Equipment offline fault diagnosis method and system, electronic equipment and storage medium

Abstract

The invention discloses a device offline fault diagnosis method and system, an electronic device and a storage medium. The method comprises the steps of obtaining first feature data of a device and second feature data of a background; wherein the first feature data originates from offline state data of the equipment; performing fusion processing on the first feature data and the second feature data to obtain mixed feature data; inputting the mixed feature data into a fault prediction model to obtain a fault diagnosis result; the fault prediction model is constructed based on a decision tree. According to the method, the equipment offline fault diagnosis is performed through the first feature data and the second feature data, so that the fault diagnosis efficiency is improved and the fault diagnosis accuracy is improved. The method can be widely applied to the technical field of computers.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, system, electronic device and storage medium for off-line fault diagnosis of a device. Background Art

[0002] With the development of 5G Internet of Things technology, the number of users and related equipment of 5G enterprise-private network dual-line platform is increasing, and the operation and maintenance costs are rising. In actual work, equipment often fails due to external environment, data configuration, software, etc. In this regard, it is necessary to quickly determine the fault diagnosis type, and then find the corresponding personnel to repair it according to the fault diagnosis type, so as to improve the operation and maintenance efficiency and user experience.

[0003] In the related art, a manual assembly line method is used for troubleshooting (front-end personnel are dispatched to make judgments in turn, and then business, software or hardware personnel are contacted for diagnosis). This method is inefficient and has poor diagnostic accuracy. Summary of the invention

[0004] The main purpose of the embodiments of the present application is to provide an efficient device offline fault diagnosis method, system, electronic device and storage medium.

[0005] To achieve the above purpose, one aspect of an embodiment of the present application proposes a method for offline fault diagnosis of a device, the method comprising: obtaining first characteristic data of the device and second characteristic data of the background; wherein the first characteristic data is derived from the status data of the device offline; fusing the first characteristic data and the second characteristic data to obtain mixed characteristic data; inputting the mixed characteristic data into a fault prediction model to obtain a fault diagnosis result; the fault prediction model is constructed based on a decision tree. The embodiment of the present application performs offline fault diagnosis of the device through the first characteristic data and the second characteristic data, which is conducive to improving the efficiency of fault diagnosis and improving the accuracy of fault diagnosis.

[0006] In some embodiments, in the method provided by the embodiments of the present application, the second feature data includes statistical feature data, model feature data and implicit feature data, and obtaining the second feature data in the background includes:

[0007] Obtaining historical data reported to the backend;

[0008] Performing statistical analysis on the historical data to obtain the statistical characteristic data;

[0009] Performing fitting processing on the historical data to obtain the model characteristic data;

[0010] The historical data is subjected to feature extraction by an encoder to obtain the implicit feature data.

[0011] In some embodiments, the method provided in the embodiments of the present application, the statistical analysis of the historical data to obtain the statistical feature data includes:

[0012] Obtaining time series data from the historical data;

[0013] Determining the mean, standard deviation, skewness, and kurtosis of the time series data;

[0014] The statistical characteristic data is determined according to the mean, the standard deviation, the skewness and the kurtosis.

[0015] In some embodiments, the method provided in the embodiments of the present application, performing fitting processing on the historical data to obtain the model feature data includes:

[0016] Performing polynomial fitting processing on the historical data to obtain polynomial features;

[0017] The model feature data is determined according to the coefficients in the polynomial feature.

[0018] In some embodiments, the method provided in the embodiments of the present application, for obtaining the first characteristic data of the device, includes:

[0019] If the device fails, obtain manually filled status data;

[0020] Feature extraction is performed on the status data to obtain first feature data corresponding to the first feature; the first feature includes whether the device continues to be offline after restarting, whether the antenna is installed correctly, whether the device lights up reasonably, or whether the device background card reading is normal.

[0021] In some embodiments, in the method provided by the embodiments of the present application, the fault prediction model is determined by the following steps:

[0022] Get mixed feature samples;

[0023] Determining the information entropy and information gain ratio of the mixed feature sample;

[0024] Determining the partition feature of the current node according to the information gain ratio, and partitioning the mixed feature sample based on the partition feature to obtain a sample subset;

[0025] Based on the information gain ratio of each sample subset, each sample subset is divided until a termination condition is met to obtain a decision tree;

[0026] The fault prediction model is constructed according to the decision tree.

[0027] In some embodiments, the method provided in the embodiments of the present application, for obtaining the first characteristic data of the device, includes:

[0028] Design a questionnaire form; the questionnaire form includes a blank corresponding to the first feature to be filled in;

[0029] Receiving the completed receipt information of the questionnaire form;

[0030] Obtaining the first characteristic data according to the filling receipt information;

[0031] Alternatively, a device status system is designed; the device status system includes a control corresponding to the first feature to be fed back;

[0032] Receiving feedback information from the device status system;

[0033] The first characteristic data is obtained according to the feedback information.

[0034] To achieve the above object, another aspect of the embodiment of the present application provides a device offline fault diagnosis system, the system comprising:

[0035] The first module is used to obtain the first characteristic data of the device and the second characteristic data of the background; wherein the first characteristic data is derived from the status data of the device being offline;

[0036] A second module is used to fuse the first feature data and the second feature data to obtain mixed feature data;

[0037] The third module is used to input the mixed feature data into a fault prediction model to obtain a fault diagnosis result; the fault prediction model is constructed based on a decision tree.

[0038] To achieve the above objective, another aspect of an embodiment of the present application provides an electronic device, the electronic device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the above method when executing the computer program.

[0039] To achieve the above objective, another aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program implements the above method when executed by a processor.

[0040] The embodiments of the present application include at least the following beneficial effects: The method provided by the embodiments of the present application includes: obtaining the first characteristic data of the device and the second characteristic data of the background; wherein the first characteristic data is derived from the status data of the device offline; fusing the first characteristic data and the second characteristic data to obtain mixed characteristic data; inputting the mixed characteristic data into a fault prediction model to obtain a fault diagnosis result; the fault prediction model is constructed based on a decision tree. The embodiments of the present application perform offline fault diagnosis of the device through the first characteristic data and the second characteristic data, which is conducive to improving the efficiency of fault diagnosis and improving the accuracy of fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a flow chart of an embodiment of the device offline fault diagnosis method provided by the present application;

[0042] Figure 2 is a flow chart of another embodiment of the device offline fault diagnosis method provided by the present application;

[0043] Figure 3 This is a flowchart of an embodiment of the process of determining the second characteristic data provided by the present application.

[0044] Figure 4 This is an architectural diagram of an embodiment of the device offline fault diagnosis method provided by the present application;

[0045] Figure 5 It is a fitting schematic diagram of an embodiment of the model feature data processing process provided by the present application;

[0046] Figure 6 It is a flowchart of an embodiment of the implicit feature data processing process provided by the present application;

[0047] Figure 7 It is a structural schematic diagram of an embodiment of a decision tree provided by the present application;

[0048] Figure 8 It is a structural schematic diagram of another embodiment of the decision tree provided by the present application;

[0049] Fig. 9 It is a structural diagram of a device offline fault diagnosis system provided in an embodiment of the present application;

[0050] Fig.10 It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail below in conjunction with the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of the present application. They are only examples of devices and methods consistent with some aspects of the embodiments of the present application as detailed in the attached claims.

[0052] It is understood that the terms "first", "second", etc. used in this application can be used to describe various concepts in this article, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another concept. For example, without departing from the scope of the embodiment of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the words "if" and "if" as used herein can be interpreted as "at the time of" or "when" or "in response to determination".

[0053] The terms "at least one", "multiple", "each", "any", etc. used in this application, at least one includes one, two or more, multiple includes two or more, each refers to each of the corresponding multiple, and any refers to any one of the multiple.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0055] With the development of 5G Internet of Things technology, the number of users of 5G commercial enterprise dual-line platforms and the number of CPE devices are increasing, and the operation and maintenance costs are rising. In actual work, equipment often fails due to external environment, data configuration, software and other reasons. In order to more efficiently troubleshoot and repair equipment, it is necessary to quickly determine the fault diagnosis type, and then find the corresponding personnel to repair it according to the fault diagnosis type, so as to improve operation and maintenance efficiency and user experience.

[0056] Since there are many and complex factors that cause failures, most of them are checked by manual assembly line method (front-end personnel are sent to make judgments in turn, and then business, software or hardware personnel are contacted for diagnosis). This method is inefficient and prone to waste of manpower.

[0057] In short, since CPE equipment may fail or work abnormally, it may be caused by factors such as communication, hardware programs, software programs, card package subscriptions, etc. In order to quickly troubleshoot and restore the equipment, it is necessary to determine the fault diagnosis type. Therefore, the solution proposed in this application can quickly identify the type of equipment failure, so as to quickly find the corresponding professional personnel to locate and repair the problem, avoid other human resources and time waste, and improve the probability of equipment repair and customer satisfaction.

[0058] Specifically, this paper proposes an offline diagnosis and discrimination scheme for equipment based on hybrid feature extraction and decision tree algorithm. By using personnel to fill in relevant information about the fault, the manually filled feature data is extracted according to the current environment and performance of the equipment; then statistical features, model features and implicit feature data are extracted from the historical data reported by the background, and the above feature data are fused and spliced ​​to obtain hybrid feature data; the hybrid feature data is used to train the decision tree model to determine the current equipment fault type. Compared with traditional algorithms and similar technologies, it is difficult to directly use the original data as the input data of the decision tree model because the original data contains a large amount of time series data; at the same time, there is a large amount of redundant information in the original data, which is easy to interfere with the judgment result, making the model's discrimination performance low, while the data features processed by the model in this paper are richer and have less interference information. It not only extracts features from the equipment reported data, but also adds feature data extracted by professionals, which greatly improves the accuracy of the judgment type; in short, the technology proposed in this paper can quickly perform offline diagnosis and discrimination, thereby reducing labor costs and improving troubleshooting efficiency.

[0059] The device offline fault diagnosis method provided in the embodiment of the present application relates to the field of computer technology. The device offline fault diagnosis method provided in the embodiment of the present application can be applied to a terminal, can also be applied to a server, and can also be software running in a terminal or a server. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, and a car terminal, etc., but is not limited to this; the server side can be configured as an independent physical server, or it can be configured as a server cluster or a distributed system composed of multiple physical servers, and can also be configured as a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network; the software can be an application that implements the device offline fault diagnosis method, etc., but is not limited to the above forms.

[0060] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0061] It should be noted that in each specific implementation of the present application, when it comes to the need to perform relevant processing based on data related to user identity or characteristics such as user information, user behavior data, user historical data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of these data will comply with relevant laws, regulations, and standards. In addition, when the embodiment of the present application needs to obtain the user's sensitive personal information, the user's separate permission or consent will be obtained through a pop-up window or by jumping to a confirmation page. After clearly obtaining the user's separate permission or consent, the necessary user-related data for the normal operation of the embodiment of the present application will be obtained.

[0062] Figure 1 This is an optional flowchart of the device offline fault diagnosis method provided in the embodiment of the present application; Figure 1 The method may include but is not limited to steps S100 to S300.

[0063] Step S100, obtaining first characteristic data of the device and second characteristic data of the background; wherein the first characteristic data is derived from the offline state data of the device;

[0064] Step S200, fusing the first feature data and the second feature data to obtain mixed feature data;

[0065] Step S300, inputting the mixed feature data into a fault prediction model to obtain a fault diagnosis result; the fault prediction model is constructed based on a decision tree.

[0066] Reference Figure 2As shown, the first characteristic data in the embodiment of the present application is derived from the manually filled data, and the second characteristic data is derived from the historical uploaded data in the background. The embodiment of the present application predicts the fault by manually filling in the data and the background historical data, thereby improving the accuracy of the prediction.

[0067] In some embodiments, reference Figure 3 As shown, in the method provided in the embodiment of the present application, the second feature data includes statistical feature data, model feature data and implicit feature data, and obtaining the second feature data in the background includes:

[0068] Step S110, obtaining historical data reported to the background;

[0069] Step S120, performing statistical analysis on the historical data to obtain statistical characteristic data;

[0070] Step S130, performing fitting processing on the historical data to obtain model feature data;

[0071] Step S140: extract features from the historical data through an encoder to obtain implicit feature data.

[0072] In some possible implementations, the second feature data in the embodiments of the present application can perform feature extraction on historical data in a variety of ways to improve the prediction accuracy of the model.

[0073] In some embodiments, the method provided in the embodiments of the present application performs statistical analysis on historical data to obtain statistical feature data, including:

[0074] Get time series data from historical data;

[0075] Determine the mean, standard deviation, skewness, and kurtosis of time series data;

[0076] Determine the statistical characteristics of the data based on the mean, standard deviation, skewness and kurtosis.

[0077] In some embodiments, the method provided in the embodiments of the present application performs fitting processing on historical data to obtain model feature data, including:

[0078] Perform polynomial fitting on historical data to obtain polynomial features;

[0079] The model characteristic data is determined based on the coefficients in the polynomial characteristics.

[0080] In some possible implementations, the background feature data of the device is extracted. The historical data uploaded by the device to the background is mostly time series data, such as CPU usage and memory usage, etc., and statistical features, model features and implicit features are extracted respectively, where the statistical features mainly include mean, standard deviation, kurtosis and skewness statistical information, etc.; the model features are mainly fitted using a univariate quartic equation model, and the parameters in the equation are obtained as model feature data; the implicit features are mainly extracted using an autoencoder neural network, and the error between the autoencoder input data and the decoded data is reduced, and the hidden layer coefficients are extracted as implicit features. Of course, the feature extraction of historical data can also be other ways, and this application does not make specific restrictions.

[0081] Depend on Figure 4 It can be seen that the diagnostic model / system proposed in this application has five major modules, namely, artificial feature data extraction module, background feature data extraction module, decision tree model construction module, database data storage module and decision result processing module.

[0082] The main function of the manual feature data extraction module is to construct the manual data format and collect the manual data; it mainly includes the current operation status of the device and the status of the card in the device. Because when the device fails, the remote background may not be able to monitor the device data, so it needs to be filled in manually. Its features are mainly shown in the following table 1:

[0083] Feature 1 Whether to continue to be offline after restart Feature 2 Is the antenna installed correctly? Feature 3 Is the device lighting correct? Feature 4 Is the device background card reading abnormal? Feature 5 Whether to trigger the network disconnection due to reaching the data limit Feature 6 Is the service suspended due to arrears? Feature 7 Can the device backend ping 180.106.148.146?

[0084] Table 1

[0085] The characteristic data listed above are obtained based on the experience of on-site diagnosis of the product, and a value of 0 indicates no, and a value of 1 indicates yes. Of course, Table 1 is an illustrative example and is not a specific limitation.

[0086] The main function of the background feature data extraction module is to build a model pre-training feature data set based on the data obtained by the background; since the data in the background database is the data reported by the device, some of the data is time series data and cannot be directly used as the input data of the decision tree model. At the same time, there is a lot of redundant information, so it is necessary to extract features from the time series data to obtain feature data as model input data. The feature extraction adopted in this article mainly includes statistical feature extraction, polynomial fitting model feature extraction and autoencoder feature extraction. The main feature extraction process is as follows:

[0087] 1) Statistical characteristics

[0088] Assume that there is one-dimensional time series data, which is expressed as follows:

[0089] X=(x1,x2,…,x n );

[0090] where x i Represents the data at the i-th moment, and performs statistical feature extraction on the one-dimensional time series data, mainly including mean feature, standard deviation feature, skewness and kurtosis feature. The specific calculation method is shown in the example of statistical feature extraction in Table 2 below:

[0091]

[0092] Table 2

[0093] The mean reflects the "average" level of the data; the standard deviation reflects the degree of dispersion of the data expressed in the same unit; the skewness mainly measures the asymmetry of the data distribution; and the kurtosis mainly measures the sharpness or flatness of the data distribution.

[0094] Reference Figure 5 As shown, the polynomial fitting model feature extraction mainly adopts the polynomial fitting method to fit the data, and the obtained polynomial coefficients are used as its features. This paper adopts cubic polynomial fitting, and its fitting formula is as follows:

[0095] y=w0+w1*x+w2*x 2 +w3*x 3 ;

[0096] Among them, w0, w1, w2, and w3 are polynomial coefficients, which are also the model feature data extracted in this paper.

[0097] For hidden features, refer to Figure 6 As shown in the figure, the autoencoder is mainly divided into two parts: encoder and decoder. The function of the encoding part is mainly to encode the input data to obtain implicit features, and then use the decoder to decode the implicit features to minimize the error between the decoded data and the original data. Its formal description is as follows:

[0098] Z=f1(W (1) *X+b (1) );

[0099] X′=f2(W (2) *Z+b (2) );

[0100] Its objective function:

[0101] J(W,b)=∑|XX′|;

[0102] The expression formula of Z is the formal description of the encoder, where X is the input data, W and b are the neural network parameters, Z is the encoded data, which is also the implicit feature data, X ′ To decode the data.

[0103] The main function of the decision result processing module is to process the results obtained from the decision tree model, notify relevant personnel to perform fault diagnosis, and finally compare them. If the diagnosis type is consistent with the decision tree diagnosis type, the database is directly saved, otherwise the diagnosis results of the relevant personnel are saved.

[0104] After extracting features from the data, fault prediction is performed using a fault prediction model.

[0105] In some embodiments, the method provided in the embodiments of the present application, for obtaining first characteristic data of a device, includes:

[0106] If the device fails, obtain manually filled status data;

[0107] Feature extraction is performed on the status data to obtain first feature data corresponding to the first feature; the first feature includes whether the device continues to be offline after restarting, whether the antenna is installed correctly, whether the device lights up reasonably, or whether the device background card reading is normal.

[0108] In some possible implementations, when a device fails, some status data of the device when it is offline is manually extracted, mainly by manually filling out forms to extract manual feature data. The main manually extracted features include: whether the device continues to be offline after restart, whether the antenna is installed correctly, whether the device lights up correctly, whether the device background card reading is abnormal, whether the device background can ping 180.106.148.146, and other feature data.

[0109] In some embodiments, in the method provided by the embodiments of the present application, the fault prediction model is determined by the following steps:

[0110] Get mixed feature samples;

[0111] Determine the information entropy and information gain ratio of the mixed feature samples;

[0112] According to the information gain ratio, the partition feature of the current node is determined, and the mixed feature samples are partitioned based on the partition feature to obtain a sample subset;

[0113] Based on the information gain ratio of each sample subset, each sample subset is divided until the termination condition is met to obtain a decision tree;

[0114] Based on the decision tree, a fault prediction model is constructed.

[0115] In some possible implementations, the manually extracted feature data and the feature data extracted from the background data are fused and spliced ​​to obtain mixed feature data; the above-mentioned feature extraction is performed on the historical data in the database, and its fault type data is obtained at the same time, mainly including hardware failures, platform failures, business failures and other types, as a training set, and then a decision tree model is constructed and trained to obtain a trained decision tree model (i.e., the fault prediction model in the embodiment of the present application).

[0116] After obtaining the trained fault prediction model, a mixed feature extraction (including manual feature extraction, statistical features, model features and implicit features) is performed on the currently faulty equipment, mainly using the above-mentioned data processing method; the above-mentioned trained decision tree model is used to discriminate the current mixed feature data, obtain the fault diagnosis type, conduct troubleshooting, obtain the final true fault diagnosis data, store the current faulty equipment feature data and the true fault diagnosis data in the database, and at the same time train the decision tree model again to update the decision tree model parameters.

[0117] Figure 4 The main function of the decision tree model building module in is to build a decision tree model based on the training data set, classify the subsequent fault data, and obtain the type diagnosis result; therefore, the decision tree model and related definitions and concepts are introduced as follows:

[0118] Decision tree is a basic classification method. The decision tree model has a tree structure. In classification problems, it represents the process of classifying instances based on features. It can be considered as a collection of if-then rules. Its advantages are that the model is readable and the classification speed is fast. When learning, the decision tree model is established based on the principle of minimizing the loss function using the training data. When predicting, the new data is classified using the decision tree model. Decision tree learning usually includes three steps: feature selection, decision tree generation, and decision tree pruning. The details are as follows:

[0119] 1) Decision tree model.

[0120] About decision tree: The classification decision tree model is a tree structure that describes the classification of instances. Figure 7 As shown in Figure 2. A decision tree consists of nodes and directed edges. There are two types of nodes: internal nodes and leaf nodes. Internal nodes represent a feature or attribute, and leaf nodes represent a type.

[0121] Using decision tree classification, we start from the root node, test a feature of the instance, and assign the instance to its child nodes based on the test results; at this time, each child node corresponds to a value of the feature. We recursively test and assign instances until we reach a leaf node. Finally, we assign the instance to the class of the leaf node.

[0122] 2) Decision trees and if-then rules.

[0123] A decision tree can be viewed as a set of if-then rules. The process of converting a decision tree into an if-then rule is as follows: each path from the root node to the leaf node of the decision tree is constructed into a rule; the characteristics of the internal nodes on the path correspond to the conditions of the rule, and the leaf nodes correspond to the conclusions of the rule. The path of a decision tree or other corresponding if-then rule set has an important property: mutual exclusion and completeness. That is, each instance is covered by one path or one rule, and only by one path or one rule.

[0124] 3) Decision Tree Learning

[0125] Assume that the given training data set is as follows

[0126] D={(x1,y1),(x2,y2),…,(x n ,y n )};

[0127] in, is the input instance, N is the number of features, y i ∈{1,2,…K} is the class label, i=1,2,…n, n is the sample size. The purpose of decision tree learning is to build a decision tree model based on a given training data set so that it can correctly classify instances.

[0128] Regarding information gain, the information gain g(D,A) of feature A on training set D is defined as the difference between the empirical entropy H(D) of set D and the empirical conditional entropy H(D|A) of D under given conditions of feature A, that is:

[0129] g(D,A)=H(D)-H(D|A);

[0130] Let the training data set be D, |D| represents the sample capacity, that is, the number of samples. Suppose there are K classes C k ,k=1,2,…,K,|C k | belongs to class C k The number of samples, Assume that feature A has n different values ​​{a1,a2,…,a n}, according to the value of feature A, D is divided into n subsets {D1, D2, ..., D n},|D i | for D i The number of sets, Note subset D i Belongs to Class C k The sample set is D ik, that is, D ik =D i ∩C k ,|D ik | for D ik The number of samples is, then the information gain algorithm is as follows:

[0131] Input: training data set D and feature A;

[0132] Output: information gain g(D,A) of feature A for training data set D;

[0133] (1) Calculate the empirical entropy H(D) of the data set D:

[0134]

[0135] (2) Calculate the empirical conditional entropy H(D|A) of feature A for data set D:

[0136]

[0137] (3) Calculate information gain:

[0138] g(D,A)=H(D)-H(D|A);

[0139] Information gain ratio: Using information gain as a feature to divide the training set has the problem of biased selection of features with more values. Using information gain ratio can improve this problem.

[0140] Information gain ratio: The information gain ratio g of feature A to training data set D r (D, A) is defined as the information gain g(D, A) and the entropy H of the training dataset D with respect to the value of feature A. A (D), that is:

[0141]

[0142] in n is the number of values ​​that feature A can take.

[0143] Decision tree algorithm, based on the previous information gain definition and decision tree introduction, we can get the decision tree generation and learning pseudo-algorithm as follows:

[0144] Input: training data set D = {(x1,y1),(x2,y2),…,(x n ,y n )}, attribute set A={a1,a2,…,a d};

[0145]

[0146] Based on the above process, the decision tree is determined through the information gain ratio, and then the fault prediction model is determined.

[0147] In some embodiments, the method provided in the embodiments of the present application, for obtaining first characteristic data of a device, includes:

[0148] Design a questionnaire form; the questionnaire form includes a blank corresponding to the first feature to be filled in;

[0149] Receive the completed receipt information of the questionnaire form;

[0150] According to the filled receipt information, the first characteristic data is obtained;

[0151] Alternatively, a device status system is designed; the device status system includes a control corresponding to the first feature to be fed back;

[0152] Receive feedback from the device status system;

[0153] According to the feedback information, first characteristic data is obtained.

[0154] The manually filled data in the embodiment of the present application can be obtained through an offline form or through an online system. The feedback information can be typed information or uploaded image signals / audio information.

[0155] The following is a detailed description and explanation of the solution of the embodiment of the present invention in conjunction with a specific application example:

[0156] As above Figure 2 As shown, the main steps of the offline diagnosis technology based on the decision tree algorithm invented in this paper are as follows:

[0157] Step 1: Extract mixed feature data through historical data, which mainly includes manually extracted features and background data features (statistical features, model features, implicit features); the extraction method has been mentioned above and will not be repeated here. The results are directly given as an example of training data shown in Table 3 below.

[0158]

[0159]

[0160] Table 3

[0161] Each column of data in Table 3 represents the model input data collected when a single device fails, mainly including the input feature data collected manually and the feature data obtained by the background at that time. The specific information is shown in the table; the diagnosis result is the diagnosis type determined after the fault is troubleshooted at that time. There are four main diagnosis types:

[0162] 0: No personnel diagnosis is required, and it is a normal operating state;

[0163] 1: Operation personnel are required to conduct investigation;

[0164] 2: Hardware engineers are required to conduct troubleshooting;

[0165] 3: Backend engineers are required to investigate.

[0166] Step 2: Use the decision tree algorithm to build a decision tree using the pre-trained data (mixed feature data) shown in step 1. The information gain is mainly used to compare the attribute features to determine whether the diagnosis type can be identified. The specific calculation details are not shown here (the calculation method uses the decision tree algorithm, formulas 6, 7, and 8, refer to the decision tree model construction module), and the results are directly displayed, such as Figure 8 As shown, the red boxes represent the internal nodes of the decision tree, and the green leaf nodes represent the diagnosis types, where X[t] represents the t+1th attribute, gini is the information gain ratio, sample represents the number of samples, and value represents the number of diagnosis type results distributed.

[0167] Step 3: According to the trained decision tree and the mixed feature data of the current faulty device, the discrimination result is obtained, as shown in the test data example in Table 4:

[0168]

[0169]

[0170] Table 4

[0171] According to the current input data and the diagnostic decision tree constructed in step 2, the diagnosis type can be obtained as normal (0) according to the if-else-then logic of the decision tree.

[0172] Based on the above description, it can be seen that in the artificial feature data extraction of the embodiment of the present application: how to extract some important feature data of the current device as model input, it is necessary for experienced professionals to design questionnaire forms and extract key feature data. In the background data feature extraction: how to construct feature data from the historical reported data of the device, how to construct statistical feature data, how to construct model feature data, how to extract implicit feature data from historical data, and how to perform mixed feature splicing.

[0173] The existing processing method mainly adopts the pipeline method, starting with the operation personnel, then to the product, software developers, and finally to the hardware developers. Compared with the current way of handling equipment failures, the method proposed in this article is more effective in improving the timeliness of equipment failure repair; compared with existing similar technologies, the traditional method directly uses the original data, and there is a lot of redundant information in the original data, which interferes with the judgment results and causes the performance of the model to decline; and this model first extracts the characteristic data of the current environment of the equipment through a questionnaire designed by professionals, and then extracts the statistical characteristics, model characteristics and implicit characteristics of the time series data through the analysis of the background historical data, and integrates these characteristic data for model training, so that the data features processed by the model are richer and the interference information is less, and the diagnosis of fault types is more extensive and the diagnosis efficiency is higher.

[0174] In summary, this paper proposes an offline diagnosis technology for commercial enterprise private network equipment based on hybrid feature extraction and decision tree algorithm. Compared with the existing processing methods, it is more efficient to troubleshoot faults and repair equipment, and can quickly determine the fault diagnosis type.

[0175] See also Fig. 9 The embodiment of the present application also provides a device offline fault diagnosis system, which can implement the above-mentioned device offline fault diagnosis method, and the system includes:

[0176] The first module 810 is used to obtain first characteristic data of the device and second characteristic data of the background; wherein the first characteristic data is derived from the status data of the device being offline;

[0177] The second module 820 is used to fuse the first feature data and the second feature data to obtain mixed feature data;

[0178] The third module 830 is used to input the mixed feature data into a fault prediction model to obtain a fault diagnosis result; the fault prediction model is constructed based on a decision tree.

[0179] It can be understood that the contents of the above method embodiments are all applicable to the present system embodiments, the functions specifically implemented by the present system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0180] The embodiment of the present application also provides an electronic device, the electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the above-mentioned device offline fault diagnosis method when executing the computer program. The electronic device can be any intelligent terminal including a tablet computer, a car computer, etc.

[0181] It can be understood that the contents of the above method embodiments are all applicable to the present device embodiments, the functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0182] See also Fig.10 , Fig.10 The hardware structure of an electronic device of another embodiment is illustrated, and the electronic device includes:

[0183] The processor 901 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application;

[0184] The memory 902 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 902 can store an operating system and other application programs. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 902, and the processor 901 calls and executes the device offline fault diagnosis method of the embodiment of this application;

[0185] Input / output interface 903, used to implement information input and output;

[0186] Communication interface 904, used to realize communication interaction between the device and other devices, which can be realized by wired mode (such as USB, network cable, etc.) or wireless mode (such as mobile network, WIFI, Bluetooth, etc.);

[0187] A bus 905 that transmits information between various components of the device (e.g., the processor 901, the memory 902, the input / output interface 903, and the communication interface 904);

[0188] The processor 901 , the memory 902 , the input / output interface 903 and the communication interface 904 are connected to each other in communication within the device via a bus 905 .

[0189] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above-mentioned device offline fault diagnosis method is implemented.

[0190] It can be understood that the contents of the above method embodiments are all applicable to the present storage medium embodiments, the functions specifically implemented by the present storage medium embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0191] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0192] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0193] Those skilled in the art will appreciate that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0194] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0195] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0196] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0197] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0198] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0199] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0200] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0201] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store programs.

[0202] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the rights of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present invention should be within the scope of the rights of the present invention.

Claims

1. A device offline fault diagnosis method, characterized in that: The method comprises: Acquire first characteristic data of the device and second characteristic data of the background; wherein the first characteristic data is derived from the status data of the device being offline; Fusing the first feature data and the second feature data to obtain mixed feature data; The mixed feature data is input into a fault prediction model to obtain a fault diagnosis result; the fault prediction model is constructed based on a decision tree.

2. The method according to claim 1, characterized in that The second feature data includes statistical feature data, model feature data and implicit feature data. Acquiring the second feature data in the background includes: Obtaining historical data reported to the backend; Performing statistical analysis on the historical data to obtain the statistical characteristic data; Performing fitting processing on the historical data to obtain the model characteristic data; The historical data is subjected to feature extraction by an encoder to obtain the implicit feature data.

3. The method according to claim 2, characterized in that The performing statistical analysis on the historical data to obtain the statistical characteristic data includes: Obtaining time series data from the historical data; Determining the mean, standard deviation, skewness, and kurtosis of the time series data; The statistical characteristic data is determined according to the mean, the standard deviation, the skewness and the kurtosis.

4. The method according to claim 2, characterized in that: The fitting process is performed on the historical data to obtain the model characteristic data, including: Performing polynomial fitting processing on the historical data to obtain polynomial features; The model feature data is determined according to the coefficients in the polynomial feature.

5. The method according to claim 1, characterized in that: Acquiring first characteristic data of the device includes: If the device fails, obtain manually filled status data; Feature extraction is performed on the status data to obtain first feature data corresponding to the first feature; the first feature includes whether the device continues to be offline after restarting, whether the antenna is installed correctly, whether the device lights up reasonably, or whether the device background card reading is normal.

6. The method according to claim 1, characterized in that The fault prediction model is determined by the following steps: Get mixed feature samples; Determining the information entropy and information gain ratio of the mixed feature sample; Determining the partition feature of the current node according to the information gain ratio, and partitioning the mixed feature sample based on the partition feature to obtain a sample subset; Based on the information gain ratio of each sample subset, each sample subset is divided until a termination condition is met to obtain a decision tree; The fault prediction model is constructed according to the decision tree.

7. The method according to claim 1, characterized in that Acquiring first characteristic data of the device includes: Design a questionnaire form; the questionnaire form includes a blank corresponding to the first feature to be filled in; Receiving the completed receipt information of the questionnaire form; Obtaining the first characteristic data according to the filling receipt information; Alternatively, a device status system is designed; the device status system includes a control corresponding to the first feature to be fed back; Receiving feedback information from the device status system; The first characteristic data is obtained according to the feedback information.

8. An equipment offline fault diagnosis system, characterized in that: The system comprises: The first module is used to obtain the first characteristic data of the device and the second characteristic data of the background; wherein the first characteristic data is derived from the status data of the device being offline; A second module is used to fuse the first feature data and the second feature data to obtain mixed feature data; The third module is used to input the mixed feature data into a fault prediction model to obtain a fault diagnosis result; the fault prediction model is constructed based on a decision tree.

9. An electronic device, characterized in that: The electronic device comprises a memory and a processor, the memory stores a computer program, and the processor implements the method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images