CN115470854A - Information system fault classification method and classification system - Google Patents
Information system fault classification method and classification system Download PDFInfo
- Publication number
- CN115470854A CN115470854A CN202211121417.4A CN202211121417A CN115470854A CN 115470854 A CN115470854 A CN 115470854A CN 202211121417 A CN202211121417 A CN 202211121417A CN 115470854 A CN115470854 A CN 115470854A
- Authority
- CN
- China
- Prior art keywords
- information
- data set
- hardware
- database
- operating system
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000013135 deep learning Methods 0.000 claims abstract description 27
- 238000012549 training Methods 0.000 claims abstract description 27
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 26
- 238000012360 testing method Methods 0.000 claims abstract description 21
- 238000012795 verification Methods 0.000 claims abstract description 15
- 230000002159 abnormal effect Effects 0.000 claims description 40
- 238000004422 calculation algorithm Methods 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 3
- 238000013145 classification model Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 9
- 238000001514 detection method Methods 0.000 description 8
- 230000006872 improvement Effects 0.000 description 8
- 238000012545 processing Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 239000011159 matrix material Substances 0.000 description 7
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000004891 communication Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
The invention discloses an information system fault classification method and a classification system, wherein the classification method comprises the following steps: acquiring a system history prompt information data set; obtaining a plurality of subdata sets from the system historical prompt information dataset; performing model training, verification and testing on the plurality of sub data sets by using a BERT deep learning network to correspondingly obtain a plurality of trained BERT convolutional neural network models; obtaining a cascade network by utilizing the plurality of models; and classifying the fault by utilizing the cascade network. The information system fault classification method and the information system fault classification system achieve the effect of classification and are high in classification accuracy and classification efficiency.
Description
Technical Field
The invention relates to the field of fault classification, and particularly provides a fault classification method and a fault classification system for an information system.
Background
The safe and stable operation of the information communication system is always an important guarantee of a power grid service system and is concerned highly. The dispatching command of the information communication system is one of core technologies of communication, communication and maintenance, and aims to realize rapid fault diagnosis, positioning and maintenance scheme optimization selection and directly influence the maintenance time of faults and the stable operation of a power grid service system.
Artificial intelligence plays an important role in data anomaly detection, can significantly improve detection speed and accuracy, and is applied in numerous fields, for example: power data. Machine learning is an important branch of artificial intelligence, deep learning is a main algorithm in machine learning, and the aim of the method is to extract high-level abstract features of data and learn potential distribution rules of the data through multi-layer nonlinear transformation, so that the capability of making reasonable judgment or prediction on new data is obtained. Deep learning relies on strong fitting ability, and starts to be a method which can also rely on deep learning in various fields, particularly in the field of data anomaly detection.
Some traditional machine learning methods can realize the intellectualization of power data anomaly detection, and the anomaly detection is roughly divided into the following four types: based on statistics, density, clustering and proximity, abnormal data of the power equipment has no definite judgment standard, and a single clustering method such as LOF, K-means and KNN is generally adopted for abnormal detection, but the problems of long calculation time, low efficiency, low precision and the like caused by high characteristic dimensionality of the power equipment cannot meet the requirement of rapid abnormal detection of the power data.
Disclosure of Invention
In view of the above, the present invention provides a method and a system for classifying a fault in an information system, so as to solve the problems of low accuracy and low classification efficiency of the conventional anomaly detection.
The invention provides a fault classification method for an information system on one hand, which comprises the following steps:
acquiring a system history prompt information data set;
obtaining a plurality of subdata sets from the system history prompt information dataset;
sequentially inputting the plurality of subdata sets into a BERT deep learning network for model training, verification and testing to correspondingly obtain a plurality of trained BERT convolutional neural network models;
building a cascade network by using a plurality of trained convolutional neural network models;
and classifying the input system prompt information data by utilizing the cascade network to obtain a classification result.
Preferably, the number of the sub data sets is 5, where the sub data set 1 is a system prompt information data set, and prompt information in the data set is divided into two types according to software information and hardware information, the sub data set 2 is a software information data set, and prompt information in the data set is divided into two types according to operating system information and database information, the sub data set 3 is a hardware information data set, and prompt information in the data set is divided into two types according to hardware exception information and hardware normal information, the sub data set 4 is an operating system information data set, and prompt information in the data set is divided into two types according to operating system exception information and operating system normal information, the sub data set 5 is a database information data set, and prompt information in the data set is divided into two types according to database exception information and database normal information.
Preferably, the cascade network is implemented in three stages, a first stage is used for predicting whether system prompt information is software information or hardware information, a second stage is used for predicting whether software information is operating system information or database information and whether hardware information is hardware normal information or hardware abnormal information, and a third stage is used for predicting whether operating system information is operating system abnormal information and operating system normal information and whether database information is database abnormal information or database normal information.
Further preferably, before the plurality of subdata sets are sequentially input into the BERT deep learning network for model training, verification and testing, the method further comprises the step of constructing the BERT deep learning network by overlapping the Transformer Encoder.
Further preferably, the parameters of the plurality of sub data sets for model training are as follows: the epoch is 30, the batch size is 128, the learning rate is 0.00005, and the optimizer is the AdamW algorithm.
The invention also provides an information system fault classification system, which comprises:
a history prompt information data set acquisition unit: the system comprises a data acquisition module, a data processing module and a data processing module, wherein the data acquisition module is used for acquiring a system historical prompt information data set;
a sub data set acquisition unit: for obtaining a plurality of subdata sets from the system history prompt information dataset;
a convolutional neural network model acquisition unit: the model test system is used for sequentially inputting the plurality of sub data sets into a BERT deep learning network for model training, verification and testing, and correspondingly obtaining a plurality of trained BERT convolutional neural network models;
a cascade network building unit: the method is used for building a cascade network by utilizing a plurality of trained convolutional neural network models;
a fault classification unit: and the system prompt information data classification device is used for classifying the input system prompt information data by utilizing the cascade network to obtain a classification result.
Preferably, the sub data set obtaining unit is configured to obtain 5 sub data sets from the system history prompt information data set, where the sub data set 1 is a system prompt information data set, prompt information in the data set is divided into two categories according to software information and hardware information, the sub data set 2 is a software information data set, prompt information in the data set is divided into two categories according to operating system information and database information, the sub data set 3 is a hardware information data set, prompt information in the data set is divided into two categories according to hardware anomaly information and hardware normal information, the sub data set 4 is an operating system information data set, prompt information in the data set is divided into two categories according to operating system anomaly information and operating system normal information, the sub data set 5 is a database information data set, and prompt information in the data set is divided into two categories according to database anomaly information and database normal information.
Preferably, the cascade network building unit is used for building a three-level cascade network by using the trained 5 convolutional neural network models, wherein the first level is used for predicting whether the system prompt information is software information or hardware information, the second level is used for predicting whether the software information is operating system information or database information and whether the hardware information is hardware normal information or hardware abnormal information, and the third level is used for predicting whether the operating system information is operating system abnormal information and operating system normal information and whether the database information is database abnormal information or database normal information.
Further preferably, the information system fault classification system further includes a BERT deep learning network construction unit, wherein the BERT deep learning network is constructed by superimposing a Transformer Encoder.
Further preferably, the parameters of the convolutional neural network model obtaining unit for performing model training on the sub-data set are as follows: the epoch is 30, the batch size is 128, the learning rate is 0.00005, and the optimizer is the AdamW algorithm.
The information system fault classification method and the classification system provided by the invention have the advantages that the cascade convolution neural network is added on the basis of the traditional deep learning network, and the effect of classification is successfully realized. The problem that the data sets cannot be accurately classified at one time due to the fact that the data sets are various in types and similar in characteristics is solved through the cascade network. The cascade network reduces the labels of the data set, improves the classification accuracy, reduces the operation work of frequently screening possible results and the text information of the current manual plan to a greater extent, greatly improves the accuracy of prediction classification, and provides a quick, accurate and reliable decision information basis for improving the classification efficiency.
According to the information system fault classification method and the information system fault classification system, the BERT deep learning network is used, various prompt messages obtained when the information system runs are input into the BERT network, a plurality of classification models can be obtained, and the information system fault types can be classified through the cascade network established by the classification models.
Drawings
The invention is described in further detail below with reference to the following figures and embodiments:
FIG. 1 is a flow chart of a method for classifying faults in an information system according to the present invention;
FIG. 2 is a diagram of a cascaded network architecture constructed in accordance with the present invention;
FIG. 3 is a comparison graph of a conventional multi-class model and 5 BERT convolutional neural network models trained by the present invention;
FIG. 4 is a confusion matrix formed by a conventional 6-class model and its classification accuracy;
fig. 5 is a confusion matrix formed by the cascade network and the classification accuracy thereof provided by the present invention.
Detailed Description
The invention will be further explained with reference to specific embodiments, without limiting the invention.
As shown in fig. 1, the present invention provides a method for classifying a fault in an information system, comprising the following steps:
acquiring a system history prompt information data set;
obtaining a plurality of subdata sets from the system history prompt information dataset;
sequentially inputting the plurality of subdata sets into a BERT deep learning network for model training, verification and testing to correspondingly obtain a plurality of trained BERT convolutional neural network models;
building a cascade network by using a plurality of trained convolutional neural network models;
and classifying the input system prompt information data by utilizing the cascade network to obtain a classification result.
As an improvement of the technical solution, the number of the sub data sets is 5, wherein the sub data set 1 is a system prompt information data set, prompt information in the data set is divided into two types according to software information and hardware information, the sub data set 2 is a software information data set, prompt information in the data set is divided into two types according to operating system information and database information, the sub data set 3 is a hardware information data set, prompt information in the data set is divided into two types according to hardware abnormal information and hardware normal information, the sub data set 4 is an operating system information data set, prompt information in the data set is divided into two types according to operating system abnormal information and operating system normal information, the sub data set 5 is a database information data set, and prompt information in the data set is divided into two types according to database abnormal information and database normal information.
As an improvement of the technical scheme, the cascade network is in three stages, the first stage is used for predicting whether system prompt information is software information or hardware information, the second stage is used for predicting whether software information is operating system information or database information and whether hardware information is hardware normal information or hardware abnormal information, and the third stage is used for predicting whether operating system information is operating system abnormal information and operating system normal information and whether database information is database abnormal information or database normal information.
As an improvement of the technical scheme, before the plurality of subdata sets are sequentially input into the BERT deep learning network for model training, verification and testing, the method further comprises the step of constructing the BERT deep learning network by overlapping a Transformer Encoder.
As an improvement of the technical solution, the parameters of the multiple subdata sets for model training are as follows: epoch is 30, batch size is 128, learning rate is 0.00005, and optimizer AdamW algorithm.
The invention also provides an information system fault classification system, which comprises:
a history prompt information data set acquisition unit: the system comprises a data acquisition module, a data processing module and a data processing module, wherein the data acquisition module is used for acquiring a system historical prompt information data set;
a sub data set acquisition unit: for obtaining a plurality of subdata sets from the system history prompt information dataset;
a convolutional neural network model acquisition unit: the model training, verifying and testing are carried out on the plurality of sub data sets sequentially input into a BERT deep learning network, and a plurality of trained BERT convolution neural network models are correspondingly obtained;
a cascade network building unit: the method comprises the steps of establishing a cascade network by utilizing a plurality of trained convolutional neural network models;
a fault classification unit: and the system prompt information data classification device is used for classifying the input system prompt information data by utilizing the cascade network to obtain a classification result.
As an improvement of the technical solution, the sub data set obtaining unit is configured to obtain 5 sub data sets from the system history prompt information data set, where the sub data set 1 is a system prompt information data set, the prompt information in the data set is divided into two types according to software information and hardware information, the sub data set 2 is a software information data set, the prompt information in the data set is divided into two types according to operating system information and database information, the sub data set 3 is a hardware information data set, the prompt information in the data set is divided into two types according to hardware exception information and hardware normal information, the sub data set 4 is an operating system information data set, the prompt information in the data set is divided into two types according to operating system exception information and operating system normal information, the sub data set 5 is a database information data set, and the prompt information in the data set is divided into two types according to database exception information and database normal information.
As an improvement of the technical scheme, the cascade network building unit is used for building a three-level cascade network by using trained 5 convolutional neural network models, wherein the first level is used for predicting whether system prompt information is software information or hardware information, the second level is used for predicting whether software information is operating system information or database information and whether hardware information is hardware normal information or hardware abnormal information, and the third level is used for predicting whether operating system information is operating system abnormal information and operating system normal information and whether database information is database abnormal information or database normal information.
As an improvement of the technical scheme, the information system fault classification system further comprises a BERT deep learning network construction unit, wherein the BERT deep learning network is constructed by superposing a Transformer Encoder.
As an improvement of the technical solution, the parameters of the convolutional neural network model acquisition unit for performing model training on the sub-data set are as follows: the epoch is 30, the batch size is 128, the learning rate is 0.00005, and the optimizer is the AdamW algorithm.
The information system fault classification method and the classification system add the cascade convolution neural network on the basis of the traditional deep learning network, and successfully realize the effect of classification. The problem that the data sets cannot be accurately classified at one time due to the fact that the data sets are various in types and similar in characteristics is solved through the cascade network. The cascade network reduces labels of a data set, improves the classification accuracy, greatly reduces the operation work of frequently screening possible results and predicting classification accuracy of the existing manual plan text information, and provides a quick, accurate and reliable decision information basis for improving the classification efficiency.
Example (b):
the method provided by the invention is utilized to construct the cascade network and verify the accuracy of the cascade network:
s1: acquiring a system historical prompt information data set, wherein the types of prompt information in the data set are 6, and the types are respectively as follows: label values of operating system abnormal information, operating system normal information, database abnormal information, database normal information, hardware abnormal information, hardware normal information and 6-type prompt information are 0,1,2,3,4 and 5 in sequence, the data set comprises 120000 pieces of data information of a training set, a verification set and a test set, wherein the training set comprises 108000 pieces of data information, and the verification set and the test set comprise 6000 pieces of data information;
s2: 5 subdata sets are obtained from the system historical prompt information data set, each subdata set comprises categories, training, verification and testing 4 txt files, and the method specifically comprises the following steps:
txt: the category name file is used for recording the category name corresponding to the label;
txt: a training set file containing training set data;
txt: a verification set file containing verification set data;
text. Txt: a test set file containing test set data;
wherein each subdata set has two tags, specifically: the subdata set 1 is a system prompt information data set, the prompt information in the data set is divided into two types according to software information and hardware information, the subdata set 2 is a software information data set, the prompt information in the data set is divided into two types according to operating system information and database information, the subdata set 3 is a hardware information data set, the prompt information in the data set is divided into two types according to hardware abnormal information and hardware normal information, the subdata set 4 is an operating system information data set, the prompt information in the data set is divided into two types according to operating system abnormal information and operating system normal information, the subdata set 5 is a database information data set, and the prompt information in the data set is divided into two types according to database abnormal information and database normal information;
s3: sequentially inputting the subdata sets 1 to 5 into a BERT deep learning network for Model training, verification and testing, and correspondingly obtaining 5 trained BERT convolutional neural network models which are respectively marked as a Model 1, a Model 2, a Model 3, a Model 4 and a Model 5, wherein the Model 1 is used for classifying input system prompt information by software information and hardware information, the Model 2 is used for classifying the input software information by an operating system and database information, the Model 3 is used for classifying the input hardware information by hardware abnormal information and hardware normal information, the Model 4 is used for classifying the input operating system information by the operating system abnormal information and the operating system normal information, and the Model 5 is used for classifying the input database information by the database abnormal information and the database normal information;
the invention uses a BERT deep learning network, uses different data sets for 5 times of training, but has the same training parameters, sets the epoch to be 30, the batch size to be 128, the learning rate to be 0.00005, and the optimizer is AdamW algorithm. After training is finished, 5 different BERT convolutional neural network models (Model 1, model 2, model 3, model 4 and Model 5) are finally obtained, and all the models are two classification models (also called two classifiers);
FIG. 3 is precision data of 5 trained BERT convolutional neural network models of the present invention; the training accuracy, the type of the classifier, the classification output result and the data set used for training are shown for 5 trained classification models (Model 1, model 2, model 3, model 4, model 5) and 1 traditional classification Model 6. It can be seen that the accuracy of the 5 binary classification models is slightly higher than the six classification models (also called multi-classifiers) used in the conventional classification network. However, the performance and accuracy of the cascade network constructed by the 5 binary classification models are not high enough to prove compared with those of the traditional six classification models, and in order to prove the performance and accuracy, a confusion matrix is calculated later, and various index data are calculated according to data obtained by the confusion matrix;
s4: building a cascade network by using the trained 5 convolutional neural network models, wherein the cascade network is three levels, as shown in fig. 2, the first level predicts whether system prompt information is software information or hardware information by using Model 1, the second level predicts whether the software information is operating system information or database information by using Model 2, predicts whether the hardware information is hardware normal information or hardware abnormal information by using Model 3, the third level predicts whether the operating system information is operating system abnormal information and operating system normal information by using Model 4, and predicts whether the database information is database abnormal information or database normal information by using Model 5;
the cascade network can divide the system prompt information into normal and abnormal information of different types;
the accuracy of the cascade network and the accuracy of the traditional classification network provided by the invention are compared through the test results as follows:
testing the accuracy of the cascade network by using a test sample, calculating the accuracy of each type, and drawing a confusion matrix (as shown in figure 5);
calculating the accuracy of the traditional classification network and drawing a confusion matrix (as shown in figure 4);
comparing the two, it can be seen that the accuracy of the cascade network provided by the invention is higher. Wherein each column of the confusion matrix represents a prediction category, and the total number of each column represents the number of data predicted for that category; each row represents a true attribution category of data, and the total number of data in each row represents the number of data instances for that category. The value in each column represents the number of classes for which real data is predicted.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein. The scheme in the embodiment of the invention can be realized by adopting various computer languages, such as object-oriented programming language Java and transliteration scripting language JavaScript.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including the preferred embodiment and all changes and modifications that fall within the scope of the invention.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (10)
1. The information system fault classification method is characterized by comprising the following steps:
acquiring a system history prompt information data set;
obtaining a plurality of subdata sets from the system history prompt information dataset;
sequentially inputting the plurality of subdata sets into a BERT deep learning network for model training, verification and testing to correspondingly obtain a plurality of trained BERT convolutional neural network models;
building a cascade network by using a plurality of trained convolutional neural network models;
and classifying the input system prompt information data by utilizing the cascade network to obtain a classification result.
2. The information system fault classification method according to claim 1, characterized in that: the number of the subdata sets is 5, wherein the subdata set 1 is a system prompt information data set, prompt information in the data set is divided into two types according to software information and hardware information, the subdata set 2 is a software information data set, the prompt information in the data set is divided into two types according to operating system information and database information, the subdata set 3 is a hardware information data set, the prompt information in the data set is divided into two types according to hardware exception information and hardware normal information, the subdata set 4 is an operating system information data set, the prompt information in the data set is divided into two types according to operating system exception information and operating system normal information, the subdata set 5 is a database information data set, and the prompt information in the data set is divided into two types according to database exception information and database normal information.
3. The information system fault classification method according to claim 2, characterized by: the cascade network is three stages, wherein the first stage is used for predicting whether the prompt information of the system is software information or hardware information, the second stage is used for predicting whether the software information is operating system information or database information and whether the hardware information is hardware normal information or hardware abnormal information, and the third stage is used for predicting whether the operating system information is operating system abnormal information and operating system normal information and whether the database information is database abnormal information or database normal information.
4. The information system fault classification method according to claim 1, characterized by: and before sequentially inputting the plurality of subdata sets into the BERT deep learning network for model training, verification and testing, the method also comprises the step of constructing the BERT deep learning network by overlapping the Transformer Encoder.
5. The information system fault classification method according to claim 1, characterized in that: the parameters of the multiple subdata sets for model training are as follows: the epoch is 30, the batch size is 128, the learning rate is 0.00005, and the optimizer is the AdamW algorithm.
6. An information system fault classification system, comprising:
a history prompt information data set acquisition unit: the system is used for acquiring a system history prompt information data set;
a sub data set acquisition unit: for obtaining a plurality of subdata sets from the system history prompt information dataset;
a convolutional neural network model acquisition unit: the model test system is used for sequentially inputting the plurality of sub data sets into a BERT deep learning network for model training, verification and testing, and correspondingly obtaining a plurality of trained BERT convolutional neural network models;
a cascade network building unit: the method comprises the steps of establishing a cascade network by utilizing a plurality of trained convolutional neural network models;
a fault classification unit: and the system prompt information data classification device is used for classifying the input system prompt information data by utilizing the cascade network to obtain a classification result.
7. The information system fault classification system according to claim 6, characterized by: the subdata set acquisition unit is used for acquiring 5 subdata sets from the system history prompt information data set, wherein the subdata set 1 is a system prompt information data set, prompt information in the data set is divided into two types according to software information and hardware information, the subdata set 2 is a software information data set, prompt information in the data set is divided into two types according to operating system information and database information, the subdata set 3 is a hardware information data set, the prompt information in the data set is divided into two types according to hardware abnormal information and hardware normal information, the subdata set 4 is an operating system information data set, the prompt information in the data set is divided into two types according to operating system abnormal information and operating system normal information, the subdata set 5 is a database information data set, and the prompt information in the data set is divided into two types according to database abnormal information and database normal information.
8. The information system fault classification system of claim 7, wherein: the cascade network building unit is used for building a three-level cascade network by utilizing the trained 5 convolutional neural network models, wherein the first level is used for predicting whether system prompt information is software information or hardware information, the second level is used for predicting whether software information is operating system information or database information and whether hardware information is hardware normal information or hardware abnormal information, and the third level is used for predicting whether operating system information is operating system abnormal information and operating system normal information and whether database information is database abnormal information or database normal information.
9. The information system fault classification system according to claim 6, characterized by: the device further comprises a BERT deep learning network construction unit, wherein the BERT deep learning network is constructed by superposing a Transformer Encoder.
10. The information system fault classification system according to claim 6, characterized by: the parameters of the convolutional neural network model acquisition unit for performing model training on the sub-data set are as follows: the epoch is 30, the batch size is 128, the learning rate is 0.00005, and the optimizer is the AdamW algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211121417.4A CN115470854A (en) | 2022-09-15 | 2022-09-15 | Information system fault classification method and classification system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211121417.4A CN115470854A (en) | 2022-09-15 | 2022-09-15 | Information system fault classification method and classification system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115470854A true CN115470854A (en) | 2022-12-13 |
Family
ID=84333228
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211121417.4A Pending CN115470854A (en) | 2022-09-15 | 2022-09-15 | Information system fault classification method and classification system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115470854A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116450399A (en) * | 2023-06-13 | 2023-07-18 | 西华大学 | Fault diagnosis and root cause positioning method for micro service system |
-
2022
- 2022-09-15 CN CN202211121417.4A patent/CN115470854A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116450399A (en) * | 2023-06-13 | 2023-07-18 | 西华大学 | Fault diagnosis and root cause positioning method for micro service system |
| CN116450399B (en) * | 2023-06-13 | 2023-08-22 | 西华大学 | Fault diagnosis and root cause positioning method for micro service system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111626336B (en) | Subway fault data classification method based on unbalanced data set | |
| US11475326B2 (en) | Analyzing test result failures using artificial intelligence models | |
| CN111427775B (en) | Method level defect positioning method based on Bert model | |
| CN111309607A (en) | Software defect positioning method of code method level | |
| CN110968703B (en) | Method and system for constructing abnormal metering point knowledge base based on LSTM end-to-end extraction algorithm | |
| CN105471647A (en) | Power communication network fault positioning method | |
| CN110704616A (en) | Equipment alarm work order identification method and device | |
| CN115373879A (en) | Intelligent operation and maintenance disk fault prediction method for large-scale cloud data center | |
| CN113221960A (en) | Construction method and collection method of high-quality vulnerability data collection model | |
| Bhardwaj et al. | Health insurance amount prediction | |
| CN116523284A (en) | Automatic evaluation method and system for business operation flow based on machine learning | |
| CN115470854A (en) | Information system fault classification method and classification system | |
| CN114021425B (en) | Power system operation data modeling and feature selection method and device, electronic equipment and storage medium | |
| CN114662405A (en) | Rock burst prediction method based on few-sample measurement and ensemble learning | |
| CN111737993B (en) | Method for extracting equipment health state from fault defect text of power distribution network equipment | |
| CN114418146A (en) | Electromechanical equipment fault diagnosis method, system and terminal based on rule engine | |
| CN113868422A (en) | Multi-label inspection work order problem traceability identification method and device | |
| CN117390407A (en) | Fault identification method, system, medium and equipment of substation equipment | |
| CN112215254A (en) | Transformer substation fault diagnosis method and diagnosis device based on improved case reasoning | |
| CN112748331A (en) | Circuit breaker mechanical fault identification method and device based on DS evidence fusion | |
| CN115438190B (en) | Power distribution network fault auxiliary decision knowledge extraction method and system | |
| CN112000708A (en) | Abnormal data processing method and system based on regulation and control adapted data fusion | |
| CN111555908A (en) | Power grid complex fault event identification method, device and system based on expert experience rule | |
| CN111385273B (en) | Internet of things business process identification method and device, electronic equipment and medium | |
| CN113891342A (en) | Base station inspection method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |