on historical printing communication execution data, standard feature data, and defect feature data; 104: Collecting real-time printing communication execution data of the printing communication device, and generating a real-time printing communication defect report based on the real-time printing communication execution data and the defect detection model.
[0006] According to a printing communication defect detection method based on machine learning provided by the present invention, historical printing communication execution data includes multiple printing communication tasks, multiple printing communication execution data and printing communication parameters of each printing communication task, and a printing communication identifier of each printing communication execution data, wherein the printing communication identifier includes a communication timestamp and a communication status identifier of the corresponding printing communication execution data.
[0007] According to a printing communication defect detection method based on machine learning provided by the present invention, defective printing communication execution data and successful printing communication execution data are determined based on historical printing communication execution data, including: Preprocessing the historical printing communication execution data, and determining the printing communication sub-data of each printing communication task based on all the printing communication execution data of each printing communication task in the preprocessed historical printing communication tasks, the printing communication identifiers of all the printing communication execution data, and the printing communication parameters; Extracting all printing communication execution data with a communication status marked as successful from the printing communication sub-data of each printing communication task, and determining successful printing communication sub-data of each printing communication task; Extracting all printing communication execution data except for the communication status marked as successful from the printing communication sub-data of each printing communication task, and determining defective printing communication sub-data of each printing communication task; Successful print communication execution data is determined based on the successful print communication sub-data of all print communication tasks, and defective print communication execution data is determined based on the defective print communication sub-data of all print communication tasks.
[0008] According to a method for detecting printing communication defects based on machine learning provided by the present invention, a preset printing standard is determined based on successful printing communication execution data, including: Based on the communication timestamps of all the printing communication execution data in the successful printing communication sub-data of each printing communication task in the successful printing communication execution data, dividing the successful printing communication sub-data of each printing communication task into a plurality of time sub-data; Extracting features from each piece of printing communication execution data in the successful printing communication execution data, and determining a success feature vector for each piece of printing communication execution data in the successful printing communication execution data; determining a fitting feature vector for each printing communication task based on the success feature vectors of all printing communication execution data in the successful printing communication sub-data of each printing communication task; determining a standard feature vector for each printed communication task based on the fitted feature vector for each printed communication task and all the time sub-data; in, represents the standard feature vector for the ith printed communication task, represents the first sub-criteria feature vector of the i-th printed communication task, represents the second sub-criteria feature vector for the i-th printed communication task, represents the third substandard feature vector of the i-th printing communication task, iN1 represents the number of time sub-data in the successful printing communication sub-data of the i-th printing communication task, ijN2 represents the number of printing communication execution data in the j-th time sub-data in the successful printing communication sub-data of the i-th printing communication task, N3 represents the number of eigenvalues in the success feature vector, represents the standard feature vector to be optimized for the i-th printed communication task, represents the dth eigenvalue of the standard eigenvector to be optimized for the i-th printed communication task, Represents the standard feature vector to be optimized for the i-th printed communication task At the dth eigenvalue The partial derivative on , represents the weight of the second sub-criteria feature vector of the i-th printed communication task, represents the second adjustment parameter, represents the dth eigenvalue of the fitted eigenvector for the ith printed communication task, represents the weight of the third sub-criteria feature vector of the i-th printed communication task, represents the first adjustment parameter, represents the time decay factor of the i-th printing communication task, represents the successful feature vector of the kth printing communication execution data in the jth time sub-data in the successful printing communication sub-data of the i-th printing communication task, represents the P-norm distance between the success feature vector of the k-th printing communication execution data in the j-th time sub-data in the successful printing communication sub-data of the i-th printing communication task and the standard feature vector to be optimized of the i-th printing communication task; Based on a standard feature vector for all printed communication tasks, standard feature data are determined.
[0009] According to a printing communication defect detection method based on machine learning provided by the present invention, defect feature data is determined based on defective printing communication execution data and standard feature data, including: Based on the communication status identifiers of all printing communication execution data in the defective printing communication sub-data of each printing communication task, defect classification is performed on the printing communication sub-data of each printing communication task to determine a plurality of categories of defective data for each printing communication task; Extracting features from each piece of defective printing communication execution data to determine a defect feature vector for each piece of defective printing communication execution data; determining a comprehensive feature vector of each category of defect data for each printing communication task based on the defect feature vectors of all printing communication execution data in each category of defect data for each printing communication task and the standard feature vector of each printing communication task in the standard feature data; in, The comprehensive feature vector of the a-th category defect data of the i-th printing communication task, The first sub-comprehensive feature vector representing the a-th category defect data of the i-th printing communication task, The second sub-comprehensive feature vector representing the a-th category defect data of the i-th printing communication task, represents the third sub-comprehensive feature vector of the a-th category defect data of the i-th printing communication task, iaN4 represents the number of printing communication execution data in the a-th category defect data of the i-th printing communication task, The defect feature vector representing the bth printing communication execution data in the ath category defect data of the i-th printing communication task, represents the comprehensive feature vector of the a-th category defect data of the i-th printing communication task, represents the weight of the second sub-comprehensive feature vector, represents the weight of the third sub-comprehensive eigenvector, represents the interval parameter, The p-norm distance between the defect feature vector of the b-th printing communication execution data in the a-th category defect data of the i-th printing communication task and the comprehensive feature vector of the a-th category defect data of the i-th printing communication task; Determining defect feature sub-data for each printing communication task based on a comprehensive feature vector of all categories of defect data for each printing communication task; Defect characteristic data are determined based on the defect characteristic sub-data of all print communication tasks.
[0010] According to a printing communication defect detection method based on machine learning provided by the present invention, a defect detection model is constructed based on historical printing communication execution data, standard feature data and defect feature data, including: Inputting standard feature data and defect feature data into a defect detection model; Divide historical printing communication execution data into training set and test set, input the training set into the defect detection model, and perform model training based on machine learning; The test set is input into the defect detection model, and the defect detection model outputs a predicted state identification of each printed communication execution data in the test set; The communication state identifier and the predicted state identifier in the printing communication identifier of each printing communication execution data in the test set are compared, and the defect detection model is evaluated and optimized based on the comparison result.
[0011] According to a printing communication defect detection method based on machine learning provided by the present invention, real-time printing communication execution data of a printing communication device is collected, and a real-time printing communication defect report is generated based on the real-time printing communication execution data and a defect detection model, including: inputting the real-time printing communication execution data into the defect detection model, the defect detection model outputting a predicted state identifier of each real-time printing communication execution data in the real-time printing communication execution data; A real-time printing communication defect report is generated based on the predicted state identification of all the real-time printing communication execution data in the real-time printing communication execution data.
[0012] On the other hand, the present invention also provides a printed communication defect detection system based on machine learning, comprising: An acquisition module: acquires historical printing communication execution data of a printing communication device, and determines defective printing communication execution data and successful printing communication execution data based on the historical printing communication execution data; Determination module: determining standard characteristic data based on successful printing communication execution data, and determining defect characteristic data based on defect printing communication execution data; Building module: Building a defect detection model based on historical printing communication execution data, standard feature data, and defect feature data; Generation module: collects real-time printing communication execution data of printing communication equipment, and generates a real-time printing communication defect report based on the real-time printing communication execution data and the defect detection model.
[0013] Compared with the prior art, the present invention has the following beneficial effects: By determining the defective printing communication execution data and the successful printing communication execution data, determining the standard feature data and the defect feature data, building a defect detection model, collecting the real-time printing communication execution data of the printing communication equipment and generating a real-time printing communication defect report, it is possible to ensure the comprehensiveness and pertinence of data processing, improve the reliability and adaptability of the model, improve the accuracy and comprehensiveness of defect classification, analysis and diagnosis, improve the defect detection, response efficiency and intelligence level, enhance the operating stability of the printing communication equipment, and achieve more accurate and intelligent equipment management. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0015] Figure 1 It is a flow chart of a printing communication defect detection method based on machine learning provided in an embodiment of the present invention.
[0016] Figure 2 It is a structural schematic diagram of a printing communication defect detection system based on machine learning provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0017] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0018] Embodiment 1: The embodiment of the present invention provides a printing communication defect detection method based on machine learning, such as Figure 1 As shown, including: 101: Acquire historical printing communication execution data of a printing communication device, and determine defective printing communication execution data and successful printing communication execution data based on the historical printing communication execution data; 102: determining standard feature data based on the successful printing communication execution data, and determining defect feature data based on the defective printing communication execution data; 103: Building a defect detection model based on historical printing communication execution data, standard feature data, and defect feature data; 104: Collecting real-time printing communication execution data of the printing communication device, and generating a real-time printing communication defect report based on the real-time printing communication execution data and the defect detection model.
[0019] In this embodiment, historical printing communication execution data is extracted from the operation record of the printing communication device, and two types of data are extracted by classifying and analyzing the acquired data: defective printing communication execution data and successful printing communication execution data.
[0020] In this embodiment, a defect detection model is constructed using a machine learning method based on historical printing communication execution data and its corresponding status labels, combined with standard feature data and defect feature data.
[0021] In this embodiment, real-time printing communication execution data generated by the current printing communication device is collected in real time and input into the defect detection model for status analysis.
[0022] The beneficial effects of the above technical solution are as follows: by determining defective printing communication execution data and successful printing communication execution data, determining standard feature data and defect feature data, building a defect detection model, collecting real-time printing communication execution data of printing communication equipment and generating a real-time printing communication defect report, the accuracy of model training can be improved, the reliability and timeliness of online detection can be ensured, the defect detection and response efficiency can be improved, and the operating stability of printing communication equipment can be enhanced.
[0023] Embodiment 2: An embodiment of the present invention provides a printing communication defect detection method based on machine learning, wherein historical printing communication execution data includes multiple printing communication tasks, multiple printing communication execution data and printing communication parameters of each printing communication task, and a printing communication identifier of each printing communication execution data, wherein the printing communication identifier includes a communication timestamp and a communication status identifier of the corresponding printing communication execution data.
[0024] In this embodiment, the historical printing communication execution data represents all operation data recorded by the device during the execution of multiple printing communication tasks.
[0025] In this embodiment, a printing communication task may be executed multiple times (including the start, processing and completion of the task), and each complete execution operation is recorded as an execution data, which includes the status of the device, communication content and other parameter information, constituting a complete operation record of the task.
[0026] In this embodiment, the core parameters of each printing communication task include information such as the target requirements, equipment settings, and operating environment of the task, which are used to describe the specific conditions and constraints of the communication task.
[0027] In this embodiment, each piece of execution data is associated with a unique identifier for tracking its task source and operation result.
[0028] In this embodiment, the printed communication identifier is composed of two key parts: a communication timestamp: recording the time point when the data is generated; and a communication status identifier: indicating the communication status, such as "success" or "failure", which is used to distinguish the task execution results.
[0029] The beneficial effects of the above technical solution are: determining the historical printing communication execution data can ensure the real-time and comprehensiveness of the data.
[0030] Embodiment 3: An embodiment of the present invention provides a printing communication defect detection method based on machine learning, which determines defective printing communication execution data and successful printing communication execution data based on historical printing communication execution data, including: Preprocessing the historical printing communication execution data, and determining the printing communication sub-data of each printing communication task based on all the printing communication execution data of each printing communication task in the preprocessed historical printing communication tasks, the printing communication identifiers of all the printing communication execution data, and the printing communication parameters; Extracting all printing communication execution data with a communication status marked as successful from the printing communication sub-data of each printing communication task, and determining successful printing communication sub-data of each printing communication task; Extracting all printing communication execution data except for the communication status marked as successful from the printing communication sub-data of each printing communication task, and determining defective printing communication sub-data of each printing communication task; Successful print communication execution data is determined based on the successful print communication sub-data of all print communication tasks, and defective print communication execution data is determined based on the defective print communication sub-data of all print communication tasks.
[0031] In this embodiment, the original historical printing communication execution data is cleaned, standardized and structured, including removing redundant information, correcting abnormal data, aligning timestamps and parameters, and other operations.
[0032] The preprocessed data is more accurate and consistent, providing a basis for subsequent data analysis and model building.
[0033] In this embodiment, all printing communication execution data of each task, its corresponding printing communication identifier (including timestamp and status identifier) and related printing communication parameters are integrated according to the pre-processed historical data to generate printing communication sub-data.
[0034] In this embodiment, all execution data with a communication status marked as "successful" are screened out from the sub-data of each task. These data reflect the normal execution of the task operation and are aggregated to form successful printing communication sub-data.
[0035] In this embodiment, all execution data with a communication status flagged as non-"successful" are screened out from the sub-data of each task. These data correspond to problems in the operation of the equipment and are aggregated to form defective printing communication sub-data.
[0036] In this embodiment, the successful printing communication sub-data of all tasks are integrated, and the defective printing communication sub-data of all tasks are integrated to extract the key features and patterns of defects or anomalies in system operation.
[0037] The beneficial effects of the above technical solution are as follows: determining defective printing communication execution data and successful printing communication execution data based on historical printing communication execution data can ensure the comprehensiveness and pertinence of data processing and improve the efficiency and reliability of defect detection and diagnosis.
[0038] Embodiment 4: The embodiment of the present invention provides a printing communication defect detection method based on machine learning, which determines a preset printing standard based on successful printing communication execution data, including: Based on the communication timestamps of all the printing communication execution data in the successful printing communication sub-data of each printing communication task in the successful printing communication execution data, dividing the successful printing communication sub-data of each printing communication task into a plurality of time sub-data; Extracting features from each piece of printing communication execution data in the successful printing communication execution data, and determining a success feature vector for each piece of printing communication execution data in the successful printing communication execution data; determining a fitting feature vector for each printing communication task based on the success feature vectors of all printing communication execution data in the successful printing communication sub-data of each printing communication task; determining a standard feature vector for each printed communication task based on the fitted feature vector for each printed communication task and all the time sub-data; in, represents the standard feature vector for the ith printed communication task, represents the first sub-criteria feature vector of the i-th printed communication task, represents the second sub-criteria feature vector for the i-th printed communication task, represents the third substandard feature vector of the i-th printing communication task, iN1 represents the number of time sub-data in the successful printing communication sub-data of the i-th printing communication task, ijN2 represents the number of printing communication execution data in the j-th time sub-data in the successful printing communication sub-data of the i-th printing communication task, N3 represents the number of eigenvalues in the success feature vector, represents the standard feature vector to be optimized for the i-th printed communication task, represents the dth eigenvalue of the standard eigenvector to be optimized for the i-th printed communication task, Represents the standard feature vector to be optimized for the i-th printed communication task At the dth eigenvalue The partial derivative on , represents the weight of the second sub-criteria feature vector of the i-th printed communication task, represents the second adjustment parameter, represents the dth eigenvalue of the fitted eigenvector for the ith printed communication task, represents the weight of the third sub-criteria feature vector of the i-th printed communication task, represents the first adjustment parameter, represents the time decay factor of the i-th printing communication task, represents the successful feature vector of the kth printing communication execution data in the jth time sub-data in the successful printing communication sub-data of the i-th printing communication task, represents the P-norm distance between the success feature vector of the k-th printing communication execution data in the j-th time sub-data in the successful printing communication sub-data of the i-th printing communication task and the standard feature vector to be optimized of the i-th printing communication task; Based on a standard feature vector for all printed communication tasks, standard feature data are determined.
[0039] In this embodiment, the value of the norm p is different. The calculation method is different. p=2 means calculating the Euclidean distance to smooth the error sensitivity; p=1 means calculating the absolute error distance to enhance the robustness to outliers; p>2 means it is more sensitive to large deviations.
[0040] In this embodiment, based on the communication timestamp of the successful printing communication sub-data of each printing communication task in the successful printing communication execution data, the sub-data is segmented in time sequence to form a plurality of time sub-data.
[0041] In this embodiment, each execution data (single successful communication record) in the successful printing communication execution data is analyzed, key features in task execution are extracted therefrom, and a corresponding success feature vector is generated to describe the characteristics of each success data.
[0042] In this embodiment, all successful feature vectors in the successful printing communication sub-data of each printing communication task are comprehensively analyzed, and a fitting feature vector representing the overall characteristics of the task is generated through fitting (such as aggregation, weighted average and other algorithms). This vector reflects the comprehensive characteristics of the task when it is successfully executed.
[0043] In this embodiment, the standard feature vectors of all printing and communication tasks are integrated to generate overall standard feature data.
[0044] The beneficial effects of the above technical solution are as follows: determining the preset printing standard based on the successful printing communication execution data can provide a reliable data basis for defect detection, effectively improve the defect identification efficiency of equipment operation, and achieve more accurate and intelligent equipment management.
[0045] Embodiment 5: An embodiment of the present invention provides a printing communication defect detection method based on machine learning, which determines defect feature data based on defective printing communication execution data and standard feature data, including: Based on the communication status identifiers of all printing communication execution data in the defective printing communication sub-data of each printing communication task, defect classification is performed on the printing communication sub-data of each printing communication task to determine a plurality of categories of defective data for each printing communication task; Extracting features from each piece of defective printing communication execution data to determine a defect feature vector for each piece of defective printing communication execution data; determining a comprehensive feature vector of each category of defect data for each printing communication task based on the defect feature vectors of all printing communication execution data in each category of defect data for each printing communication task and the standard feature vector of each printing communication task in the standard feature data; in, The comprehensive feature vector of the a-th category defect data of the i-th printing communication task, The first sub-comprehensive feature vector representing the a-th category defect data of the i-th printing communication task, The second sub-comprehensive feature vector representing the a-th category defect data of the i-th printing communication task, represents the third sub-comprehensive feature vector of the a-th category defect data of the i-th printing communication task, iaN4 represents the number of printing communication execution data in the a-th category defect data of the i-th printing communication task, The defect feature vector representing the bth printing communication execution data in the ath category defect data of the i-th printing communication task, represents the comprehensive feature vector of the a-th category defect data of the i-th printing communication task, represents the weight of the second sub-comprehensive feature vector, represents the weight of the third sub-comprehensive eigenvector, represents the interval parameter, The p-norm distance between the defect feature vector of the b-th printing communication execution data in the a-th category defect data of the i-th printing communication task and the comprehensive feature vector of the a-th category defect data of the i-th printing communication task; Determining defect feature sub-data for each printing communication task based on a comprehensive feature vector of all categories of defect data for each printing communication task; Defect characteristic data are determined based on the defect characteristic sub-data of all print communication tasks.
[0046] In this embodiment, the printing communication sub-data of each task is classified according to the communication status identifier of each execution data in the defective printing communication sub-data, and the defective execution data in different states are classified into multiple categories of defect data (such as hardware failure, communication timeout, etc.), and each category of defect data corresponds to a defect type.
[0047] In this embodiment, each execution data in the defective printing communication execution data is analyzed, key information reflecting the defect characteristics is extracted, and a defect feature vector is generated to describe the specific characteristics of each defective execution data.
[0048] In this embodiment, the comprehensive feature vectors of all categories of defect data of each task are integrated to generate defect feature sub-data of the task.
[0049] In this embodiment, the defect feature sub-data of all tasks are integrated and summarized to generate overall defect feature data.
[0050] The beneficial effects of the above technical solution are as follows: by determining defect feature data based on defective printing communication execution data and standard feature data, the accuracy and comprehensiveness of defect classification, analysis and diagnosis can be improved, providing an important basis for intelligent defect detection.
[0051] Embodiment 6: The embodiment of the present invention provides a printing communication defect detection method based on machine learning, which builds a defect detection model based on historical printing communication execution data, standard feature data and defect feature data, including: Inputting standard feature data and defect feature data into a defect detection model; Divide historical printing communication execution data into training set and test set, input the training set into the defect detection model, and perform model training based on machine learning; The test set is input into the defect detection model, and the defect detection model outputs a predicted state identification of each printed communication execution data in the test set; The communication state identifier and the predicted state identifier in the printing communication identifier of each printing communication execution data in the test set are compared, and the defect detection model is evaluated and optimized based on the comparison result.
[0052] In this embodiment, standard feature data and defect feature data are provided as input to a defect detection model, wherein the standard feature data represents normal operating characteristics and the defect feature data represents abnormal characteristics.
[0053] In this embodiment, the historical printing communication execution data is divided into a training set (for model learning) and a test set (for model verification) in proportion. The data in the training set contains known communication status identifiers, and the test set is used to evaluate the prediction accuracy of the model.
[0054] In this embodiment, the training set is input into the defect detection model, and the model is trained using a machine learning algorithm (such as supervised learning) in combination with the standard feature data and the defect feature data. During the training process, the model learns how to accurately predict the communication status identifier (such as "normal" or a certain type of "defect") based on the feature data.
[0055] In this embodiment, the test set is input into a trained defect detection model, and the model predicts the communication status identifier of each data in the test set, such as normal or a specific defect category, based on the learned rules.
[0056] In this embodiment, the actual communication status identifiers in the test set are compared one by one with the status identifiers predicted by the model, and the prediction accuracy, recall rate and other indicators are calculated to evaluate the model performance. Based on the comparison results, the deficiencies of the model are analyzed and the model parameters are adjusted.
[0057] The beneficial effects of the above technical solution are as follows: constructing a defect detection model based on historical printing communication execution data, standard feature data and defect feature data can improve the reliability and adaptability of the model, thereby improving the defect detection efficiency and intelligence level of printing communication tasks.
[0058] Embodiment 7: The embodiment of the present invention provides a printing communication defect detection method based on machine learning, which collects real-time printing communication execution data of a printing communication device, and generates a real-time printing communication defect report based on the real-time printing communication execution data and a defect detection model, including: inputting the real-time printing communication execution data into the defect detection model, the defect detection model outputting a predicted state identifier of each real-time printing communication execution data in the real-time printing communication execution data; A real-time printing communication defect report is generated based on the predicted state identification of all the real-time printing communication execution data in the real-time printing communication execution data.
[0059] In this embodiment, the printing communication execution data generated in real time is input into the defect detection model that has been trained.
[0060] In this embodiment, the defect detection model analyzes each piece of real-time communication data based on the rules learned in previous training and outputs a predicted state identifier.
[0061] In this embodiment, the predicted status indicator indicates the communication status of the data, such as "normal" or some type of "defect" (such as hardware failure, timeout exception, etc.).
[0062] In this embodiment, the detection results can be classified and counted based on the predicted status identification of all real-time data, the quantity and distribution of normal data and various defect data can be listed, the defect data can be summarized by category, the frequency of occurrence, impact range and other information can be counted, and a real-time printing communication defect report can be generated.
[0063] The beneficial effects of the above technical solution are: collecting real-time printing communication execution data of printing communication equipment, generating real-time printing communication defect reports based on real-time printing communication execution data and defect detection models, which can improve operational reliability and quickly respond to real-time defects.
[0064] Embodiment 8: The embodiment of the present invention provides a printing communication defect detection system based on machine learning, such as Figure 2 As shown, including: An acquisition module: acquires historical printing communication execution data of a printing communication device, and determines defective printing communication execution data and successful printing communication execution data based on the historical printing communication execution data; Determination module: determining standard characteristic data based on successful printing communication execution data, and determining defect characteristic data based on defect printing communication execution data; Building module: Building a defect detection model based on historical printing communication execution data, standard feature data, and defect feature data; Generation module: collects real-time printing communication execution data of printing communication equipment, and generates a real-time printing communication defect report based on the real-time printing communication execution data and the defect detection model.
[0065] The beneficial effects of the above technical solution are as follows: by determining defective printing communication execution data and successful printing communication execution data, determining standard feature data and defect feature data, building a defect detection model, collecting real-time printing communication execution data of printing communication equipment and generating real-time printing communication defect reports, it is possible to ensure the comprehensiveness and pertinence of data processing, improve the reliability and adaptability of the model, improve the accuracy and comprehensiveness of defect classification, analysis and diagnosis, improve defect detection, response efficiency and intelligence level, enhance the operating stability of printing communication equipment, and achieve more accurate and intelligent equipment management.
[0066] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.
[0067] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0068] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A printed communication defect detection method based on machine learning, characterized in that: include: 101: Acquire historical printing communication execution data of a printing communication device, and determine defective printing communication execution data and successful printing communication execution data based on the historical printing communication execution data; 102: determining standard feature data based on the successful printing communication execution data, and determining defect feature data based on the defective printing communication execution data; 103: Building a defect detection model based on historical printing communication execution data, standard feature data, and defect feature data; 104: Collecting real-time printing communication execution data of the printing communication device, and generating a real-time printing communication defect report based on the real-time printing communication execution data and the defect detection model.
2. The method for detecting printed communication defects based on machine learning according to claim 1, characterized in that: The historical printing communication execution data includes multiple printing communication tasks, multiple printing communication execution data and printing communication parameters of each printing communication task, and a printing communication identifier of each printing communication execution data, wherein the printing communication identifier includes a communication timestamp and a communication status identifier of the corresponding printing communication execution data.
3. The method for detecting printed communication defects based on machine learning according to claim 2, characterized in that: Determining defective printing communication execution data and successful printing communication execution data based on historical printing communication execution data includes: Preprocessing the historical printing communication execution data, and determining the printing communication sub-data of each printing communication task based on all the printing communication execution data of each printing communication task in the preprocessed historical printing communication tasks, the printing communication identifiers of all the printing communication execution data, and the printing communication parameters; Extracting all printing communication execution data with a communication status marked as successful from the printing communication sub-data of each printing communication task, and determining successful printing communication sub-data of each printing communication task; Extracting all printing communication execution data except for the communication status marked as successful from the printing communication sub-data of each printing communication task, and determining defective printing communication sub-data of each printing communication task; Successful print communication execution data is determined based on the successful print communication sub-data of all print communication tasks, and defective print communication execution data is determined based on the defective print communication sub-data of all print communication tasks.
4. The method for detecting printed communication defects based on machine learning according to claim 3, characterized in that: Determine preset printing standards based on successful print communication execution data, including: Based on the communication timestamps of all the printing communication execution data in the successful printing communication sub-data of each printing communication task in the successful printing communication execution data, dividing the successful printing communication sub-data of each printing communication task into a plurality of time sub-data; Extracting features from each piece of printing communication execution data in the successful printing communication execution data, and determining a success feature vector for each piece of printing communication execution data in the successful printing communication execution data; determining a fitting feature vector for each printing communication task based on the success feature vectors of all printing communication execution data in the successful printing communication sub-data of each printing communication task; determining a standard feature vector for each printed communication task based on the fitted feature vector for each printed communication task and all the time sub-data; in, represents the standard feature vector for the ith printed communication task, represents the first sub-criteria feature vector of the i-th printed communication task, represents the second sub-criteria feature vector for the i-th printed communication task, represents the third substandard feature vector of the i-th printing communication task, iN1 represents the number of time sub-data in the successful printing communication sub-data of the i-th printing communication task, ijN2 represents the number of printing communication execution data in the j-th time sub-data in the successful printing communication sub-data of the i-th printing communication task, N3 represents the number of eigenvalues in the success feature vector, represents the standard feature vector to be optimized for the i-th printed communication task, represents the dth eigenvalue of the standard eigenvector to be optimized for the i-th printed communication task, Represents the standard feature vector to be optimized for the i-th printed communication task At the dth eigenvalue The partial derivative on , represents the weight of the second sub-criteria feature vector of the i-th printed communication task, represents the second adjustment parameter, represents the dth eigenvalue of the fitted eigenvector for the ith printed communication task, represents the weight of the third sub-criteria feature vector of the i-th printed communication task, represents the first adjustment parameter, represents the time decay factor of the i-th printing communication task, represents the successful feature vector of the kth printing communication execution data in the jth time sub-data in the successful printing communication sub-data of the i-th printing communication task, represents the P-norm distance between the success feature vector of the k-th printing communication execution data in the j-th time sub-data in the successful printing communication sub-data of the i-th printing communication task and the standard feature vector to be optimized of the i-th printing communication task; Based on a standard feature vector for all printed communication tasks, standard feature data are determined.
5. The method for detecting printed communication defects based on machine learning according to claim 4, characterized in that: Defect characteristic data is determined based on the defective printing communication execution data and the standard characteristic data, including: Based on the communication status identifiers of all printing communication execution data in the defective printing communication sub-data of each printing communication task, defect classification is performed on the printing communication sub-data of each printing communication task to determine a plurality of categories of defective data for each printing communication task; Extracting features from each piece of defective printing communication execution data to determine a defect feature vector for each piece of defective printing communication execution data; determining a comprehensive feature vector of each category of defect data for each printing communication task based on the defect feature vectors of all printing communication execution data in each category of defect data for each printing communication task and the standard feature vector of each printing communication task in the standard feature data; in, The comprehensive feature vector of the a-th category defect data of the i-th printing communication task, The first sub-comprehensive feature vector representing the a-th category defect data of the i-th printing communication task, The second sub-comprehensive feature vector representing the a-th category defect data of the i-th printing communication task, represents the third sub-comprehensive feature vector of the a-th category defect data of the i-th printing communication task, iaN4 represents the number of printing communication execution data in the a-th category defect data of the i-th printing communication task, The defect feature vector representing the bth printing communication execution data in the ath category defect data of the i-th printing communication task, represents the comprehensive feature vector of the a-th category defect data of the i-th printing communication task, represents the weight of the second sub-comprehensive feature vector, represents the weight of the third sub-comprehensive eigenvector, represents the interval parameter, represents the p-norm distance between the defect feature vector of the b-th printing communication execution data in the a-th category defect data of the i-th printing communication task and the comprehensive feature vector of the a-th category defect data of the i-th printing communication task; Determining defect feature sub-data for each printing communication task based on a comprehensive feature vector of all categories of defect data for each printing communication task; Defect characteristic data are determined based on the defect characteristic sub-data of all print communication tasks.
6. The method for detecting printed communication defects based on machine learning according to claim 1, characterized in that: Construct defect detection models based on historical printing communication execution data, standard feature data, and defect feature data, including: Inputting standard feature data and defect feature data into a defect detection model; Divide historical printing communication execution data into training set and test set, input the training set into the defect detection model, and perform model training based on machine learning; The test set is input into the defect detection model, and the defect detection model outputs a predicted state identification of each printed communication execution data in the test set; The communication state identifier and the predicted state identifier in the printing communication identifier of each printing communication execution data in the test set are compared, and the defect detection model is evaluated and optimized based on the comparison result.
7. The method for detecting printed communication defects based on machine learning according to claim 1, characterized in that: Collect real-time printing communication execution data of the printing communication device, and generate a real-time printing communication defect report based on the real-time printing communication execution data and the defect detection model, including: inputting the real-time printing communication execution data into the defect detection model, the defect detection model outputting a predicted state identifier of each real-time printing communication execution data in the real-time printing communication execution data; A real-time printing communication defect report is generated based on the predicted state identification of all the real-time printing communication execution data in the real-time printing communication execution data.
8. A printed communication defect detection system based on machine learning, characterized in that: include: An acquisition module: acquires historical printing communication execution data of a printing communication device, and determines defective printing communication execution data and successful printing communication execution data based on the historical printing communication execution data; Determination module: determining standard characteristic data based on successful printing communication execution data, and determining defect characteristic data based on defect printing communication execution data; Building module: Building a defect detection model based on historical printing communication execution data, standard feature data, and defect feature data; Generation module: collects real-time printing communication execution data of printing communication equipment, and generates a real-time printing communication defect report based on the real-time printing communication execution data and the defect detection model.