CN106487592A - A kind of Faults in Distributed Systems diagnostic method based on data cube - Google Patents
A kind of Faults in Distributed Systems diagnostic method based on data cube Download PDFInfo
- Publication number
- CN106487592A CN106487592A CN201610921292.1A CN201610921292A CN106487592A CN 106487592 A CN106487592 A CN 106487592A CN 201610921292 A CN201610921292 A CN 201610921292A CN 106487592 A CN106487592 A CN 106487592A
- Authority
- CN
- China
- Prior art keywords
- data
- test
- value
- test data
- failure probability
- 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.)
- Granted
Links
- 238000002405 diagnostic procedure Methods 0.000 title claims abstract description 8
- 238000012360 testing method Methods 0.000 claims abstract description 88
- 238000000034 method Methods 0.000 claims abstract description 38
- 238000003745 diagnosis Methods 0.000 claims abstract description 28
- 238000012544 monitoring process Methods 0.000 claims abstract description 22
- 230000008569 process Effects 0.000 claims description 11
- 238000010606 normalization Methods 0.000 claims description 7
- 230000004888 barrier function Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000000547 structure data Methods 0.000 claims 1
- 230000000875 corresponding effect Effects 0.000 description 11
- 239000000523 sample Substances 0.000 description 11
- 238000005516 engineering process Methods 0.000 description 8
- 238000007405 data analysis Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 230000002159 abnormal effect Effects 0.000 description 4
- 238000003066 decision tree Methods 0.000 description 3
- 238000012631 diagnostic technique Methods 0.000 description 3
- 238000007619 statistical method Methods 0.000 description 3
- 238000011985 exploratory data analysis Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000004141 dimensional analysis Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000007935 neutral effect Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
- H04L41/0677—Localisation of faults
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Environmental & Geological Engineering (AREA)
- Test And Diagnosis Of Digital Computers (AREA)
- Debugging And Monitoring (AREA)
Abstract
The present invention provides a kind of Faults in Distributed Systems diagnostic method based on data cube, and its method and step is as follows：Obtain distributed system test data, it includes external test data and internal monitoring location data；Test data pretreatment, sets up test data set；Build data cube：By analytical data cube, diagnose Faults in Distributed Systems；Execute fault diagnosis again.The efficiency being accurately positioned fault can be significantly improved by the method, substantially reduce the requirement to fault location personnel.
Description
Technical field
The invention belongs to data mining technology field is and in particular to a kind of Faults in Distributed Systems based on data cube
Diagnostic method.
Background technology
Distributed system is collectively formed by the software and hardware facilities being deployed in different geographical, to the Internet user of zones of different
Service is provided.Monitoring for distributed system generally includes the external testing of analog subscriber and to internal system equipment and software
Deng monitoring.Because distributed system self structure is complicated, a system failure may cause multinomial monitoring index abnormal and big
Amount test failure, system manager is difficult to rapidly and accurately tracing trouble reason.Meanwhile, the prison to equipment and software for the internal system
Survey and often there is blind spot so that internal monitoring index is all normal, but still thrashing occurs.Therefore, by external test data
Concluded with internal monitoring data and analyzed, the invention auxiliary that effectively Faults in Distributed Systems based on data analysiss diagnoses
Instrument is highly desirable to.
Technology related to the present invention includes intelligent trouble diagnosis, distributed information system FLT, data are divided
Analysis technology.
Fault diagnosis technology is a kind of to utilize equipment current state information and historical situation, by certain analysis method to setting
The status recognition technique that standby state is evaluated.Intelligent trouble diagnosis is to utilize artificial intelligence technology, by system current state and
Historical situation is described as mathematical symbol, determines the technology at guilty culprit substantially position by methods such as mathematical logic, machine learning.
Intelligent Fault Diagnosis Technique includes fault tree analysiss, rule-based reasoning, reasoning by cases, decision tree, neutral net, diagnosis Bayesian network
Network diagnostic techniquess etc..Fault Tree Analysis have powerful knowledge representation ability, but obtain tired for complication system diagnostic knowledge
Difficult.Rule-based reasoning and reasoning by cases diagnostic techniquess explicitly represent, store and process knowledge with sign format, represent directly perceived, easily
Understand, but shortcoming to be diagnostic techniquess knowledge acquisition based on symbolic reasoning difficult, inference speed is slow, being not particularly suitable for real-time diagnosis will
Seek higher diagnostic field.Decision tree and Neural Network Diagnosis technology are represented and stored knowledge with numerical matrix form, calculate
Journey is equivalent to reasoning process, is not required to human intervention, and inference speed is fast, but shortcoming is decision tree and diagnoses Bayesian network method no
Method diagnoses unknown failure, and the practical significance of neural network model parameter is difficult to explain.
Developing rapidly with distributed information system, data interaction behavior becomes increasingly complex, and the system failure constantly rises,
The intelligent maintenance of distributed information system also increasingly receives publicity.Generally the monitoring to system running state includes monitoring system
Hardware performance index, the attribute of monitor operating system, handling capacity of monitoring application program etc., and dug using data further
Pick, machine learning and statistical method carry out the failture evacuation of automatization and reduce manual intervention.Distributed based on event handling
Information system FLT passes through to build a kind of event flow model based on set, event is carried out with formal definitions and determines
Justice operation is so that user defines the diagnosis rule of complexity by grasping several simple set operations.IBM research worker is passed through
Active probe carries out fault diagnosis with reference to Bayesian network, proposes the approximate solution that sequential multiple faults method tries to achieve problem.Above two
The method of kind remains the diagnosis to known fault type.
Data analysiss refer to be analyzed to collecting the mass data come with suitable statistical analysis technique, extract useful letter
Breath and form conclusion and the process to data in addition research and summary in detail.In practicality, data analysiss can help people
Judge, to take appropriate action.In field of statistics, some people by data analysiss be divided into descriptive statistical analysis,
Exploratory data analysis and confirmatory data analysiss；Wherein, exploratory data analysis lay particular emphasis on find among data new
Feature, and confirmatory data analysiss then lay particular emphasis on the confirmation assumed or falsfication.Analysis method mainly has tabulating method, mapping
Method, and simple mathematical computing, statistics, fast Fourier transform, smooth, baseline analysis etc..In recent years, with the big data epoch
Arrival, the acquisition of data and storage capacity obtained unprecedented development, and the value of data increasingly highlights, and data analysiss exist
The industries such as medical treatment, communication, the energy are goed deep into and are widely applied.
Data cube is a class multidimensional data, allows user to explore and analytical data collection from multiple angles.Distributed system
Test data set not only includes analog subscriber from different geographical, operator, accesses the test of the combination of dissimilar service, test
Data set yet forms both and covers distributed system internal difference deployment area, device type, software type, operating system
The attributes such as state, have therefore naturally constituted multidimensional data.Have not yet to see and examined based on the distributed fault of data cube
Disconnected method.
Content of the invention
In order to make up above-mentioned deficiency, the present invention provides a kind of Faults in Distributed Systems diagnosis side based on data cube
Method, describes test data set by obtaining data cube, and carries out dimensional analysis to the data cube building；Thus it is quick
Diagnose Faults in Distributed Systems exactly, effectively prevent puzzlement and the economic loss of thrashing generation.
In order to realize foregoing invention purpose, the present invention adopts the following technical scheme that：
A kind of Faults in Distributed Systems diagnostic method based on data cube, methods described comprises the steps：
(1) obtain distributed system test data, it includes external test data and internal monitoring location data；
(2) test data pretreatment, sets up test data set；
(3) build data cube：
(4) Faults in Distributed Systems is diagnosed by analytical data cube；
(5) execute fault diagnosis again.
Preferably, the external test data of described step (1) is covered by distributed system different operators and test
Point periodically carries out measure of merit acquisition, tests corresponding testing location, test operator, COS and test knot including each
Really.
Preferably, in described step (1), obtain internal monitoring location data and include：Determine distributed system inner track,
The network equipment of record flow process, server, software type, and server CPU usage and running status value.
Preferably, in described step (2), test data pretreatment includes：By external test data and internal monitoring positioning
Data summarization；If comprising successive value in its property value, execution data generaliza-tion is processed so as to be changed into centrifugal pump；Obtain attribute to take
When value is different, the testing time in test data and Failure count；Generate test data set.
Preferably, described step (3) builds data cube and includes：Define data cube and (a is constituted by n dimension1,
a2,...,an), each attribute corresponds to a dimension；Define the value of each dimension, be separately recorded under different values, test number
According to the testing time of collection, Failure count and failure probability.
Preferably, described step (4) diagnosis Faults in Distributed Systems includes：Test sample is projected to data cube not
With in the corresponding dimension of attribute；When a certain equipment or software break down or running status is abnormal, sample project to each
Failure probability in dimension increases therewith, and wherein, failure probability difference is the source of trouble to the maximum；By calculating testing time, inefficacy
Number of times and failure probability, determine the source of trouble；It is specially：
Defining the failure probability Diversity measure index in the corresponding each dimension of internal monitoring item is Si；Select SiIn
Little value Sg；If SgLess than decision threshold, then write down its attribute ag, and to agJudge further；Otherwise it is assumed that there is not event in system
Barrier, diagnosis process terminates；
Determine attribute agDifferent values xg,jMiddle failure probability maximum, analyzes equipment corresponding to this value, software or fortune
The failure cause of row state.
Further, described failure probability Diversity measure index SiExpression formula be：
In formula, P 'ijFor similar information entropy, by normalization failure probability PijObtain；ΩiRepresent attribute aiThe collection of value
Close；J is ΩiIn j-th value.
Further, described xjFailure probability be P (C | ai=xi,j,xi,j∈Ωi), wherein ΩiFor attribute aiValue collection
Close, C represents that test was lost efficacy；
By P (C | ai=xi,j,xi,j∈Ωi) it is denoted as Pij, failure probability PijObtained according to the measuring and calculation that test data is concentrated
, its expression formula is：
In formula, nm (A)And nm (C)Represent the testing time of different attribute value combination and the Failure count of serial number m respectively.
Further, by PijNormalization, obtains similar information entropy P 'ij's
Expression formula：
Preferably, described step (5) executes fault diagnosis again and includes：Delete faulty equipment, software or running status phase
After the test sample closed, remaining test sample is carried out again with the fault diagnosis of step (4), until failure probability Diversity measure
When index is all higher than decision threshold, diagnosis terminates.
Compared with prior art, the beneficial effect that the present invention reaches is：
1) present invention projects to test sample in the corresponding dimension of data cube different attribute；When certain equipment, soft
When part fault or running status are abnormal, the failure probability that correlated sampless project in each dimension has increased, especially the source of trouble
Failure probability difference in corresponding dimension is the most notable.The statistical nature of test data when exactly utilizing said system fault, can
Rapidly and accurately to find fault；Thus significantly improving the efficiency being accurately positioned fault, greatly reduce to fault location personnel
Requirement.
2) this method is passed through to count the diversity tracing trouble of each dimension failure probability, is applicable not only to find definitely to draw
Play the factor of thrashing, apply also for the factor finding to break down with greater probability.There is phase between the attribute of system
During closing property, the attribute of the maximum that preferentially finds differences, the attribute of higher Concept Hierarchies, the event of the higher Concept Hierarchies of the minority of discovery
Barrier, more tallies with the actual situation.
Brief description
Fig. 1 is the Faults in Distributed Systems diagnostic method flow chart based on data cube；
Specific embodiment
Below with reference to accompanying drawing, the specific embodiment of the present invention is described in further detail.
As shown in figure 1, a kind of Faults in Distributed Systems diagnostic method based on data cube, methods described includes following
Step：
(1) obtain distributed system test data, it includes external test data and internal monitoring location data；Outside survey
The different operators that examination data is covered by distributed system and test point periodically carry out measure of merit acquisition, test including each
Corresponding testing location, test operator, COS and test result.
Obtain internal monitoring location data to include：Determine distributed system inner track, the network that record flow passes through sets
Standby, server, software type, and server CPU usage and running status value.
(2) test data pretreatment, sets up test data set；Test data pretreatment includes：By external test data and
Internal monitoring location data collects；If comprising successive value in its property value, execution data generaliza-tion is processed so as to be changed into discrete
Value；Obtain attribute value different when, the testing time in test data and Failure count；Generate test data set；As table 1 institute
Show.
Table 1 test data set example
Note：N in table(A)Represent testing time, n(C)Represent Failure count.
(3) build data cube：Define data cube and (a is constituted by n dimension1,a2,...,an), each attribute pair
Answer a dimension；Define the value of each dimension, be separately recorded under different values, the testing time of test data set, inefficacy
Number of times and failure probability.
(4) Faults in Distributed Systems is diagnosed by analytical data cube；Concrete steps include：
Test sample is projected in the corresponding dimension of data cube different attribute；When a certain equipment or software occur event
When barrier or running status are abnormal, the failure probability in each dimension that sample projects to increases therewith；Wherein, failure probability difference
It is the source of trouble to the maximum；By calculating testing time, Failure count and failure probability, determine the source of trouble；It is specially：
Defining the failure probability Diversity measure index in the corresponding each dimension of internal monitoring item is Si；Select SiIn
Little value Sg；If SgLess than decision threshold, then write down its attribute ag, and to agJudge further；Otherwise it is assumed that there is not event in system
Barrier, diagnosis process terminates；For example, COS in table 1, system region, the network equipment, computation attribute is poor in the dimension of server
Different value, respectively 0.93,0.76,0.74,0.82,0.99, minima 0.74 is less than given threshold value 0.9, therefore, it is determined that system is deposited
In fault.
Determine attribute agDifferent values xg,jMiddle failure probability maximum, analyzes equipment corresponding to this value, software or fortune
The failure cause of row state.For example, in table 1 network equipment N11, N12, N21, N22 crash rate be respectively 0.38,0.08,
0.05th, 0.08 it is possible to determine that be that equipment N11 breaks down.
Wherein, failure probability Diversity measure index SiExpression formula be：
In formula, P 'ijFor similar information entropy, by normalization failure probability PijObtain；ΩiRepresent attribute aiThe collection of value
Close；J is ΩiIn j-th value.
Described xjFailure probability be P (C | ai=xi,j,xi,j∈Ωi), wherein ΩiFor attribute aiValue set, C represents survey
Examination was lost efficacy；
By P (C | ai=xi,j,xi,j∈Ωi) it is denoted as Pij, failure probability PijObtained according to the measuring and calculation that test data is concentrated
, its expression formula is：
In formula, nm (A)And nm (C)Represent the testing time of different attribute value combination and the Failure count of serial number m respectively.
PijNormalization, obtains similar information entropy P 'ijExpression formula：
Specifically the operating process of fault diagnosis is：Calculate the corresponding each dimension (a of internal monitoring item1,a2,...,an) on
Failure probability Diversity measure index Si；For attribute ai, this property value is xjWhen, failure probability be P (C | ai=xi,j,xi,j∈
Ωi), wherein ΩiIt is attribute aiThe set of value, C represents that test was lost efficacy.For convenience, by P (C | ai=xi,j,xi,j∈
Ωi) it is abbreviated as Pij.Failure probability PijValue pass through data set in test record calculate,Wherein nm (A)And nm (C)Represent the testing time of different attribute value combination and the Failure count of serial number m respectively.By PijNormalization, obtains class
Like the computing formula of comentropy, that is,
Then attribute aiThe computing formula of diversity be：
S in formulaiSpan is [0,1].SiValue less mark difference is more obvious, is worth and represents that difference is the most obvious for 0,
Namely aiWhen only taking some value, crash rate is not 0, takes and is 0 during other values.SiIt is worth, i.e. P ' minimum for 1 mark differenceij=
1/|Ωi|.Obtain SiMinima S in valuegIf, SgLess than decision threshold STWhen be determined that fault, if SgMore than judgement
Threshold value STThen it is assumed that system does not have fault, diagnosis process terminates.
(5) execute fault diagnosis again.It is specially：Delete the related test sample of faulty equipment, software or running status
Afterwards, remaining test sample is carried out again with the fault diagnosis of step (4), until failure probability Diversity measure index is all higher than sentencing
When determining threshold value, diagnosis terminates.It is N that such as sample in table 1 removes the network equipment11Record, calculate the difference in each dimension again
Different value is 0.99,0.99,0.99,0.97,0.99 successively, is all higher than threshold value ST=0.9, thus judge that event is not had on remaining dimension
Barrier.
Finally it should be noted that：Above example is only not intended to limit in order to technical scheme to be described, institute
The those of ordinary skill in genus field still the specific embodiment of the present invention can be modified with reference to above-described embodiment or
Equivalent, these are all applying for pending this without departing from any modification of spirit and scope of the invention or equivalent
Within bright claims.
Claims (10)
1. a kind of Faults in Distributed Systems diagnostic method based on data cube it is characterised in that methods described include following
Step：
(1) obtain distributed system test data, it includes external test data and internal monitoring location data；
(2) test data pretreatment, sets up test data set；
(3) build data cube：
(4) Faults in Distributed Systems is diagnosed by analytical data cube；
(5) execute fault diagnosis again.
2. method as claimed in claim 1 is it is characterised in that the external test data of described step (1) passes through distributed system
The different operators covering and test point periodically carry out measure of merit acquisition, test corresponding testing location, test including each
Operator, COS and test result.
3. method as claimed in claim 1 is it is characterised in that in described step (1), obtaining internal monitoring location data and include：
Determine that distributed system inner track, the network equipment of record flow process, server, software type, and server CPU account for
With rate and running status value.
4. method as claimed in claim 1 is it is characterised in that in described step (2), test data pretreatment includes：By outside
Test data and internal monitoring location data collect；If comprising successive value in its property value, execution data generaliza-tion process so as to
It is changed into centrifugal pump；Obtain attribute value different when, the testing time in test data and Failure count；Generate test data set.
5. method as claimed in claim 1 is it is characterised in that described step (3) structure data cube includes：Define data to stand
Cube constitutes (a by n dimension1,a2,...,an), each attribute corresponds to a dimension；Define the value of each dimension, remember respectively
Record under different values, the testing time of test data set, Failure count and failure probability.
6. method as claimed in claim 1 is it is characterised in that described step (4) diagnosis Faults in Distributed Systems includes：To test
Sample projects in the corresponding dimension of data cube different attribute；When a certain equipment or software breaks down or running status is different
Chang Shi, the failure probability in each dimension that sample projects to increases therewith, and wherein, failure probability difference is the source of trouble to the maximum；
By calculating testing time, Failure count and failure probability, determine the source of trouble；It is specially：
Defining the failure probability Diversity measure index in the corresponding each dimension of internal monitoring item is Si；Select SiIn minima
Sg；If SgLess than decision threshold, then write down its attribute ag, and to agJudge further；Otherwise it is assumed that system does not have fault, examine
Disconnected process terminates；
Determine attribute agDifferent values xg,jMiddle failure probability maximum, analyzes the equipment corresponding to this value, software or runs shape
The failure cause of state.
7. method as claimed in claim 6 is it is characterised in that described failure probability Diversity measure index SiExpression formula be：
In formula, P 'ijFor similar information entropy, by normalization failure probability PijObtain；ΩiRepresent attribute aiThe set of value；J is
ΩiIn j-th value.
8. method as claimed in claim 6 is it is characterised in that described xi,jFailure probability be P (C | ai=xi,j,xi,j∈Ωi),
Wherein ΩiFor attribute aiValue set, C represents that test was lost efficacy；By P (C | ai=xi,j,xi,j∈Ωi) it is denoted as Pij, lost efficacy general
Rate PijObtained according to the measuring and calculation that test data is concentrated, its expression formula is：
In formula, nm (A)And nm (C)Represent the testing time of different attribute value combination and the Failure count of serial number m respectively.
9. method as claimed in claim 7 is it is characterised in that by PijNormalization, obtains similar information entropy P 'ij's
Expression formula：
10. method as claimed in claim 1 is it is characterised in that described step (5) executes fault diagnosis again includes：Delete event
After the related test sample of barrier equipment, software or running status, the fault carrying out step (4) again to remaining test sample is examined
Disconnected, when failure probability Diversity measure index is all higher than decision threshold, diagnosis terminates.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610921292.1A CN106487592B (en) | 2016-10-21 | 2016-10-21 | A kind of Faults in Distributed Systems diagnostic method based on data cube |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610921292.1A CN106487592B (en) | 2016-10-21 | 2016-10-21 | A kind of Faults in Distributed Systems diagnostic method based on data cube |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106487592A true CN106487592A (en) | 2017-03-08 |
| CN106487592B CN106487592B (en) | 2019-09-27 |
Family
ID=58270340
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610921292.1A Expired - Fee Related CN106487592B (en) | 2016-10-21 | 2016-10-21 | A kind of Faults in Distributed Systems diagnostic method based on data cube |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106487592B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109558258A (en) * | 2017-09-27 | 2019-04-02 | 国家计算机网络与信息安全管理中心 | A kind of method and device of distributed system root fault location |
| CN111612770A (en) * | 2020-05-22 | 2020-09-01 | 中国科学院自动化研究所 | Active screening-based focus detection system of semi-supervised focus detection network |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101170729A (en) * | 2007-11-30 | 2008-04-30 | 中国移动通信集团重庆有限公司 | A GSM network alarming analysis system and network alarm analysis method |
| US20100204857A1 (en) * | 2006-09-18 | 2010-08-12 | Bombardier Transportation Gmbh | Diagnostic system and method for monitoring a rail system |
| CN102521417A (en) * | 2011-12-30 | 2012-06-27 | 南京柏梭信息科技有限公司 | Method for processing multi-dimensional data based on virtual data cube and system of method |
| CN103676794A (en) * | 2012-09-04 | 2014-03-26 | 上海杰之能信息科技有限公司 | Energy consumption monitoring method and energy consumption monitoring system |
| CN104182830A (en) * | 2014-08-14 | 2014-12-03 | 天津大学 | A method for mining a weak reliability link of a power distribution system based on multi-dimensional analysis |
| CN104616092A (en) * | 2014-12-16 | 2015-05-13 | 国家电网公司 | Distributed log analysis based distributed mode handling method |
| CN105679309A (en) * | 2014-11-21 | 2016-06-15 | 科大讯飞股份有限公司 | Method and device for optimizing speech recognition system |
-
2016
- 2016-10-21 CN CN201610921292.1A patent/CN106487592B/en not_active Expired - Fee Related
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100204857A1 (en) * | 2006-09-18 | 2010-08-12 | Bombardier Transportation Gmbh | Diagnostic system and method for monitoring a rail system |
| CN101170729A (en) * | 2007-11-30 | 2008-04-30 | 中国移动通信集团重庆有限公司 | A GSM network alarming analysis system and network alarm analysis method |
| CN102521417A (en) * | 2011-12-30 | 2012-06-27 | 南京柏梭信息科技有限公司 | Method for processing multi-dimensional data based on virtual data cube and system of method |
| CN103676794A (en) * | 2012-09-04 | 2014-03-26 | 上海杰之能信息科技有限公司 | Energy consumption monitoring method and energy consumption monitoring system |
| CN104182830A (en) * | 2014-08-14 | 2014-12-03 | 天津大学 | A method for mining a weak reliability link of a power distribution system based on multi-dimensional analysis |
| CN105679309A (en) * | 2014-11-21 | 2016-06-15 | 科大讯飞股份有限公司 | Method and device for optimizing speech recognition system |
| CN104616092A (en) * | 2014-12-16 | 2015-05-13 | 国家电网公司 | Distributed log analysis based distributed mode handling method |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109558258A (en) * | 2017-09-27 | 2019-04-02 | 国家计算机网络与信息安全管理中心 | A kind of method and device of distributed system root fault location |
| CN109558258B (en) * | 2017-09-27 | 2023-01-24 | 国家计算机网络与信息安全管理中心 | Method and device for positioning root fault of distributed system |
| CN111612770A (en) * | 2020-05-22 | 2020-09-01 | 中国科学院自动化研究所 | Active screening-based focus detection system of semi-supervised focus detection network |
| CN111612770B (en) * | 2020-05-22 | 2021-08-31 | 中国科学院自动化研究所 | Active screening-based focus detection system of semi-supervised focus detection network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106487592B (en) | 2019-09-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106354118B (en) | A kind of train fault diagnostic system and method based on fault tree | |
| Mohan et al. | Cascading spatio-temporal pattern discovery | |
| Liu et al. | Fault diagnosis and cause analysis using fuzzy evidential reasoning approach and dynamic adaptive fuzzy Petri nets | |
| CN105677791B (en) | For analyzing the method and system of the operation data of wind power generating set | |
| CN107146018A (en) | Data Management Analysis method and system based on electric network state monitoring system | |
| CN109555566A (en) | A kind of turbine rotor method for diagnosing faults based on LSTM | |
| CN106776208B (en) | Fault Locating Method when a kind of running software | |
| Wu et al. | Microdiag: Fine-grained performance diagnosis for microservice systems | |
| CN104765965A (en) | GIS fault diagnosis and reliability analysis method based on fuzzy Petri | |
| CN114021932A (en) | Energy efficiency evaluation and diagnosis method, system and medium for wind turbine generator | |
| Huang et al. | Review of data-driven prognostics and health management techniques: lessions learned from PHM data challenge competitions | |
| CN107301118A (en) | A kind of fault indices automatic marking method and system based on daily record | |
| CN103678881B (en) | Composite fault diagnosis method based on combination of artificial immunity and evidence theory | |
| CN112906764B (en) | Communication safety equipment intelligent diagnosis method and system based on improved BP neural network | |
| Momtazpour et al. | Analyzing invariants in cyber-physical systems using latent factor regression | |
| Li et al. | Quantitative evaluations on saliency methods: An experimental study | |
| CN117609836A (en) | Electromagnetic sensitivity prediction and health management method for integrated module | |
| CN106487592B (en) | A kind of Faults in Distributed Systems diagnostic method based on data cube | |
| Sönmez et al. | Anomaly detection using data mining methods in it systems: a decision support application | |
| Yuan et al. | Learning-based anomaly cause tracing with synthetic analysis of logs from multiple cloud service components | |
| Pan | An improved Apriori algorithm for association mining between physical fitness indices of college students | |
| Chen et al. | Interpretable fault diagnosis with shapelet temporal logic: Theory and application | |
| CN106909136A (en) | A kind of method for diagnosing faults based on index regularization kernel linear discriminant analysis | |
| Dwivedi et al. | DynamoPMU: A Physics Informed Anomaly Detection, Clustering and Prediction Method using Non-linear Dynamics on μPMU Measurements | |
| Ye et al. | An intrusion detection approach based on system call sequences and rules extraction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190927 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |