CN108667648A - A kind of alarm sequential pattern mining method based on network and time-constrain - Google Patents
A kind of alarm sequential pattern mining method based on network and time-constrain Download PDFInfo
- Publication number
- CN108667648A CN108667648A CN201810289630.3A CN201810289630A CN108667648A CN 108667648 A CN108667648 A CN 108667648A CN 201810289630 A CN201810289630 A CN 201810289630A CN 108667648 A CN108667648 A CN 108667648A
- Authority
- CN
- China
- Prior art keywords
- alarm
- sequence
- time
- network
- sequential pattern
- 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
Classifications
-
- 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
- H04L41/0631—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
- H04L41/064—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis involving time analysis
-
- 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
- H04L41/0631—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
-
- 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
- H04L41/0631—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
- H04L41/065—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis involving logical or physical relationship, e.g. grouping and hierarchies
Abstract
The alarm sequential pattern mining method based on network and time-constrain that the invention discloses a kind of, including alarm initiation sequence generate：The alarm of certain time in Network Management System is obtained, alarm initiation sequence is generated；Network constraint is analyzed：It is several alarm sequences by alarm initiation sequence dismantling, every alarm sequence corresponds to one group, and there are the equipment of network constraint relationship；Time Constraints Analysis：Setting time window filters out the alarm except time window in alarm sequence, obtains new alarm sequence；Alarm sequential pattern mining：Based on new alarm sequence, alarm sequence pattern is excavated；Specification is alerted to generate：Based on alarm sequence pattern, final alarm regulation is generated.The present invention solves dependence of the conventional alarm correlation analysis for priori, pass through network constraint and time-constrain simultaneously, ensure the specific aim and efficiency of alarm sequential pattern mining, power telecom network alarm correlation analysis and trouble diagnosibility are improved, communication operation and management level is improved.
Description
Technical field
The alarm sequential pattern mining method based on network and time-constrain that the present invention relates to a kind of belonging to power communication neck
Domain.
Background technology
With the fast development of powerline network and business, power system telecommunications technology has become a weight of power generation
Link, the reliability of electric power communication transmission network network is wanted to become the important component for affecting power generation safety.Alarm is event
The external presentation of barrier has the characteristics that transitivity and many-sided, too many levels show [2], i.e. a root alarm can cause a large amount of times
Derive alarm, in some instances it may even be possible to cause " alarm windstorm ".Operation maintenance personnel as can at the first time to system alarm accurate judgement and and
When handle, fault handling time can be greatly reduced, it is rapid to restore operation system normal use, therefore alarm correlation analysis be therefore
The basis of barrier analysis and processing.
However there are the following problems for the traditional means of alarm correlation analysis：
1) prioris such as traditional means Main Basiss alarm regulation tree, white list, need to be proficient in related service and technology
Expert's support, and the rule of each producer, each device type is different, and huge expert team is needed to support
2) electric power transmission network is complicated and changeable, and alarm quantity is huge, real-time, and traditional analysis means can not adapt to, rule
The difficulty and workload then safeguarded are very big, can not automatically update rule tree, white list, may result in alarm correlation analysis
Accuracy and efficiency substantially reduce.
Invention content
The alarm sequence pattern based on network and time-constrain that in order to solve the above technical problem, the present invention provides a kind of
Method for digging.
In order to achieve the above object, the technical solution adopted in the present invention is：
A kind of alarm sequential pattern mining method based on network and time-constrain, includes the following steps,
Initiation sequence is alerted to generate：The alarm of certain time in Network Management System is obtained, alarm initiation sequence is generated；
Network constraint is analyzed：It is several alarm sequences by alarm initiation sequence dismantling, every alarm sequence corresponds to one group
There are the equipment of network constraint relationship；
Time Constraints Analysis：Setting time window filters out the alarm except time window in alarm sequence, obtains new alarm
Sequence；
Alarm sequential pattern mining：Based on new alarm sequence, alarm sequence pattern is excavated；
Specification is alerted to generate：Based on alarm sequence pattern, final alarm regulation is generated.
Alarm is pre-processed, alarm initiation sequence is generated, alerting pretreated process is,
Alarm is temporally ranked up；
Different alarm types is mapped with different parameters；
An alarm is indicated with parameter and time, and temporally forms alarm initiation sequence.
The representation of alarm is K (i), wherein K is parameter, indicates that alarm type, i are the relative time that alarm occurs,
The difference for the time that the time and first alarm that i.e. the alarm occurs occur.
According to Network Management System service fulfillment situation, the equipment collection for being related to alarm is divided into several groups, every group of correspondence one
A business, every alarm sequence correspond to one group of equipment.
Frequent mode threshold value C is set, all frequent modes more than threshold value are found out.
Confidence level Conf is set, alarm sequence pattern is based on, generates final alarm regulation, that is, the alarm of user is supplied to close
Join knowledge.
The advantageous effect that the present invention is reached：1, the present invention without knowing alarm regulation in advance, to O＆M and administrative staff
Requirement substantially reduce；2, automatically updating for alarm association knowledge may be implemented in the present invention, improves alarm correlation analysis
Accuracy and efficiency；3, the present invention improves the specific aim and efficiency of alarm sequential pattern mining by network and time-constrain,
Relative to general alarm sequential pattern mining method, accuracy and efficiency higher.
Description of the drawings
Fig. 1 is the flow chart of the present invention；
Fig. 2 is that powerline network transmits topological embodiment exemplary view；
Fig. 3 be sequential mode mining embodiment exemplary view.
Specific implementation mode
The invention will be further described below in conjunction with the accompanying drawings.Following embodiment is only used for clearly illustrating the present invention
Technical solution, and not intended to limit the protection scope of the present invention.
As shown in Figure 1, a kind of alarm sequential pattern mining method based on network and time-constrain, includes the following steps：
Step 1, alarm initiation sequence generates：The alarm for obtaining certain time in Network Management System carries out alarm pre-
Processing generates alarm initiation sequence.
Alerting pretreated process is：
Alarm is temporally ranked up；
Different alarm types is mapped with different parameters；
An alarm is indicated with parameter and time, and temporally forms alarm initiation sequence；The representation of alarm is K
(i), wherein K is parameter, indicates that alarm type, i are the relative time that alarm occurs, i.e., the time and first that the alarm occurs
The difference for the time that a alarm occurs.
Step 2, network constraint is analyzed：It is several alarm sequences, every alarm sequence pair by alarm initiation sequence dismantling
Answering one group, there are the equipment of network constraint relationship.
According to Network Management System service fulfillment situation, the equipment collection for being related to alarm is divided into several groups, every group of correspondence one
A business, every alarm sequence correspond to one group of equipment.
Step 3, Time Constraints Analysis：Setting time window filters out the alarm except time window in alarm sequence, obtains new
Alarm sequence；
Step 4, alarm sequential pattern mining：Based on new alarm sequence, alarm sequence pattern is excavated.
Frequent mode threshold value C is set, finds out all frequent modes more than threshold value, the discovery procedure of frequent mode includes repeatedly
Two stages in generation, frequent mode and beta pruning, obtain 1- item collections, 2- item collections, 3- item collections ..., l- item collections successively；Given support
Threshold xi, if numbers of the sequence x in sequence library S is not less than ξ, sequence x is referred to as sequence pattern, the pattern that length is 1
As 1- patterns.
Step 5, alarm specification generates：Confidence level Conf is set, alarm sequence pattern is based on, generates final alarm regulation,
It is supplied to the alarm association knowledge of user.
It is illustrated with reference to specific embodiment：
Certain grid company section time alarm record data are obtained, as shown in table 1, by taking this alarm data record as an example, are given
Go out the embodiment of the alarm sequential pattern mining method of the present invention based on network and time-constrain.
Certain section of time alarm record of table 1
Step 1, alarm initiation sequence generate；
It is apparent intuitive to state, alarm type is mapped, as shown in table 2：
2 alarm type mapping table of table
Based on alarm type mapping table, alarm initiation sequence={ A (0), A (1), B (1), C (2), C (2), D can be obtained
(2), (2) G, E (3), G (4), F (5), F (6), E (10), F (12), D (35) }, relative time is removed, table 3 can be reduced to
Method indicates：
Table 3 alerts initiation sequence
Wherein, the expression in bracket occurs simultaneously.
Step 2, network constraint analysis；
It alerts and comes from equipment collection M, wherein equipment collection M=equipment 1, and equipment 2, equipment 3, equipment 4, equipment 5, equipment 6, if
Standby 7 }, it is as shown in Fig. 2 to obtain equipment connecting relation in conjunction with Network Management System.
In conjunction with Network Management System service fulfillment situation, include in figure two circuits can be obtained, respectively：Equipment 1<->
Equipment 3<->Equipment 4；Equipment 2<->Equipment 5<->Equipment 6<->Equipment 7；Therefore alarm initiation sequence can be split as two strip sequences
Row, as shown in table 4.
Sequence under 4 network constraint of table
Step 3, Time Constraints Analysis；
It is assumed that time window 15s, i.e., the rear 15s since alerting first, filter out the alarm except time window, when can be obtained
Between sequence under constraints, as shown in table 5.
Sequence under 5 time-constrain of table
Step 4, alarm sequential pattern mining；
It is assumed that frequent mode threshold value 2, finds out all frequent modes more than threshold value.The discovery procedure of frequent mode includes repeatedly
Two stages in generation, frequent mode and beta pruning, obtain 1- item collections successively, 2- item collections, 3- item collections, 4- item collections, as shown in Fig. 3.
Step 5, alarm regulation generate
It is assumed that confidence level is 70%, the alarm sequence pattern obtained based on a upper trifle<A,C,E,F>, generate final alarm
Specification：R_LOS->MS_AIS->AU_AIS->TU_AIS.
The above method solves dependence of the conventional alarm correlation analysis for priori, at the same by network constraint and
Time-constrain, it is ensured that the specific aim and efficiency of alarm sequential pattern mining improve power telecom network alarm correlation analysis
And trouble diagnosibility, improve communication operation and management level.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, without departing from the technical principles of the invention, several improvement and deformations can also be made, these improvement and deformations
Also it should be regarded as protection scope of the present invention.
Claims (6)
1. a kind of alarm sequential pattern mining method based on network and time-constrain, it is characterised in that：Include the following steps,
Initiation sequence is alerted to generate：The alarm of certain time in Network Management System is obtained, alarm initiation sequence is generated；
Network constraint is analyzed：It is several alarm sequences by alarm initiation sequence dismantling, every alarm sequence corresponds to one group of presence
The equipment of network constraint relationship；
Time Constraints Analysis：Setting time window filters out the alarm except time window in alarm sequence, obtains new alarm sequence
Row；
Alarm sequential pattern mining：Based on new alarm sequence, alarm sequence pattern is excavated；
Specification is alerted to generate：Based on alarm sequence pattern, final alarm regulation is generated.
2. a kind of alarm sequential pattern mining method based on network and time-constrain according to claim 1, feature
It is：Alarm is pre-processed, alarm initiation sequence is generated, alerting pretreated process is,
Alarm is temporally ranked up；
Different alarm types is mapped with different parameters；
An alarm is indicated with parameter and time, and temporally forms alarm initiation sequence.
3. a kind of alarm sequential pattern mining method based on network and time-constrain according to claim 2, feature
It is：The representation of alarm is K (i), wherein K is parameter, indicates that alarm type, i are the relative time that alarm occurs, i.e.,
The difference for the time that the time and first alarm that the alarm occurs occur.
4. a kind of alarm sequential pattern mining method based on network and time-constrain according to claim 1, feature
It is：According to Network Management System service fulfillment situation, the equipment collection for being related to alarm is divided into several groups, one industry of every group of correspondence
Business, every alarm sequence correspond to one group of equipment.
5. a kind of alarm sequential pattern mining method based on network and time-constrain according to claim 1, feature
It is：Frequent mode threshold value C is set, all frequent modes more than threshold value are found out.
6. a kind of alarm sequential pattern mining method based on network and time-constrain according to claim 1, feature
It is：Confidence level Conf is set, alarm sequence pattern is based on, generates final alarm regulation, that is, be supplied to the alarm association of user
Knowledge.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810289630.3A CN108667648A (en) | 2018-04-03 | 2018-04-03 | A kind of alarm sequential pattern mining method based on network and time-constrain |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810289630.3A CN108667648A (en) | 2018-04-03 | 2018-04-03 | A kind of alarm sequential pattern mining method based on network and time-constrain |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108667648A true CN108667648A (en) | 2018-10-16 |
Family
ID=63782890
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810289630.3A Pending CN108667648A (en) | 2018-04-03 | 2018-04-03 | A kind of alarm sequential pattern mining method based on network and time-constrain |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108667648A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111122945A (en) * | 2019-12-31 | 2020-05-08 | 南京天溯自动化控制系统有限公司 | High-precision alarm filtering method and device for hospital logistics monitoring system |
| CN112312443A (en) * | 2020-10-13 | 2021-02-02 | 西安电子科技大学 | Mass alarm data processing method, system, medium, computer equipment and application |
| WO2021254244A1 (en) * | 2020-06-16 | 2021-12-23 | 中兴通讯股份有限公司 | Alarm mining model determination method and apparatus, device and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100211192A1 (en) * | 2009-02-17 | 2010-08-19 | Honeywell International Inc. | Apparatus and method for automated analysis of alarm data to support alarm rationalization |
| CN102136949A (en) * | 2011-03-24 | 2011-07-27 | 国网电力科学研究院 | Method and system for analyzing alarm correlation based on network and time |
| CN107835087A (en) * | 2017-09-14 | 2018-03-23 | 北京科东电力控制系统有限责任公司 | A kind of safety means alarm regulation extraction method based on Frequent Pattern Mining |
| CN107844540A (en) * | 2017-10-25 | 2018-03-27 | 电子科技大学 | A kind of time series method for digging for electric power data |
-
2018
- 2018-04-03 CN CN201810289630.3A patent/CN108667648A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100211192A1 (en) * | 2009-02-17 | 2010-08-19 | Honeywell International Inc. | Apparatus and method for automated analysis of alarm data to support alarm rationalization |
| CN102136949A (en) * | 2011-03-24 | 2011-07-27 | 国网电力科学研究院 | Method and system for analyzing alarm correlation based on network and time |
| CN107835087A (en) * | 2017-09-14 | 2018-03-23 | 北京科东电力控制系统有限责任公司 | A kind of safety means alarm regulation extraction method based on Frequent Pattern Mining |
| CN107844540A (en) * | 2017-10-25 | 2018-03-27 | 电子科技大学 | A kind of time series method for digging for electric power data |
Non-Patent Citations (2)
| Title |
|---|
| 姚伟力等: "基于序列模式挖掘的告警相关性分析算法", 《北京邮电大学学报》 * |
| 闫生超等: "基于网络和时间关联关系的告警相关性分析", 《电力系统自动化》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111122945A (en) * | 2019-12-31 | 2020-05-08 | 南京天溯自动化控制系统有限公司 | High-precision alarm filtering method and device for hospital logistics monitoring system |
| WO2021254244A1 (en) * | 2020-06-16 | 2021-12-23 | 中兴通讯股份有限公司 | Alarm mining model determination method and apparatus, device and storage medium |
| CN112312443A (en) * | 2020-10-13 | 2021-02-02 | 西安电子科技大学 | Mass alarm data processing method, system, medium, computer equipment and application |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3975048A1 (en) | Method for constructing cloud network alarm root cause relational tree model, device, and storage medium | |
| CN110493025B (en) | Fault root cause diagnosis method and device based on multilayer digraphs | |
| CN108536770B (en) | Distribution network ring net automatic generation method based on distribution GIS model | |
| CN109189736B (en) | Method and device for generating alarm association rule | |
| US6349334B1 (en) | Telecommunications network management method and system | |
| CN108667648A (en) | A kind of alarm sequential pattern mining method based on network and time-constrain | |
| US6816576B2 (en) | Tree hierarchy and description for generated logs | |
| CN116184961A (en) | Injection molding automatic laminating production line safety control system based on industrial Internet of things | |
| CN112685393B (en) | Intelligent alarm merging method and system for power communication network | |
| CN111722976A (en) | Fault flow analysis method, device and medium based on intelligent operation and maintenance | |
| CN113505048A (en) | Unified monitoring platform based on application system portrait and implementation method | |
| CN116248389A (en) | Data supervision platform based on enterprise data analysis | |
| CN114971583A (en) | Digital modeling system for complex regulation and control rule of power grid | |
| CN112711493A (en) | Scenario root cause analysis application | |
| CN107548087A (en) | A kind of method and device of warning association analysis | |
| CN108073582A (en) | A kind of Computational frame selection method and device | |
| CN106506190A (en) | A kind of communication network runtime and operation method | |
| CN110582091B (en) | Method and apparatus for locating wireless quality problems | |
| CN111476547A (en) | Method for analyzing assets of distribution network line ring-out point arrangement in user | |
| CN106294750A (en) | ICD model file familial defect upgrade-system and method | |
| CN110941910A (en) | Intelligent auxiliary method and system for power grid three-dimensional design review | |
| CN114866396B (en) | Method for realizing network fault location under inaccurate resources based on text similarity | |
| CN104244293B (en) | Network node alarm method and system | |
| CN116151433A (en) | Intelligent operation risk management and control method and device based on key process model | |
| CN114298467A (en) | Intelligent supervision system and method for civil aviation operation management automation system |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181016 |
|
| RJ01 | Rejection of invention patent application after publication |