CN117555501A

CN117555501A - Cloud printer operation and data processing method based on edge calculation and related device

Info

Publication number: CN117555501A
Application number: CN202410005446.7A
Authority: CN
Inventors: 李晨晨; 刘丹
Original assignee: Zhuhai Xinye Electronic Technology Co Ltd
Current assignee: Zhuhai Xinye Electronic Technology Co Ltd
Priority date: 2024-01-03
Filing date: 2024-01-03
Publication date: 2024-02-13
Anticipated expiration: 2044-01-03
Also published as: CN117555501B

Abstract

The embodiment of the invention provides a cloud printer operation and data processing method based on edge calculation and a related device, and belongs to the technical field of printers. The method comprises the following steps: the cloud node performs feature extraction on the target operation data to obtain first feature data; performing association analysis on each feature in the first feature data to determine cross-correlation weights between the first feature data; calculating the association degree of each feature in the first feature data and the monitoring object to obtain association weight corresponding to the first feature data; performing feature screening on the first feature data according to the cross-correlation weight and the association weight to obtain second feature data; performing feature screening on the second feature data to obtain target feature data; the edge node obtains a fault detection model according to the target characteristic data, and then performs fault detection on the current operation data according to the fault detection model to obtain a corresponding fault type, so that the operation and maintenance service platform determines a fault processing strategy of the monitoring object according to the fault type.

Description

Cloud printer operation and data processing method based on edge calculation and related device

Technical Field

The invention relates to the technical field of printers, in particular to a cloud printer operation and data processing method based on edge calculation and a related device.

Background

Printers have been widely used office equipment in daily life. With the development of printing technology, printers have made great progress in technology and function, and are also being affected by the trend of digitization, and printers have also been developed toward more efficient and intelligent directions. However, when the operation and maintenance management is performed on the printer system, the central server is adopted to collect the printer operation data and then perform fault analysis uniformly, the central server performs calculation uniformly to solve the problems of high delay, unstable network, low bandwidth and the like, and meanwhile, the problem of high concentration of risks is also solved, so that the interaction and interference between the printer operation data are caused, the operation and maintenance result is difficult to accurately capture the real fault cause, and the accuracy of the operation and maintenance result is reduced.

Disclosure of Invention

The embodiment of the invention mainly aims to provide a cloud printer operation and maintenance data processing method based on edge calculation and a related device, and aims to solve the problems that an operation and maintenance result is difficult to accurately capture a real failure cause and the accuracy of the operation and maintenance result is reduced in the related technology.

In a first aspect, an embodiment of the present invention provides a cloud printer operation and maintenance data processing method based on edge computing, which is applied to a cloud printer operation and maintenance data processing system based on edge computing, where the system includes a plurality of monitoring nodes, a plurality of edge nodes communicatively connected with the plurality of monitoring nodes, a cloud node communicatively connected with the plurality of edge nodes, and an operation and maintenance service platform communicatively connected with the edge nodes and the monitoring nodes, and includes:

the operation and maintenance service platform sends data acquisition requests to a plurality of monitoring nodes;

the monitoring node receives the data acquisition request, acquires the operation data of a monitoring object according to the data acquisition request, obtains target operation data corresponding to the monitoring object, and sends the target operation data to the cloud node through the corresponding edge node; each monitoring node corresponds to one monitoring object;

the cloud node receives the target operation data, performs feature extraction on the target operation data, and obtains first feature data corresponding to the target operation data; performing association analysis on each feature in the first feature data to determine the cross-correlation weight between the first feature data; calculating the association degree of each feature in the first feature data and the monitoring object to obtain association weight corresponding to the first feature data; performing feature screening on the first feature data according to the cross-correlation weight and the association weight to obtain second feature data corresponding to the target operation data; performing feature screening on the second feature data to obtain target feature data corresponding to the target operation data; transmitting the target feature data to the corresponding edge node so that the edge node trains a fault detection model according to the target feature data;

The edge node receives the target feature data, classifies historical operation data corresponding to the monitoring object according to the target feature data, and obtains target training data corresponding to the monitoring node; training the fault detection model according to the target training data to obtain the fault detection model corresponding to the monitoring node;

the monitoring node obtains current operation data corresponding to the monitoring object and sends the current operation data to the edge node corresponding to the monitoring node;

the edge node receives the current operation data, performs fault detection on the current operation data according to the fault detection model, obtains a fault type corresponding to the monitoring object, and sends the fault type corresponding to the monitoring object to the operation and maintenance service platform;

the operation and maintenance service platform receives the fault type corresponding to the monitoring object, determines a fault processing strategy of the monitoring object according to the fault type corresponding to the monitoring object, and sends the fault processing strategy to a target terminal in communication connection with the operation and maintenance service platform, so that the target terminal executes corresponding processing operation according to the fault processing strategy.

In a second aspect, an edge-computing-based cloud printer operation and maintenance data processing device is applied to an edge-computing-based cloud printer operation and maintenance data processing system, where the system includes a plurality of monitoring nodes, a plurality of edge nodes communicatively connected to the monitoring nodes, a cloud node communicatively connected to the edge nodes, and an operation and maintenance service platform communicatively connected to the edge nodes and the monitoring nodes, and the operation and maintenance service platform includes:

the data request module is used for sending data acquisition requests to the plurality of monitoring nodes by the operation and maintenance service platform;

the data acquisition module is used for receiving the data acquisition request by the monitoring node, acquiring the operation data of the monitoring object according to the data acquisition request, obtaining target operation data corresponding to the monitoring object, and sending the target operation data to the cloud node through the corresponding edge node; each monitoring node corresponds to one monitoring object;

the feature extraction module is used for receiving the target operation data by the cloud node, extracting features of the target operation data and obtaining first feature data corresponding to the target operation data; performing association analysis on each feature in the first feature data to determine the cross-correlation weight between the first feature data; calculating the association degree of each feature in the first feature data and the monitoring object to obtain association weight corresponding to the first feature data; performing feature screening on the first feature data according to the cross-correlation weight and the association weight to obtain second feature data corresponding to the target operation data; performing feature screening on the second feature data to obtain target feature data corresponding to the target operation data; transmitting the target feature data to the corresponding edge node so that the edge node trains a fault detection model according to the target feature data;

The model training module is used for receiving the target feature data by the edge node, classifying the historical operation data corresponding to the monitoring object according to the target feature data, and obtaining target training data corresponding to the monitoring node; training the fault detection model according to the target training data to obtain the fault detection model corresponding to the monitoring node;

the data acquisition module is used for acquiring current operation data corresponding to the monitoring object by the monitoring node and transmitting the current operation data to the edge node corresponding to the monitoring node;

the fault monitoring module is used for receiving the current operation data by the edge node, carrying out fault detection on the current operation data according to the fault detection model, obtaining a fault type corresponding to the monitoring object, and sending the fault type corresponding to the monitoring object to the operation and maintenance service platform;

the fault processing module is used for receiving the fault type corresponding to the monitoring object by the operation and maintenance service platform, determining a fault processing strategy of the monitoring object according to the fault type corresponding to the monitoring object, and sending the fault processing strategy to a target terminal in communication connection with the operation and maintenance service platform so that the target terminal executes corresponding processing operation according to the fault processing strategy.

In a third aspect, an embodiment of the present invention further provides a terminal device, where the terminal device includes a processor, a memory, a computer program stored on the memory and executable by the processor, and a data bus for implementing connection communication between the processor and the memory, where the computer program, when executed by the processor, implements the steps of any of the edge-computing-based cloud printer operation data processing methods provided in the present specification.

In a fourth aspect, an embodiment of the present invention further provides a storage medium, where the storage medium stores one or more programs, where the one or more programs are executable by one or more processors to implement the steps of any edge-computing-based cloud printer operation data processing method as provided in the present specification.

The embodiment of the invention provides a cloud printer operation and data processing method based on edge calculation and a related device, wherein the method comprises the following steps: the operation and maintenance service platform sends data acquisition requests to a plurality of monitoring nodes; the monitoring node receives a data acquisition request, acquires the operation data of the monitoring object according to the data acquisition request, acquires target operation data corresponding to the monitoring object, and sends the target operation data to the cloud node through the corresponding edge node; and each monitoring node corresponds to one monitoring object; the cloud node receives target operation data, performs feature extraction on the target operation data, and obtains first feature data corresponding to the target operation data; performing association analysis on each feature in the first feature data to determine cross-correlation weights between the first feature data; calculating the association degree of each feature in the first feature data and the monitoring object to obtain association weight corresponding to the first feature data; feature screening is carried out on the first feature data according to the cross-correlation weight and the association weight, and second feature data corresponding to the target operation data are obtained; performing feature screening on the second feature data to obtain target feature data corresponding to the target operation data; transmitting the target feature data to the corresponding edge nodes so that the edge nodes train the fault detection model according to the target feature data; the edge node receives the target feature data, classifies historical operation data corresponding to the monitoring object according to the target feature data, and obtains target training data corresponding to the monitoring node; training a fault detection model according to the target training data to obtain a fault detection model corresponding to the monitoring node; the monitoring node obtains current operation data corresponding to the monitoring object and sends the current operation data to an edge node corresponding to the monitoring node; the edge node receives the current operation data, performs fault detection on the current operation data according to the fault detection model, obtains a fault type corresponding to the monitoring object, and sends the fault type corresponding to the monitoring object to the operation and maintenance service platform; the operation and maintenance service platform receives the fault type corresponding to the monitoring object, determines a fault processing strategy of the monitoring object according to the fault type corresponding to the monitoring object, and sends the fault processing strategy to a target terminal in communication connection with the operation and maintenance service platform, so that the target terminal executes corresponding processing operation according to the fault processing strategy. According to the cloud node, the target operation data are analyzed, so that the target feature data corresponding to the monitoring object are obtained, the edge node further obtains the fault detection model corresponding to the monitoring object according to the target feature data, and therefore after the edge node obtains the current operation data of the monitoring object corresponding to the monitoring node, the fault type of the monitoring object can be accurately obtained according to the fault detection model, and further efficient fault detection of the monitoring object is achieved, and reliability and stability of the operation and maintenance system are improved. The problem that the operation and maintenance result in the related technology is difficult to accurately capture the real fault cause and the accuracy of the operation and maintenance result is reduced is solved.

In addition, the following technical effects can be achieved by the embodiment of the invention:

1. computing power is deployed near the device side, and the device requests real-time response;

2. the work is migrated to be closer to a user or a data acquisition terminal, so that the influence caused by the limitation of network bandwidth can be reduced;

3. and the data is locally collected, locally analyzed and locally processed, so that the opportunity that the data is exposed in a public network is effectively reduced, and the data privacy is protected.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a schematic diagram of a cloud printer operation and data processing method system based on edge computing according to an embodiment of the present invention;

fig. 2 is a schematic flow chart of a cloud printer operation and data processing method based on edge computing according to an embodiment of the present invention;

fig. 3 is a schematic block diagram of a cloud printer operation and data processing device based on edge calculation according to an embodiment of the present invention;

Fig. 4 is a schematic block diagram of a structure of a terminal device according to an embodiment of the present invention.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

The flow diagrams depicted in the figures are merely illustrative and not necessarily all of the elements and operations/steps are included or performed in the order described. For example, some operations/steps may be further divided, combined, or partially combined, so that the order of actual execution may be changed according to actual situations.

It is to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

The embodiment of the invention provides a cloud printer operation and data processing method based on edge calculation and a related device. The cloud printer operation and maintenance data processing method based on edge calculation can be applied to a cloud printer operation and maintenance data processing system based on edge calculation, and as shown in fig. 1, the cloud printer operation and maintenance data processing system based on edge calculation comprises a plurality of monitoring nodes, a plurality of edge nodes in communication connection with the monitoring nodes, cloud nodes in communication connection with the edge nodes, and an operation and maintenance service platform in communication connection with the edge nodes and the monitoring nodes. The monitoring node collects operation data of a monitoring object in real time, the monitoring object can be a cloud printer or a cloud printing server, the cloud printer can be a thermal printer, a bill printer, a bar code printer and the like, the monitoring node can be a local server or other local Internet of things equipment, the edge node can be a local server or other local Internet of things equipment, and the cloud node can be a cloud server or a cloud server cluster; the operation and maintenance service platform can be a cloud server or a cloud server cluster.

In the existing print scenario: after a customer places a bill through the take-out platform such as the beauty group/hunger, the order is pushed to the restaurant order receiving system, the cloud printing server and the restaurant order receiving system perform data synchronization, take-out order information is synchronized to a cloud printer associated with a merchant, and the cloud printer prints the take-out order information. Therefore, a merchant can manufacture dishes according to the printed paper content, the merchant confirms the completion of the dishes by scanning the two-dimensional code on the paper content through the extremely-fast reporting function provided by the cloud printer, and feeds information back to the restaurant order receiving system, and once all the dishes are manufactured, the merchant wraps the dishes according to order information and notifies a takeaway rider of taking the dishes.

Under the existing printing scene, when a printing fault occurs, all operation data of a cloud printer and a cloud printing server in the printing scene are uploaded to a central server for fault analysis, the central server uniformly calculates the problems of high delay, unstable network, low bandwidth and the like, and meanwhile, the problem of high concentration of risks exists, so that when fault diagnosis is performed, the real fault cause is difficult to capture quickly and accurately, the accuracy of operation and maintenance results is reduced, and the fault problem cannot be effectively solved.

Therefore, the monitoring objects (namely, cloud printing and cloud servers) in the cloud printer operation and data processing system based on the edge calculation are respectively monitored by introducing a plurality of monitoring nodes and a plurality of edge nodes, so that fault types (such as fault types such as online abnormality of a cloud printer, abnormal receiving and sending of a printer order, abnormal execution of a printer order and the like) occurring in the cloud printer operation and data processing system based on the edge calculation can be efficiently and accurately obtained, and further the fault processing efficiency of the system is improved.

Some embodiments of the invention are described in detail below with reference to the accompanying drawings. The following embodiments and features of the embodiments may be combined with each other without conflict.

Referring to fig. 2, fig. 2 is a flow chart of a cloud printer operation and data processing method based on edge computing according to an embodiment of the present invention.

As shown in fig. 2, the cloud printer operation data processing method based on edge calculation includes steps S101 to S107.

And step S101, the operation and maintenance service platform sends data acquisition requests to a plurality of monitoring nodes.

Illustratively, the operation and maintenance service platform is a centralized management platform, and aims to improve the reliability and stability of the system and reduce the maintenance and support cost through means of automation, standardization, monitoring and the like. In the operation and maintenance service platform, an administrator can perform visual management and monitoring on the whole system so as to check the state of the system, obtain the fault type of a monitored object and timely inform a responsible person of the fault type.

For example, when the monitoring object is a cloud printer, the fault type of the monitoring object may be online abnormality, abnormal receiving and sending of a printer order, abnormal execution of the printer order, and the like; when the monitoring object is a cloud print server, the fault type of the monitoring object may be that the cloud print server itself fails or is under network attack.

The cloud printer operation and maintenance data processing system based on edge computing comprises a plurality of monitoring objects, corresponding monitoring nodes are respectively arranged on the monitoring objects, and after the operation and maintenance service platform sends data acquisition requests to the plurality of monitoring nodes, the monitoring nodes acquire operation data of the monitoring objects according to the data acquisition requests.

Step S102, the monitoring node receives the data acquisition request, acquires the operation data of a monitoring object according to the data acquisition request, obtains target operation data corresponding to the monitoring object, and sends the target operation data to the cloud node through the corresponding edge node; each monitoring node corresponds to one monitoring object.

The monitoring node receives the data acquisition request, and the monitoring node performs data acquisition on the operation data of the monitoring object of the monitoring node, so as to obtain the target operation data corresponding to the monitoring object.

In an exemplary embodiment, in the cloud printer operation and data processing system based on edge computing, a corresponding monitoring node is set for a monitored object, and meanwhile, a corresponding edge node is set for the monitored node, so that a support is provided for a fault type of a subsequently obtained monitored object, transmission time for obtaining the fault type is shortened, and transmission efficiency for obtaining the fault type is improved.

The monitoring node sends the target operation data to the edge node after obtaining the target operation data corresponding to the monitoring objects, so that the edge node uploads the target operation data to the cloud node, the cloud node receives the target operation data corresponding to each monitoring object, and the association relationship between the target operation data is obtained, so that good support is provided for the accuracy of the fault detection model corresponding to the monitoring objects.

In addition, because the operation and maintenance service platform sends data acquisition requests to the plurality of monitoring nodes in the process, the monitoring nodes receive the data acquisition requests and acquire the operation data of the monitoring objects according to the data acquisition requests, the purpose of acquiring the target operation data corresponding to the monitoring objects is to provide support for the fault detection model corresponding to the monitoring objects to be acquired later, and then the target operation data can be directly sent to the cloud nodes through the monitoring nodes.

Step S103, the cloud node receives the target operation data, performs feature extraction on the target operation data, and obtains first feature data corresponding to the target operation data; performing association analysis on each feature in the first feature data to determine the cross-correlation weight between the first feature data; calculating the association degree of each feature in the first feature data and the monitoring object to obtain association weight corresponding to the first feature data; performing feature screening on the first feature data according to the cross-correlation weight and the association weight to obtain second feature data corresponding to the target operation data; performing feature screening on the second feature data to obtain target feature data corresponding to the target operation data; and sending the target feature data to the corresponding edge node so that the edge node trains a fault detection model according to the target feature data.

The cloud node receives target operation data corresponding to each monitoring object, and the cloud node includes all target operation data corresponding to all monitoring objects in the cloud printer operation data processing system based on edge calculation.

The cloud node performs feature extraction on the target operation data by using a feature extraction model to obtain first feature data corresponding to the target operation data.

Illustratively, the feature extraction model may be constructed using a deep learning framework (e.g., tensorFlow, pyTorch, etc.), and the target operation data is input into the feature extraction model to obtain a plurality of feature representations corresponding to the target operation data, so as to select and reduce the extracted plurality of features to exclude redundant or irrelevant features in the plurality of feature representations, thereby obtaining the first feature data corresponding to the target operation data.

For example, when the first feature data corresponding to the target operation data is obtained, all the operation data corresponding to all the monitoring objects may be taken as the target operation data, so as to provide support for the target feature data required for fault analysis of the monitoring objects to be obtained later. The first feature data is used to characterize all features affecting the monitored object in the edge-computing-based cloud printer operation data processing system.

Illustratively, the cloud node prepares a dataset containing first feature data and ensures that features in the dataset have undergone the necessary data processing operations such as preprocessing and normalization. And further uses the correlation coefficient to measure the degree of correlation between features. And calculating the cross-correlation weight between every two features by using the correlation coefficient to obtain a cross-correlation matrix. And further, based on the values in the cross-correlation matrix, cross-correlation weights between features can be determined. The greater the weight, the stronger the correlation between the features.

Alternatively, the most common correlation coefficient is the pearson correlation coefficient, which can measure the linear relationship between features. The correlation coefficient has a value ranging from-1 to 1, wherein 1 represents a complete positive correlation, -1 represents a complete negative correlation, and 0 represents no linear relationship.

The cloud node obtains the relevant data containing the first feature data in the historical operation data corresponding to the monitoring object, further obtains the monitoring state corresponding to the monitoring object in the relevant data, further determines the association degree between each feature in the first feature data and the monitoring object according to the monitoring state, and further obtains the association weight corresponding to the first feature data.

The feature screening is performed on the first feature data according to the cross-correlation weight and the association weight, so that corresponding second feature data in the target operation data of the monitoring object is obtained.

For example, the first feature data is screened for the first time according to the association weight, when the association weight is lower than a first preset weight, the feature is deleted if the association between the feature and the monitored object is lower, so that when the cross-correlation weight corresponding to the feature in the rest feature data is greater than the first preset weight, the feature is also used as an influence feature corresponding to the monitored object, and therefore second feature data corresponding to the monitored object is obtained.

After the cloud node obtains the second feature data corresponding to the monitoring object, in order to reduce the feature quantity and reduce the size of the model, feature screening can be performed on the second feature data again so as to obtain target feature data corresponding to the target operation data, and the target feature data is further sent to the corresponding edge node, so that the edge node trains the fault detection model according to the target feature data.

In some embodiments, the cloud node performs association analysis on each feature in the first feature data, and determines a cross-correlation weight between the first feature data, including: the cloud node obtains history associated data corresponding to the first characteristic data; the cloud node determines first probability information corresponding to each feature in the first feature data according to the history associated data, and determines second probability information between any two features in the first feature data according to the history associated data; the cloud node determines the cross-correlation weight between the first feature data according to the first probability information and the second probability information, wherein the cross-correlation weight is used for representing the degree of interdependence between the first feature data.

Illustratively, obtaining the cross-correlation weight is used to characterize the degree of inter-dependence between any two features in the first feature data, or to characterize that the amount of information in one feature contained in another feature in the first feature data is a measure of inter-feature inter-dependence.

The cloud node obtains, from a database, history associated data including first feature data, thereby obtaining first frequency information of each feature in the first feature data in the history associated data, and divides the first frequency information by the total number of the history associated data, thereby determining first probability information corresponding to the feature.

The cloud node obtains any two features from the first feature data, calculates second frequency information of the two features in the history associated data, and divides the second frequency information by the total number of the history associated data, so as to determine second probability information corresponding to the two features.

Illustratively, after obtaining the first probability information corresponding to each feature in the first feature data and the second probability information between any two features in the first feature data, the cross-correlation weight between the first feature data is calculated using the following formula:

；

Wherein,

representing the cross-correlation weight between feature x and feature y in the first feature data, +.>Second probability information between feature x and feature y in the first feature data, +.>First probability information representing a feature x in the first feature data, ++>And the first probability information corresponding to the feature y in the first feature data is represented.

Specifically, the cloud node is utilized to acquire historical association data and calculate the cross-correlation weight between any two features in the first feature data, so that potential association relations between the features can be revealed, deeper insight and guidance are provided, and more beneficial information is provided for subsequent fault detection.

In some embodiments, the cloud node calculates a degree of association between each feature in the first feature data and the monitored object, to obtain an association weight corresponding to the first feature data, including: the cloud node performs feature representation on each feature in the first feature data to obtain a feature matrix corresponding to the first feature data; the cloud node determines the monitoring position of the monitoring object, and determines a reference vector corresponding to the monitoring object according to the monitoring position; the cloud node analyzes the association degree of the feature matrix and the reference vector to obtain queuing information corresponding to the first feature data corresponding to the reference vector; the cloud node determines the association weight corresponding to the first characteristic data in the monitoring position of the monitoring object according to the queuing information, wherein the association weight characterizes the consistency degree of the influence of the first characteristic data on the monitoring position.

The feature representation model is used for carrying out feature representation on each feature in the first feature data, so that feature vectors corresponding to each feature in the first feature data are obtained, and then the feature vectors are combined in rows or columns to obtain a feature matrix corresponding to the first feature data.

Optionally, the feature representation model may be a word2vec model or a one-hot model, which is not specifically limited in this application, and may be set by a user according to actual needs.

For example, after determining the monitoring position corresponding to the monitored object, the cloud node needs to define the reference vector of the monitored object. A reference vector is a vector used to identify a monitored object and typically includes a plurality of features or attributes. The specific definition of reference vectors needs to be dependent on the specific scenario and application requirements.

For example, if the monitoring object is a cloud printer and receives print content, the monitoring position is a link of the cloud printer receiving a print instruction, and the reference vector needs to include a communication state of a receiving end of the cloud printer, a communication state of a transmitting end transmitting the print instruction, a state of whether the print instruction contains reasonable content or not, and the like.

The cloud node obtains the historical data corresponding to the cloud printer operation and maintenance system from the database, analyzes the historical data, extracts the feature vector of the monitoring object, and calculates statistics such as an average value or a weighted average value of the feature vector as a reference vector. In addition, in the process of calculating the reference vector by using the history data, the history data conforming to the monitoring position and the required reference vector need only be selected.

For example, various similarity measurement methods may be used to measure the similarity between the reference vector and the feature matrix, such as cosine similarity, euclidean distance, and correlation coefficient, to calculate the similarity between the reference vector and the feature matrix. And calculating the similarity between each feature vector and the reference vector by using a similarity measurement method to obtain a similarity matrix between the whole feature matrix and the reference vector.

Illustratively, each reference vector is ordered according to a similarity matrix, resulting in a queue of information representing a similarity ranking between each reference vector and the feature matrix. This queuing information can be used to obtain the first characteristic data corresponding to the reference vector to be analyzed.

By way of example, by evaluating the queuing information, when any two features in the first feature data are adjacent in the queuing information, it is indicated that the two features have a similar influence on the monitoring position of the monitoring object. And determining the corresponding similarity value in the similarity matrix as the corresponding association weight of the first characteristic data in the monitoring position of the monitored object, wherein the association weight represents the consistency degree of the influence of the first characteristic data on the monitoring position.

Specifically, the associated weight may be used for the degree of influence of any feature in the first feature data on the monitored location of the monitored object. Better support can be provided for subsequent fault detection.

In some embodiments, the cloud node performs feature screening on the first feature data according to the cross-correlation weight and the association weight to obtain second feature data corresponding to the target operation data, including: the cloud node calculates the feature importance of each feature in the first feature data according to the cross-correlation weight and the association weight, and obtains the feature association degree corresponding to each feature in the first feature data; the cloud node sorts the first feature data according to the feature association degree to obtain a target sorting result; and the cloud node performs feature screening on the first feature data according to the target sequencing result to obtain second feature data corresponding to the target operation data.

The cross-correlation weight reflects the influence degree between any two features in the first feature data, and the association weight reflects the consistent influence degree of any features in the first feature data on the monitoring position of the monitored object, so that the cross-correlation weight and the association weight are combined, and feature screening is performed by adopting a feature optimization mode of feature association degree, and the problem that a single screening mode screening result is unreliable is solved.

Illustratively, the feature importance calculation is performed on each feature in the first feature data according to the cross-correlation weight and the association weight, so as to obtain the feature association degree corresponding to each feature in the first feature data, which can be obtained by adopting the following formula:

；

wherein,representing the feature association degree corresponding to the ith feature in the first feature data,/for>Representing the cross-correlation weight between the ith feature and the jth feature in the first feature data,/th feature>Representing the associated weight corresponding to the ith feature in the first feature data,/for>And (3) representing the association weight corresponding to the jth feature in the first feature data, and n represents the value corresponding to the subtraction of 1 from the corresponding feature quantity in the first feature data.

Illustratively, the first feature data may be ranked according to the calculated feature association degree to obtain the target ranking result. In general, the higher the degree of correlation of features, the more front the features are in the ranking result. According to the target sorting result, the feature ranked at the top can be selected as the most important feature, the first feature data is screened, and only the important features are reserved to obtain second feature data corresponding to the target operation data.

It should be noted that the choice of feature importance calculations and feature screening methods depends on the particular problem and dataset. Different methods may have different sensitivities and accuracies to the relevance and importance of features. Therefore, a proper calculation method is required to be selected according to actual conditions, and parameter adjustment and model training optimization are required to obtain the best result.

In some embodiments, the cloud node performs feature screening on the second feature data to obtain target feature data corresponding to the target operation data, including: the cloud node obtains the maximum value of the feature association degree in the second feature data from the target sorting result, adds the target feature corresponding to the maximum value into a target set, and deletes the target feature from the second feature data; the cloud node sequentially adds the remaining features in the second feature data to the target set according to the target sorting result, and when the preferential weight corresponding to any feature in the target set meets a preset condition, the target feature data are obtained according to the target set; wherein the preferred weights are obtained according to the following formula:

；

representing the preferred weight corresponding to the jth feature in the target set,/->Representing said feature association degree corresponding to the jth feature,/th feature>And representing the sum of the feature association degrees corresponding to all the features in the target set.

The cloud node obtains the maximum value of the feature association degree in the second feature data from the target sorting result, adds the target feature corresponding to the maximum value to the target set, and deletes the target feature from the second feature data.

For example, if the second feature data is v1, v2, v3, v4, v5, v6, or v7, the feature association degree J corresponding to v1, v2, v3, v4, v5, v6, or v7 is obtained ₁ 、J ₂ 、J ₃ 、J ₄ 、J ₅ 、J ₆ 、J _7， Then obtain J ₁ 、J ₂ 、J ₃ 、J ₄ 、J ₅ 、J ₆ 、J ₇ The maximum value of (1) is assumed to be J ₆ Then J is arranged ₆ And adding the corresponding characteristic v6 as a target characteristic into the target set, deleting the v6 from the second characteristic data, and then obtaining the current second characteristic data as v1, v2, v3, v4, v5 and v7.

The cloud node sequentially adds the remaining features in the second feature data to the target set according to the target sorting result, and obtains target feature data according to the target set when the preferential weight corresponding to any feature in the target set meets a preset condition.

Illustratively, whenWhen the threshold value is smaller than the threshold value, the feature screening is finished, and when +.>If the feature is greater than or equal to the threshold value, the feature is further screened.

For example, if the current feature of the target set is v6, the remaining second feature data v1, v2, v3, v4, v5, v7 are ranked from large to small according to the correlation degree, v7, v1, v5, v3, v2, v4 can be obtained, v7 is added to the target set, the available target set includes v6 and v7, and then calculation is performedIt can be seen that- >And then W is ₇ Comparing with threshold value, when W ₇ When less than the threshold, then the target set includes v6 and v7, when W ₇ If the value is greater than or equal to the threshold value, then the addition of the feature v5 to the target set is continued until the calculated +.>Less than a threshold.

Optionally, the setting of the threshold is not particularly limited, and the user can set the threshold according to actual requirements.

Illustratively, after the target set is obtained, the features in the target set are determined as target feature data corresponding to the monitoring position of the monitored object. After the target feature data is obtained, the target feature data is sent to the edge node corresponding to the monitoring node of the monitoring object, so that after the edge node receives the target feature data, a fault detection model corresponding to the monitoring object can be obtained according to the target feature data.

Step S104, the edge node receives the target feature data, classifies the historical operation data corresponding to the monitoring object according to the target feature data, and obtains target training data corresponding to the monitoring node; and training the fault detection model according to the target training data to obtain the fault detection model corresponding to the monitoring node.

The edge node corresponding to the monitoring node of the monitoring object receives the target feature data, classifies the historical operation data corresponding to the monitoring object according to the target feature data, so that the historical operation data contains target training data corresponding to the target feature data, and further the edge node performs fault detection model training according to the target training data, so that a fault detection model corresponding to the monitoring node is obtained.

In some embodiments, the fault detection model includes a feature representation network, a feature fusion network, and a feature classification network, and the edge node performs training of the fault detection model according to the target training data to obtain the fault detection model corresponding to the monitoring node, including: the edge node inputs the target training data to the feature representation network to obtain a feature representation vector corresponding to the target training data; the edge node inputs the characteristic representation vector to the characteristic fusion network to obtain a target representation vector corresponding to the target training data; the edge node inputs the target expression vector to the feature classification network to obtain a target classification result corresponding to the target training data; the edge node calculates a classification loss value corresponding to the target training data according to the target classification result; the edge node judges whether the classification loss value meets a preset value, and when the classification loss value does not meet the preset value, model parameters corresponding to the fault detection model are updated, and the target training data are reused for training; and when the classification loss value meets the preset value, obtaining the fault detection model corresponding to the monitoring node.

Illustratively, the edge node inputs the target training data into the feature representation network, and obtains a feature representation vector corresponding to the target training data through forward propagation of the network. The feature representation network is typically a neural network consisting of a series of layers and activation functions for extracting meaningful features from raw data.

Illustratively, the edge node inputs the feature representation vector into the feature fusion network, and obtains a target representation vector corresponding to the target training data through forward propagation of the network. The feature fusion network may further integrate feature representation vectors, capturing higher level feature and target representations.

Illustratively, the edge node inputs the target expression vector into the feature classification network, and obtains a target classification result corresponding to the target training data through forward propagation of the network. The feature classification network may classify and predict the target based on the target representation vector. And the edge node calculates a classification loss value corresponding to the target training data according to the target classification result and the real label. Common loss functions include cross entropy loss functions, mean square error loss functions, and the like.

Illustratively, the edge node determines whether a preset value is satisfied according to the classification loss value. If the classification loss value does not meet the preset value, the performance of the model is not ideal, and the parameters of the fault detection model need to be updated. The model parameters may be updated according to the gradient of the loss function using an optimization algorithm such as gradient descent. And after the model parameters are updated, the edge node trains the fault detection model by reusing the target training data. This includes inputting the target training data into a feature representation network, a feature fusion network, and a feature classification network, and back-propagating and parameter updating.

Illustratively, by iterating the above steps in a loop, the edge node may continually optimize the fault detection model so that it can accurately classify the target training data. It should be noted that the specific implementation may involve the selection of network architecture, adjustment of parameters, and design of training strategies, which need to be adjusted and optimized according to the specific scenario and data set.

In other embodiments, the fault detection model may be a fault detection model based on a GRU-GAN, where the fault detection model includes a recurrent neural network (GRU) and a generation countermeasure network (GAN), and the collected target training data is input to the GRU-GAN model to obtain an abnormal classification result corresponding to the target training data, so as to determine a fault corresponding to the monitored object according to the abnormal classification result.

Step S105, the monitoring node obtains current operation data corresponding to the monitored object, and sends the current operation data to the edge node corresponding to the monitoring node.

The monitoring node monitors the monitoring object, so that current operation data corresponding to the monitoring object in the current operation process are obtained, the current operation data are sent to the edge node corresponding to the monitoring node, and the edge node is enabled to monitor the operation state of the monitoring object.

And step S106, the edge node receives the current operation data, performs fault detection on the current operation data according to the fault detection model, obtains a fault type corresponding to the monitoring object, and sends the fault type corresponding to the monitoring object to the operation and maintenance service platform.

The edge node corresponding to the monitoring node receives current operation data, and then performs fault detection on the current operation data according to a fault detection model deployed in the edge node, so as to obtain a fault type corresponding to the monitoring object, and sends the fault type corresponding to the monitoring object to the operation and maintenance service platform.

In some embodiments, the edge node performs fault detection on the current operation data according to the fault detection model, to obtain a fault type corresponding to the monitored object, including: the edge node performs data screening on the current operation data according to the target characteristic data to obtain current associated data corresponding to the monitoring object; and the edge node performs fault analysis by utilizing the fault detection model according to the current associated data to obtain the fault type corresponding to the monitoring object.

For example, in order to reduce the interference of irrelevant data in the current operation data, the target feature data is utilized to perform data screening on the current operation data, so as to obtain current associated data corresponding to the monitored object. The target feature data are key features for screening, are associated with the monitored object and can reflect the current state of the monitored object. And screening the current operation data according to the target characteristic data, wherein the data screening can be realized by using a rule-based, statistical method or machine learning method, so as to obtain the current associated data corresponding to the monitoring object. And the edge node performs fault analysis by using a fault detection model according to the current associated data to obtain the fault type corresponding to the monitoring object.

For example, the fault types may include cloud printer online anomalies, printer order transception anomalies, printer order execution anomalies, and so forth.

Step S107, the operation and maintenance service platform receives the fault type corresponding to the monitoring object, determines a fault processing strategy of the monitoring object according to the fault type corresponding to the monitoring object, and sends the fault processing strategy to a target terminal in communication connection with the operation and maintenance service platform, so that the target terminal executes corresponding processing operation according to the fault processing strategy.

The operation and maintenance service platform is provided with a mapping table corresponding to the fault type and the fault processing strategy, and then after the operation and maintenance service platform receives the fault type corresponding to the monitoring object, the operation and maintenance service platform can search in the mapping table according to the fault type so as to determine the fault processing strategy of the monitoring object, and then the fault processing strategy is sent to a target terminal in communication connection with the operation and maintenance service platform, so that the target terminal executes corresponding processing operation according to the fault processing strategy.

Referring to fig. 3, fig. 3 is a cloud printer operation and maintenance data processing device 200 based on edge computing, provided in an embodiment of the present application, applied to a cloud printer operation and maintenance data processing system based on edge computing, where the system includes a plurality of monitoring nodes, a plurality of edge nodes communicatively connected with the monitoring nodes, a cloud node communicatively connected with the edge nodes, and an operation and maintenance service platform communicatively connected with the edge nodes and the monitoring nodes, and the cloud printer operation and maintenance data processing device 200 based on edge computing includes: the system comprises a data request module 201, a data acquisition module 202, a feature extraction module 203, a model training module 204, a data acquisition module 205, a fault monitoring module 206 and a fault processing module 207, wherein the data request module 201 is used for sending data acquisition requests to a plurality of monitoring nodes by the operation and maintenance service platform; the data acquisition module 202 is configured to receive the data acquisition request by the monitoring node, acquire operation data of a monitored object according to the data acquisition request, obtain target operation data corresponding to the monitored object, and send the target operation data to the cloud node through the corresponding edge node; each monitoring node corresponds to one monitoring object; the feature extraction module 203 is configured to receive the target operation data, perform feature extraction on the target operation data, and obtain first feature data corresponding to the target operation data; performing association analysis on each feature in the first feature data to determine the cross-correlation weight between the first feature data; calculating the association degree of each feature in the first feature data and the monitoring object to obtain association weight corresponding to the first feature data; performing feature screening on the first feature data according to the cross-correlation weight and the association weight to obtain second feature data corresponding to the target operation data; performing feature screening on the second feature data to obtain target feature data corresponding to the target operation data; transmitting the target feature data to the corresponding edge node so that the edge node trains a fault detection model according to the target feature data; the model training module 204 is configured to receive the target feature data by the edge node, and classify historical operation data corresponding to the monitored object according to the target feature data, so as to obtain target training data corresponding to the monitored node; training the fault detection model according to the target training data to obtain the fault detection model corresponding to the monitoring node; the data acquisition module 205 is configured to obtain current operation data corresponding to the monitoring object by using the monitoring node, and send the current operation data to the edge node corresponding to the monitoring node; the fault monitoring module 206 is configured to receive the current operation data by using the edge node, perform fault detection on the current operation data according to the fault detection model, obtain a fault type corresponding to the monitored object, and send the fault type corresponding to the monitored object to the operation and maintenance service platform; the fault handling module 207 is configured to receive a fault type corresponding to the monitored object, determine a fault handling policy of the monitored object according to the fault type corresponding to the monitored object, and send the fault handling policy to a target terminal communicatively connected to the operation and maintenance service platform, so that the target terminal performs a corresponding processing operation according to the fault handling policy.

In some embodiments, the feature extraction module 203 performs association analysis on each feature in the first feature data by the cloud node, and in determining the cross-correlation weight between the first feature data, performs:

the cloud node obtains history associated data corresponding to the first characteristic data;

the cloud node determines first probability information corresponding to each feature in the first feature data according to the history associated data, and determines second probability information between any two features in the first feature data according to the history associated data;

the cloud node determines the cross-correlation weight between the first feature data according to the first probability information and the second probability information, wherein the cross-correlation weight is used for representing the degree of interdependence between the first feature data.

In some embodiments, the feature extraction module 203 calculates, at a cloud node, a degree of association between each feature in the first feature data and the monitored object, and performs, in a process of obtaining an association weight corresponding to the first feature data:

the cloud node performs feature representation on each feature in the first feature data to obtain a feature matrix corresponding to the first feature data;

The cloud node determines the monitoring position of the monitoring object, and determines a reference vector corresponding to the monitoring object according to the monitoring position;

the cloud node analyzes the association degree of the feature matrix and the reference vector to obtain queuing information corresponding to the first feature data corresponding to the reference vector;

the cloud node determines the association weight corresponding to the first characteristic data in the monitoring position of the monitoring object according to the queuing information, wherein the association weight characterizes the consistency degree of the influence of the first characteristic data on the monitoring position.

In some embodiments, the feature extraction module 203 performs, in a process of the cloud node performing feature screening on the first feature data according to the cross-correlation weight and the association weight to obtain second feature data corresponding to the target operation data, the following steps:

the cloud node calculates the feature importance of each feature in the first feature data according to the cross-correlation weight and the association weight, and obtains the feature association degree corresponding to each feature in the first feature data;

the cloud node sorts the first feature data according to the feature association degree to obtain a target sorting result;

And the cloud node performs feature screening on the first feature data according to the target sequencing result to obtain second feature data corresponding to the target operation data.

In some embodiments, the feature extraction module 203 performs, in a process of performing feature screening on the second feature data by using a cloud node to obtain target feature data corresponding to the target operation data, the following steps:

the cloud node obtains the maximum value of the feature association degree in the second feature data from the target sorting result, adds the target feature corresponding to the maximum value into a target set, and deletes the target feature from the second feature data;

the cloud node sequentially adds the remaining features in the second feature data to the target set according to the target sorting result, and when the preferential weight corresponding to any feature in the target set meets a preset condition, the target feature data are obtained according to the target set;

wherein the preferred weights are obtained according to the following formula:

；/>

representing the preferred weight corresponding to the jth feature in the target set,/->Representing said feature association degree corresponding to the jth feature,/th feature >And representing the sum of the feature association degrees corresponding to all the features in the target set.

In some embodiments, the fault detection model includes a feature representation network, a feature fusion network, and a feature classification network, and the model training module 204 performs, in the process that the edge node performs the fault detection model training according to the target training data to obtain the fault detection model corresponding to the monitoring node:

the edge node inputs the target training data to the feature representation network to obtain a feature representation vector corresponding to the target training data;

the edge node inputs the characteristic representation vector to the characteristic fusion network to obtain a target representation vector corresponding to the target training data;

the edge node inputs the target expression vector to the feature classification network to obtain a target classification result corresponding to the target training data;

the edge node calculates a classification loss value corresponding to the target training data according to the target classification result;

the edge node judges whether the classification loss value meets a preset value, and when the classification loss value does not meet the preset value, model parameters corresponding to the fault detection model are updated, and the target training data are reused for training; and when the classification loss value meets the preset value, obtaining the fault detection model corresponding to the monitoring node.

In some embodiments, the fault monitoring module 206 performs, in the process that the edge node performs fault detection on the current operation data according to the fault detection model to obtain the fault type corresponding to the monitored object, the following steps:

the edge node performs data screening on the current operation data according to the target characteristic data to obtain current associated data corresponding to the monitoring object;

and the edge node performs fault analysis by utilizing the fault detection model according to the current associated data to obtain the fault type corresponding to the monitoring object.

In some embodiments, the cloud printer operation data processing apparatus 200 based on edge computing may be applied to a terminal device.

It should be noted that, for convenience and brevity of description, the specific working process of the above-described cloud printer operation data processing apparatus 200 based on edge computing may refer to the corresponding process in the foregoing embodiment of the cloud printer operation data processing method based on edge computing, which is not described herein again.

Referring to fig. 4, fig. 4 is a schematic block diagram of a structure of a terminal device according to an embodiment of the present invention.

As shown in fig. 4, the terminal device 300 includes a processor 301 and a memory 302, the processor 301 and the memory 302 being connected by a bus 303, such as an I2C (Inter-integrated Circuit) bus.

In particular, the processor 301 is used to provide computing and control capabilities, supporting the operation of the entire terminal device. The processor 301 may be a central processing unit (Central Processing Unit, CPU), the processor 301 may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field-programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. Wherein the general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

Specifically, the Memory 302 may be a Flash chip, a Read-Only Memory (ROM) disk, an optical disk, a U-disk, a removable hard disk, or the like.

It will be appreciated by those skilled in the art that the structure shown in fig. 4 is merely a block diagram of a portion of the structure related to the embodiment of the present invention, and does not constitute a limitation of the terminal device to which the embodiment of the present invention is applied, and that a specific server may include more or less components than those shown in the drawings, or may combine some components, or have a different arrangement of components.

The processor is used for running a computer program stored in the memory, and implementing any one of the cloud printer operation data processing methods based on edge computing provided by the embodiment of the invention when the computer program is executed.

In an embodiment, the processor is configured to run a computer program stored in a memory and to implement the following steps when executing the computer program:

In some embodiments, the processor 301 performs, in the process of determining the cross-correlation weights between the first feature data by performing association analysis between the cloud node and each feature in the first feature data, the following steps:

In some embodiments, the processor 301 calculates, at a cloud node, a degree of association between each feature in the first feature data and the monitored object, and performs:

In some embodiments, the processor 301 performs, in a process of the cloud node performing feature screening on the first feature data according to the cross-correlation weight and the association weight to obtain second feature data corresponding to the target operation data, the following steps:

In some embodiments, in the process that the cloud node performs feature screening on the second feature data to obtain the target feature data corresponding to the target operation data, the processor 301 performs:

wherein the preferred weights are obtained according to the following formula:

；

In some embodiments, the fault detection model includes a feature representation network, a feature fusion network, and a feature classification network, and the processor 301 performs, in the process that the edge node performs the training of the fault detection model according to the target training data to obtain the fault detection model corresponding to the monitoring node:

In some embodiments, the processor 301 performs, in the process that the edge node obtains the fault type corresponding to the monitored object, performing fault detection on the current operation data according to the fault detection model:

It should be noted that, for convenience and brevity of description, a specific working process of the above-described terminal device may refer to a corresponding process in the foregoing embodiment of the operation and data processing method of the cloud printer based on edge computing, which is not described herein again.

Embodiments of the present invention also provide a storage medium for computer readable storage, where the storage medium stores one or more programs, where the one or more programs are executable by one or more processors to implement the steps of any of the edge-computing-based cloud printer operation and dimension processing methods as provided in the embodiments of the present invention.

The storage medium may be an internal storage unit of the terminal device according to the foregoing embodiment, for example, a hard disk or a memory of the terminal device. The storage medium may also be an external storage device of the terminal device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card) or the like, which are provided on the terminal device.

Those of ordinary skill in the art will appreciate that all or some of the steps, systems, functional modules/units in the apparatus, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. In a hardware embodiment, the division between the functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be performed cooperatively by several physical components. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as known to those skilled in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Furthermore, as is well known to those of ordinary skill in the art, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media.

It should be understood that the term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations. It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.

The foregoing embodiment numbers of the present invention are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments. While the invention has been described with reference to certain preferred embodiments, it will be understood by those skilled in the art that various changes and substitutions may be made therein without departing from the spirit and scope of the invention as defined by the appended claims. Therefore, the protection scope of the invention is subject to the protection scope of the claims.

Claims

1. The cloud printer operation and maintenance data processing method based on edge calculation is characterized by being applied to a cloud printer operation and maintenance data processing system based on edge calculation, wherein the system comprises a plurality of monitoring nodes, a plurality of edge nodes in communication connection with the monitoring nodes, a cloud node in communication connection with the edge nodes, and an operation and maintenance service platform in communication connection with the edge nodes and the monitoring nodes, and the method comprises the following steps:

2. The method of claim 1, wherein the cloud node performs association analysis on each feature in the first feature data, and determining a cross-correlation weight between the first feature data comprises:

3. The method of claim 1, wherein the cloud node calculates a degree of association between each feature in the first feature data and the monitored object, and obtains an association weight corresponding to the first feature data, including:

4. The method of claim 1, wherein the cloud node performs feature screening on the first feature data according to the cross-correlation weight and the association weight to obtain second feature data corresponding to the target operation data, and the method comprises:

5. The method of claim 4, wherein the feature screening the second feature data by the cloud node to obtain target feature data corresponding to the target operation data comprises:

wherein the preferred weights are obtained according to the following formula:

；

6. The method according to claim 1, wherein the fault detection model includes a feature representation network, a feature fusion network, and a feature classification network, the edge node performs the training of the fault detection model according to the target training data, and the obtaining the fault detection model corresponding to the monitoring node includes:

7. The method according to claim 5, wherein the edge node performs fault detection on the current operation data according to the fault detection model to obtain a fault type corresponding to the monitored object, including:

8. The cloud printer operation and maintenance data processing device based on edge calculation is characterized by being applied to a cloud printer operation and maintenance data processing system based on edge calculation, wherein the system comprises a plurality of monitoring nodes, a plurality of edge nodes in communication connection with the monitoring nodes, a cloud node in communication connection with the edge nodes, and an operation and maintenance service platform in communication connection with the edge nodes and the monitoring nodes, and the operation and maintenance service platform comprises the following components:

9. A terminal device, characterized in that the terminal device comprises a processor and a memory;

the memory is used for storing a computer program;

the processor is configured to execute the computer program and implement the cloud printer operation data processing method based on edge computing as claimed in any one of claims 1 to 7 when the computer program is executed.

10. A computer storage medium for computer storage, wherein the computer storage medium stores one or more programs executable by one or more processors to implement the steps of the edge computing-based cloud printer dimension processing method of any of claims 1 to 7.