CN105759784B - A kind of method for diagnosing faults based on DEA - Google Patents
A kind of method for diagnosing faults based on DEA Download PDFInfo
- Publication number
- CN105759784B CN105759784B CN201610080503.3A CN201610080503A CN105759784B CN 105759784 B CN105759784 B CN 105759784B CN 201610080503 A CN201610080503 A CN 201610080503A CN 105759784 B CN105759784 B CN 105759784B
- Authority
- CN
- China
- Prior art keywords
- test data
- envelope
- data
- multiple groups
- original test
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0224—Process history based detection method, e.g. whereby history implies the availability of large amounts of data
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/24—Pc safety
- G05B2219/24048—Remote test, monitoring, diagnostic
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Test And Diagnosis Of Digital Computers (AREA)
Abstract
The present invention relates to a kind of method for diagnosing faults based on DEA, it is characterised in that includes the following steps: typing multiple groups original test data;Time unifying processing is carried out to multiple groups original test data, generates the upper and lower line of envelope of multiple groups original test data;Input needs the test data diagnosed;Test data is judged whether in the envelope region of the upper and lower line of envelope, if being determined as qualification in envelope region, if there is the data being located at outside envelope region, is then determined as failure.A large amount of historical datas that the present invention is accumulated according to carrier rocket carry out the historical data Envelope Analysis of system, efficiently solve the modeling bottleneck of conventional failure method for diagnosing faults, solve the problems, such as conventional failure tree modeling complexity and heavy workload.
Description
Technical field
The present invention relates to a kind of method for diagnosing faults based on DEA, belong to fault diagnosis field.
Background technique
Carrier rocket needs to carry out status monitoring to it using fault diagnosis system in test and emission process, tradition
Method for diagnosing faults be using the diagnostic method based on fault tree mostly.On the basis of establishing object outages tree-model,
When object actual motion, fault tree is searched using failure method for searching, and complete fault diagnosis.
Common failure method for searching has logic method (using method for searching from top to down, from fault tree top event
Start, first test initial intermediate event, test next stage intermediate event is removed further according to the test result of intermediate event, until surveying
Try bottom event, complete fault diagnosis) with Minimal Cut Set (minimal cut corresponding with fault mode in test failure tree one by one
Collection completes fault diagnosis) two kinds.
It is intuitive simple that fault tree diagnosis has the advantages that diagnostic mode, but building correct comprehensive fault tree is fault tree
The core and key of diagnosis, and fault tree models the bottleneck of exactly fault diagnosis, it usually needs very comprehensive fault tree
Analysis could establish the fault tree of system perfecting, and workload is huge, needs to consume a large amount of manpower and material resources, and once event
Barrier tree establish it is incorrect or not comprehensively, then fault tree diagnosis will largely fail.
Summary of the invention
It is an object of the invention to overcome the deficiencies of the prior art and provide a kind of fault diagnosises based on DEA
Method breaches the modeling bottleneck of conventional fault diagnosis method, can be with automatically generated data packet by DATA ENVELOPMENT ANALYSIS METHOD
Winding thread carries out the real-time comparison of test data, completes fault parameter detection and fault location function.
The object of the invention is achieved by following technical solution:
There is provided a kind of method for diagnosing faults based on DEA, it is characterised in that include the following steps:
(1) typing multiple groups original test data;
(2) time unifying processing is carried out to multiple groups original test data, the envelope for generating multiple groups original test data is upper and lower
Line;
(3) input needs the test data diagnosed;
(4) test data is judged whether in the envelope region of the upper and lower line of envelope, if determined in envelope region
Then it is determined as failure if there is the data being located at outside envelope region for qualification.
Preferably, the upper and lower line method of envelope of multiple groups original test data is generated in the step (2) are as follows: for any
Moment t takes maximum value in multiple groups original test data, and it is online to form envelope;Minimum value in multiple groups original test data is taken, is formed
Envelope is offline;
Preferably, typing multiple groups original test data in step (1), and establish the database instance of MySQL database;?
Data in MySQL database example are imported into Hive tool;When realizing that multiple groups original test data carries out by Hive tool
Between registration process and generate multiple groups original test data the upper and lower line of envelope.
Preferably, importing task is executed using the open source component sqoop in Hive.
Preferably, time unifying processing is carried out to multiple groups original test data in the step (2), generates the original survey of multiple groups
Try the upper and lower line of envelope of data method particularly includes: the sampling time is divided into equal section of N number of time, takes each section
Multiple groups original test data maximum value, carry out straight line fitting formed envelope it is online;Take the original test of the multiple groups in each section
It is offline to carry out straight line fitting formation envelope for minimum value in data.
Preferably, time unifying processing is carried out to multiple groups original test data in the step (2), generates the original survey of multiple groups
Try the upper and lower line of envelope of data method particularly includes: be fitted respectively to multiple groups original test data and obtain a plurality of fitting song
Line;It selects one group of sampled data as benchmark, calculates each sampling instant of this group of sampled data and correspond in a plurality of matched curve respectively
The maximum value of point, is fitted, and it is online to form envelope;It calculates each sampling instant of this group of sampled data and corresponds to a plurality of matched curve
The minimum value of upper each point, is fitted, and it is offline to form envelope.
Preferably, the multiple groups original test data of typing selects the original test data in 1 year in step (1).
Preferably, if it is decided that then further include step (5) for qualification, the test data for needing to diagnose is inputted into MySQL number
According in library.
Preferably, multiple groups original test data is greater than 5 groups.
The invention has the following advantages over the prior art:
(1) the present invention is based on a large amount of history numbers that the method for diagnosing faults of DEA is accumulated according to carrier rocket
According to carrying out the historical data Envelope Analysis of system, efficiently solve the modeling bottleneck of conventional failure method for diagnosing faults, solve
The problem of conventional failure tree modeling complexity and heavy workload;Conventional failure method for diagnosing faults depends on expertise, needs to protect
Demonstrate,prove the integrality of fault tree, method for diagnosing faults of the invention, the historical test data based on conformity testing, without establishing failure
Tree, reduces the difficulty of fault diagnosis;
(2) the present invention is based on multiple groups historical test datas to be analyzed, and probability of miscarriage of justice is low.
(3) present invention carries out alignment of data processing, can truly reflect the fluctuation of initial data, improve on envelope,
Offline precision.
(4) present invention realizes that multiple groups original test data carries out time unifying processing and generation envelope using Hive tool
Upper and lower line, analysis speed is fast, and processing capability in real time is strong.
(5) present invention realizes the intelligence of fault diagnosis using envelope region intelligent alarm is exceeded.
Detailed description of the invention
Fig. 1 is that Fig. 1 overall procedure of the present invention plans schematic diagram;
Fig. 2 is that Fig. 2 data prediction of the present invention is aligned schematic diagram;
Fig. 3 is that the present invention is based on the fault diagnosis schematic diagrames of envelope;
Fig. 4 is that the present invention sends out time A, B data schematic diagram;
Fig. 5 is Envelope Analysis of embodiment of the present invention schematic diagram.
Specific embodiment
1, flow of task is planned
Overall workflow is parsed text and led first as shown in Figure 1, initial data store in the form of DAT file
In the MySQL database entered, it is transferred to Hive database, is analyzed by data prediction and data, excavates and believes in historical data
Breath, is stored in HBase database, and web server inquires HBase, returns to the data of needs.
2, typing initial data
Initial data is organized according to the hierarchical structure of model, hair, parameter, and each DAT file usually has 2-3 column,
Thousands of rows are differed between hundreds of thousands row, are imported into MySQL by Python program batch.
3, data are shifted to Hive
Hive is a kind of Tool for Data Warehouse based on Hadoop, the data file of structuring can be mapped as a number
According to library table, and provide SQL query function.Program work is excavated for the ease of big data, MySQL data are imported into Hive,
Importing task is executed parallel using open source component sqoop.
4, data prediction and Envelope Analysis
What the data sampling time of the same model difference hair time was not perfectly matched to, for subsequent Envelope Analysis, need
It is foundation time close alignment data processing, while guarantees the authenticity of data, is illustrated in figure 2 section alignment algorithm.
DATA ENVELOPMENT ANALYSIS METHOD is a kind of anomaly parameter diagnostic method based on historical data, by going through to multiple rocket
Relative time is calculated in the historical data of synchronization in history test data, obtains the envelope upper limit and envelope lower limit, finally,
All envelope upper limit moment points are linked to be curve, the envelope upper limit is generated, all envelope lower limit moment points is linked to be curve, it is raw
At envelope lower limit, envelope domain is ultimately generated, when carrying out the fault diagnosis based on DEA, is surveyed by test data
Value is compared with envelope domain, and interpretation test parameter is with the presence or absence of abnormal.Envelope Analysis chooses the ginseng of history similar in multiple task
Number data, determine history envelope of the parameter under the task by the maximum value and minimum value of finding out each moment, using big
Data platform can obtain the parameters history envelope for reliably reflecting truth, with a large amount of historical datas of comprehensive analysis convenient in the future
The abnormal conditions of timely monitoring parameter.
But due to the possible different from of each hair subtask, the same parameter may present different under different hairs time
The characteristic of sample will lead to envelope and degenerate if arbitrarily selection hair time generates envelope, wider between bound, it is difficult to reflection tool
The feature of body parameter, as shown in Figure 3.
Therefore, close hair should be selected when generating envelope, for example, by using the data in 1 year, makes every effort to accomplish envelope both
It can accomplish to reflect true common scenario, and will not be as the feature of degeneration itself.By calculating envelope contribution degree and abnormal packet
Network rejects incoherent hair.
The mode of the first alignment are as follows: the sampling time is divided into equal section of N number of time, takes the multiple groups in each section
It is online to carry out straight line fitting formation envelope for the maximum value of original test data;It takes in the multiple groups original test data in each section
Minimum value, carry out straight line fitting formed envelope it is offline.
The mode of second of alignment are as follows: multiple groups original test data is fitted respectively and obtains a plurality of matched curve;Choosing
One group of sampled data is selected as benchmark, each sampling instant of this group of sampled data is calculated and corresponds in a plurality of matched curve each point most
Big value, is fitted, and it is online to form envelope;It calculates each sampling instant of this group of sampled data and corresponds to each point in a plurality of matched curve
Minimum value, be fitted, formed envelope it is offline.
5, front end is shown
It is upper, envelope is offline and diagnostic message is shown, it is checked in order to facilitate data analysis result, using B/S mode
Show result.Using uWSGI as web server, the request of browser is forwarded in the webService that Python writes,
Using Happybase module accesses HBase database, query result returns to browser via uWSGI.
DATA ENVELOPMENT ANALYSIS METHOD is realized with program, plug-in unit is generated, is integrated into big data platform, ultimately generates and be based on
The fault diagnosis system of DEA.
6, embodiment
By big data platform, flying quality is entered into big data platform, Envelope Analysis program is added to big data
Platform forms stronger analysis ability, is verified by certain model data, and the known time A that sends out is there are abnormal conditions referring to fig. 4,
Using the hair time A as verify data, using in addition data compare verification platform point with conventional analysis as historical basis data several times
Function is analysed, test parameter chooses certain temperature (XX) in existing analysis report.
The Envelope Analysis of parameter XX is as shown in Figure 5.In conventional analysis figure as can be seen that in 1240s or so, parameter XX with
B hair time is abnormal compared to occurring.In Envelope Analysis figure as can be seen that in 580s or so, sends out time A parameter and slight cross the border occur
There is serious cross-border phenomenon in 1240s or so in phenomenon.
Can be seen that conventional method of analysis by the comparative analysis of parameter XX could not judge the failure of 500s or so, only
Breakdown judge can be provided after severely subnormal occurs in 1240s.Envelope Analysis Method can judge the failure of 500s or so.
The present invention is to realize that the effective means of real-time online fault detection and positioning, while pole are tested and emitted to carrier rocket
Application of the fault diagnosis system in carrier rocket industry has been pushed greatly, has been had broad application prospects latent with huge market
Power.
The above, optimal specific embodiment only of the invention, but scope of protection of the present invention is not limited thereto,
In the technical scope disclosed by the present invention, any changes or substitutions that can be easily thought of by anyone skilled in the art,
It should be covered by the protection scope of the present invention.
The content that description in the present invention is not described in detail belongs to the well-known technique of professional and technical personnel in the field.
Claims (6)
1. a kind of method for diagnosing faults of the DEA based on carrier rocket historical data, it is characterised in that including as follows
Step:
(1) typing multiple groups carrier rocket original test data;
(2) time unifying processing is carried out to multiple groups original test data, generates the upper and lower line of envelope of multiple groups original test data;
(3) input needs the test data diagnosed;
(4) test data is judged whether in the envelope region of the upper and lower line of envelope, if being judged to closing in envelope region
Lattice are then determined as failure if there is the data being located at outside envelope region;
Time unifying processing is carried out to multiple groups original test data in the step (2), generates the packet of multiple groups original test data
The upper and lower line of network method particularly includes: acquisition is fitted respectively to every group of original test data in multiple groups original test data
A plurality of matched curve;It selects one group of sampled data as benchmark, calculates each sampling instant of this group of sampled data and correspond to a plurality of intend
The maximum value for closing each point on curve, is fitted, and it is online to form envelope;It is more to calculate each sampling instant correspondence of this group of sampled data
The minimum value of each point, is fitted in matched curve, and it is offline to form envelope.
2. method according to claim 1, it is characterised in that: typing multiple groups original test data in step (1), and establish
The database instance of MySQL database;Data in MySQL database example are imported into Hive tool;In Hive tool
It is middle to realize that multiple groups original test data carries out time unifying processing and generates the upper and lower line of envelope of multiple groups original test data.
3. method according to claim 2, it is characterised in that: execute importing task using the open source component sqoop in Hive.
4. method according to claim 1, it is characterised in that: the multiple groups original test data selection one of typing in step (1)
Original test data in year.
5. method according to claim 4, it is characterised in that: if it is determined that qualified, then further include step (5), will need
In the test data input MySQL database of diagnosis.
6. method according to claim 1, it is characterised in that: multiple groups original test data is greater than 5 groups.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610080503.3A CN105759784B (en) | 2016-02-04 | 2016-02-04 | A kind of method for diagnosing faults based on DEA |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610080503.3A CN105759784B (en) | 2016-02-04 | 2016-02-04 | A kind of method for diagnosing faults based on DEA |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105759784A CN105759784A (en) | 2016-07-13 |
CN105759784B true CN105759784B (en) | 2019-04-09 |
Family
ID=56330599
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610080503.3A Active CN105759784B (en) | 2016-02-04 | 2016-02-04 | A kind of method for diagnosing faults based on DEA |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105759784B (en) |
Families Citing this family (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20180107202A1 (en) * | 2016-10-18 | 2018-04-19 | Micron Technology, Inc. | System and method for detecting fault events |
CN107229234A (en) * | 2017-05-23 | 2017-10-03 | 深圳大学 | The distributed libray system and method for Aviation electronic data |
CN107545145B (en) * | 2017-09-08 | 2020-04-07 | 国网湖南省电力有限公司 | Power grid forest fire disaster danger degree super-efficiency envelope analysis method and system |
CN109597399B (en) * | 2018-11-28 | 2020-09-18 | 北京宇航系统工程研究所 | Information control platform for informatization rocket launching |
CN110187631B (en) * | 2019-06-25 | 2021-04-13 | 北京临近空间飞行器系统工程研究所 | Time alignment method and system for control system |
CN110646212B (en) * | 2019-10-23 | 2022-01-25 | 成都飞机工业(集团)有限责任公司 | Novel method for calibrating aircraft engine |
CN110851497A (en) * | 2019-11-01 | 2020-02-28 | 唐山钢铁集团有限责任公司 | Method for detecting whether converter oxygen blowing is not ignited |
CN111486920B (en) * | 2020-04-15 | 2022-06-14 | 上海航天精密机械研究所 | Method, system and medium for judging and analyzing volume measurement data of carrier rocket storage tank |
CN111770002B (en) * | 2020-06-12 | 2022-02-25 | 南京领行科技股份有限公司 | Test data forwarding control method and device, readable storage medium and electronic equipment |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3103193B2 (en) * | 1992-04-15 | 2000-10-23 | 東京電力株式会社 | Diagnostic equipment for rotating machinery |
CN1834607A (en) * | 2005-03-16 | 2006-09-20 | 欧姆龙株式会社 | Inspection method and inspection apparatus |
CN103269345A (en) * | 2013-05-30 | 2013-08-28 | 沈阳师范大学 | Intelligent display device and method based on Modbus protocol |
CN103868694A (en) * | 2014-03-26 | 2014-06-18 | 东南大学 | Embedded variable-rotation-speed bearing fault diagnosis device |
CN105092239A (en) * | 2014-05-09 | 2015-11-25 | 潍坊学院 | Method for detecting early stage fault of gear |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP5395041B2 (en) * | 2010-11-30 | 2014-01-22 | トヨタ自動車株式会社 | Vehicle, internal combustion engine abnormality determination method, and internal combustion engine abnormality determination device |
CN203350691U (en) * | 2013-05-30 | 2013-12-18 | 沈阳师范大学 | Intelligent display device based on Modbus protocol |
CN103698637B (en) * | 2013-12-25 | 2016-05-18 | 云南电力调度控制中心 | A kind of electric power critical Indexes Abnormality method for quick and device |
-
2016
- 2016-02-04 CN CN201610080503.3A patent/CN105759784B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3103193B2 (en) * | 1992-04-15 | 2000-10-23 | 東京電力株式会社 | Diagnostic equipment for rotating machinery |
CN1834607A (en) * | 2005-03-16 | 2006-09-20 | 欧姆龙株式会社 | Inspection method and inspection apparatus |
CN103269345A (en) * | 2013-05-30 | 2013-08-28 | 沈阳师范大学 | Intelligent display device and method based on Modbus protocol |
CN103868694A (en) * | 2014-03-26 | 2014-06-18 | 东南大学 | Embedded variable-rotation-speed bearing fault diagnosis device |
CN105092239A (en) * | 2014-05-09 | 2015-11-25 | 潍坊学院 | Method for detecting early stage fault of gear |
Also Published As
Publication number | Publication date |
---|---|
CN105759784A (en) | 2016-07-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105759784B (en) | A kind of method for diagnosing faults based on DEA | |
CN111985561A (en) | Fault diagnosis method and system for intelligent electric meter and electronic device | |
CN106777101B (en) | Data processing engine | |
CN101093559B (en) | Method for constructing expert system based on knowledge discovery | |
CN111459799A (en) | Software defect detection model establishing and detecting method and system based on Github | |
CN109204389A (en) | A kind of subway equipment fault diagnosis and self-healing method, system | |
CN106776208A (en) | Fault Locating Method during a kind of running software | |
CN116771576A (en) | Comprehensive fault diagnosis method for hydroelectric generating set | |
CN109213773A (en) | A kind of diagnostic method, device and the electronic equipment of online failure | |
CN109936479A (en) | Control plane failure diagnostic system and its implementation based on Differential Detection | |
CN108897686A (en) | It is complete to record separately automated testing method and device | |
CN116467674A (en) | Intelligent fault processing fusion updating system and method for power distribution network | |
CN109507992A (en) | A kind of failure prediction method, device and the equipment of locomotive braking system component | |
Lam | Formal analysis of BPMN models: a NuSMV-based approach | |
Karakostas et al. | Industrial data services for quality control in smart manufacturing–the i4q framework | |
Siddique et al. | Hybrid Framework To Exclude Similar and Faulty Test Cases In Regression Testing | |
CN117376087A (en) | Method, device, equipment and storage medium for delimiting network quality problems | |
Li et al. | Active learning empirical research on cross-version software defect prediction datasets | |
CN111143432A (en) | Data analysis early warning system and method for event processing result | |
CN112416761B (en) | Test case generation method and device based on breadth-first search | |
CN110008245B (en) | Method suitable for searching equipment fault early warning model time period | |
Li et al. | An improvement scheme for the overall line effectiveness of a production line: A case study | |
Ma et al. | Data management of salt cavern gas storage based on data model | |
Liu et al. | A Web Back-End Database Leakage Incident Reconstruction Framework Over Unlabeled Logs | |
Han et al. | Research of Fault Diagnosis System for Remote Sensing Satellite Receiving System Base on Fault Tree |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |