CN107544337A - A kind of method of system for monitoring loosening part trigger data intelligent classification - Google Patents
A kind of method of system for monitoring loosening part trigger data intelligent classification Download PDFInfo
- Publication number
- CN107544337A CN107544337A CN201710843795.6A CN201710843795A CN107544337A CN 107544337 A CN107544337 A CN 107544337A CN 201710843795 A CN201710843795 A CN 201710843795A CN 107544337 A CN107544337 A CN 107544337A
- Authority
- CN
- China
- Prior art keywords
- data
- passage
- loosening part
- trigger data
- monitoring
- 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
Landscapes
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention discloses a kind of method of system for monitoring loosening part trigger data intelligent classification, methods described includes：Obtain the original trigger data of system for monitoring loosening part；Based on original trigger data, data corresponding to each passage are obtained；Waveshape feature abstraction is carried out to the data of each passage, obtains the characteristic vector of each passage；Establish single channel disaggregated model, characteristic vector based on single channel disaggregated model and each passage, the data classification results of each passage are obtained, real loosening part is hit into making a distinction, count and managing for signal data and false triggering data intelligence, the efficiency of lifting loosening part data analysis.
Description
Technical field
The present invention relates to reactor and primary Ioops loosening part monitoring field, in particular it relates to which a kind of loosening part monitors
The method of system trigger data intelligence classification.
Background technology
Include in-pile component and fuel assembly inside nuclear reactor pressure container, in-pile component is mainly by holding down assembly, hanging
Basket assembly, reactor core lower support component etc. are formed, and the part and part in each component are most of to be connected by screw, pin etc..
Heap under continuous service condition, because of the influence of water impact and Flow vibration, may produce some parts and loosen even for a long time
Come off, so as to form loose piece in primary Ioops.And when constructing, reloading or repairing, also have and leave gold in primary Ioops system
Belong to the possibility of part (claiming outer parcel received).If these loosening parts find and disposed not in time, the lighter causes huge economic loss,
Severe one will have a strong impact on that nuclear plant safety is run, and consequence is hardly imaginable.Therefore, domestic all nuclear power stations have been respectively mounted loosening part
Monitoring system, for on-line real time monitoring reactor and the loosening part of primary Ioops.
The principle of nuclear power plant reactor and primary Ioops system for monitoring loosening part is to work as reactor, main pipeline, main pump, steam
When the loosening element of the travelling part and equipment in cooling agent occurs in generator, loose piece will touch with equipment wall or internals
Hit, produce " bending wave " signal, signal is propagated along wall and tube wall, detected using the accelerometer on wall this
Acoustic signals, and signal is handled and analyzed, thus judge that loosening part whether there is.Due to reactor and primary Ioops system
Interference and unit peripheral operation (such as CRDM motion, peripheral operation or valve event, noise level liter be present in system
High, System self-test etc.), system " false triggering " may be caused, produce substantial amounts of trigger data, so as to generate to come two and ask
Topic：
First, real loosening part, which hits data, may be submerged in these " false triggering " data, have impact on and release
The analysis promptness and accuracy of part；
Second, system operation maintenance personnel is big to the workload of daily loosening part system trigger data analysis and mechanical.
In summary, present inventor has found above-mentioned technology extremely during the present application technical scheme is realized
Following technical problem less be present：
In the prior art, existing system for monitoring loosening part, which exists, easily causes false triggering, analyzes promptness and standard
True property is poor, the relatively low technical problem of trigger data analysis efficiency.
The content of the invention
The invention provides a kind of method of system for monitoring loosening part trigger data intelligent classification, will interference or system from
Data automatic rejection caused by inspection, to improve the analysis efficiency and intelligent level of loosening part system trigger signal data.
For achieving the above object, this application provides a kind of system for monitoring loosening part trigger data intelligent classification
Method, methods described include：
Obtain the original trigger data of system for monitoring loosening part；
Based on original trigger data, data corresponding to each passage are obtained；
Waveshape feature abstraction is carried out to the data of each passage, obtains the characteristic vector of each passage；
Single channel disaggregated model is established, the characteristic vector based on single channel disaggregated model and each passage, is obtained each logical
The data classification results in road.
The present invention is directed to the system for monitoring loosening part trigger data of magnanimity, extracts effective characteristic quantity, utilizes engineering
The method of habit establishes system for monitoring loosening part trigger data intelligent classification model, and real loosening part is hit into signal data
Make a distinction, count and manage with " false triggering " data intelligence, lift the efficiency of loosening part data analysis.
It can be implemented function such as by this method：
(1) intelligent classification of system for monitoring loosening part trigger data；
(2) statistic of classification of system for monitoring loosening part trigger data, display and management；
(3) intelligent classification software can open multiple program process simultaneously on the same host, and not mutual between process
Influence.
The complete Wave data of system for monitoring loosening part trigger data one (deposit by system for monitoring loosening part trigger data
Storage form is binary d at patterns file) it is made up of 13 channel datas, each passage includes 5000 data points.Therefore, it is right
The prediction of whole Wave data is dependent on the prediction to each single-channel data.Therefore, using the method for machine learning, for list
The different wave data of individual passage are learnt, train classification models.Finally each passage using the model to Wave data
Prediction result merge, formed finally to the prediction result of complete waveform data.
System for monitoring loosening part trigger data Characteristic Extraction：
When studying complex object, primary object can be typically mapped in characteristic vector space by feature extraction.It is special
Each dimension of sign vector corresponds to a feature of primary object.Primary object is mapped into characteristic vector space two
Key benefit：
1) by feature extraction, the dimension of primary object can be effectively reduced.System for monitoring loosening part trigger data
One raw channel data has 5000 points, and data dimension can be reduced to 46 dimensions by feature extraction.The data of low dimensional
Machine learning algorithm can be allowed more easily to be trained, so as to produce higher-quality model.
2) preparatory condition or data easily can be dissolved into machine learning algorithm by feature extraction.Feature is often skill
Definition of the art personnel to research object key element, plays important role in machine learning task.Utilize technical staff
The default high-quality characteristics provided, machine learning algorithm can obtain more preferable result.
3) model for training to obtain by characteristic has higher interpretation.
System for monitoring loosening part trigger data model of mind is established：
The model that can be classified to single channel Wave data is produced by machine learning algorithm.The present invention is using random
Forest (Random Forest) algorithm carries out disaggregated model training.Random forests algorithm is generated using integrated study technological learning
More decision trees, and the prediction result of all decision trees is merged, obtain result.Expertise is converted into logic rule
Then, merged with single pass prediction result, ultimately form the intelligent classification model of system for monitoring loosening part trigger data.
Performance indications are as shown in table 1：
Table 1
One or more technical schemes that the application provides, have at least the following technical effects or advantages：
By combining art technology knowledge base information, 49 have been designed and Implemented based on the waveform of time domain and frequency domain spy
Sign.These wave characters are respectively provided with height interpretation, significant for checking, interpretation model effect；
Distinctive random forests algorithm in the feature of waveform and machine learning is effectively combined, it is different in waveform
Feature is showed by different decision situations, and integration technology knowledge base information, obtains disaggregated model to the end；
Model is subjected to software implementation, and develops human-computer interaction interface, realizes intelligent classification, the number of loosening part trigger data
According to management and display.
Brief description of the drawings
Accompanying drawing described herein is used for providing further understanding the embodiment of the present invention, forms one of the application
Point, do not form the restriction to the embodiment of the present invention；
The loosening part trigger data single channel waveform recognition model training process of accompanying drawing 1；
The loosening part trigger data assorting process of accompanying drawing 2；
The data of accompanying drawing 3 classification software flow.
Embodiment
The invention provides a kind of method of system for monitoring loosening part trigger data intelligent classification, will interference or system from
Data automatic rejection caused by inspection, to improve the analysis efficiency and intelligent level of loosening part system trigger signal data.
It is below in conjunction with the accompanying drawings and specific real in order to be more clearly understood that the above objects, features and advantages of the present invention
Mode is applied the present invention is further described in detail.It should be noted that in the case where not conflicting mutually, the application's
Feature in embodiment and embodiment can be mutually combined.
Many details are elaborated in the following description to facilitate a thorough understanding of the present invention, still, the present invention may be used also
To be implemented using other different from the other modes in the range of being described herein, therefore, protection scope of the present invention is not by under
The limitation of specific embodiment disclosed in face.
Whole Algorithm for Training flow is as shown in figure 1, model application flow is as shown in Figure 2.
Effective channel data extraction：
System for monitoring loosening part trigger data classification have control rod hit signal, Typical Components come off hit signal,
System self-test signal, pulse spike, water conservancy fluctuation and external disturbance totally six class signal.The mark of these data is also with whole ripple
Graphic data is what unit was labeled, for example, a loosening part trigger data is made up of 13 passages, each 5000, passage
Data point, it is water conservancy fluctuation signal that a Wave data for being noted as water conservancy fluctuation, which may only have 1 passage, and other 12
Passage is normal noise signal, so as to which the channel data for really being able to represent Wave data classification is called effective port number
According to.
Feature extraction：
Loosening part trigger data is mapped in characteristic vector space by feature extraction.Each dimension of characteristic vector
A feature of the degree corresponding to primary object.Primary object is mapped into characteristic vector space three key benefits：
By feature extraction, the dimension of primary object can be effectively reduced.For example, a raw channel data has 5000
It is individual, data dimension can be reduced to by feature extraction by 49 dimensions.The data of low dimensional can allow machine learning algorithm more to hold
Change places and be trained, so as to produce higher-quality model.
Art technology knowledge base information easily can be dissolved into machine learning algorithm by feature extraction.Feature is usually
It is definition of the those skilled in the art to research object key element, important role is play in machine learning task.Profit
The high-quality characteristics provided with art technology knowledge base information, machine learning algorithm can obtain more preferable result.
The model for training to obtain by characteristic has higher interpretation.
To loosening part trigger signal, based on expertise, the peak-to-peak amplitude of the basic passage of extraction triggering section response, on
Rise time, time delay, dominant frequency, the peak-to-peak amplitude of minimum time difference passage and time small time difference passage, the rise time, time delay, when
The affair character amount such as difference.For the signature analysis of waveform itself, extract some and do not possess loosening part shock physical significance, but energy
The characteristic quantity of effective district partial wave shape, because each feature corresponds to a time domain and a frequency domain character, all feature quantities respectively
For 49
Model training：
The single channel Wave data disaggregated model of loosening part trigger data can be established by being produced by machine learning algorithm.
Art technology knowledge base information is then based on, single pass prediction result is merged, is formed finally to original waveform number
According to prediction result.
The present invention carries out model training using random forest (Random Forest) algorithm.Random forest is learned using integrated
Practise technological learning and generate more decision trees, and the prediction result of all decision trees is merged, form final result.Machine forest
Algorithm trained n decision tree altogether, and then last sample Instance output result is to all n decision trees pair
Instance result carries out what ballot was drawn.It is noted herein that random forests algorithm is when training each decision tree
Use and be characterized in different (quantity may also be different).For example, when training first decision tree, what is used is probably
Preceding 10 features, and train second decision tree use be probably the 9 to 15th feature.
By above-mentioned model, the intelligent classification of loosening part trigger data single-channel data is realized, for 13 passages
Data, combine expertise and single-channel data result judged, for being not belonging to 6 style numbers described above, will
The dtd--data type definition is other, finally realizes the final classification of data.
Fig. 3 is refer to, software is realized：
Waveform intelligent classification is one and exempts from installation, cross-platform, without network, with graphic user interface (GUI) desktop
Application program.It provides a user a friendly interactive interface, for aiding in carrying out the identification of high-volume waveform and tying in real time
Fruit counts.
The application is based on advanced desktop application constructing technology, using single programming language, open source technology framework, with reference to stream
Capable WEB technologies, shield the difference between different operating system, it is possible to achieve first encoding, multi-platform operation.
After user decompresses corresponding platform and the applied compression bag of CPU architecture, double-click executable program and open Graphics Application circle
Face.In the interface, user needs to configure two parameters：Wave data file and prediction result outgoing route to be predicted.
Program reads Wave data file to be predicted, the lower All Files of display this document folder and one-level subfile automatically
Title is pressed from both sides, and counts their quantity.User can also point to mouse some Fileview its creation time, modification time and text
Part size.
User's click starts after predicting button, two input parameters of program ver-ify, then calls Java CLI and incoming use
Family input paramete information, while built-in model file path is provided, prediction process transfers to java applet to handle.
After Java receives prediction instruction, Waveform forecast is proceeded by, prediction result is constantly produced in outgoing route.Program
Start timer, poll prediction result, and real-time display waveform classification, quantity, running status, beginning/end time, identified
Wave number amount, the time-consuming information of average identification.
Because reactor and primary Ioops system have interference and unit peripheral operation, system for monitoring loosening part can be caused
" false triggering ", the method for utilizing machine learning using the trigger data for magnanimity, the present invention, by trigger data waveform
Analysis, is extracted effective characteristic quantity, employs random forests algorithm, by constantly training and combining expertise, establishes
System for monitoring loosening part trigger data intelligent classification model, by the prototype software, effectively realizes loosening part and hits
Differentiation, statistics and the management of signal data and " false triggering " data intelligence are hit, lifts the efficiency of loosening part data analysis.
Although preferred embodiments of the present invention have been described, but those skilled in the art once know basic creation
Property concept, then can make other change and modification to these embodiments.So appended claims be intended to be construed to include it is excellent
Select embodiment and fall into having altered and changing for the scope of the invention.
Obviously, those skilled in the art can carry out the essence of various changes and modification without departing from the present invention to the present invention
God and scope.So, if these modifications and variations of the present invention belong to the scope of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to comprising including these changes and modification.
Claims (8)
- A kind of 1. method of system for monitoring loosening part trigger data intelligent classification, it is characterised in that methods described includes：Obtain the original trigger data of system for monitoring loosening part；Based on original trigger data, data corresponding to each passage are obtained；Waveshape feature abstraction is carried out to the data of each passage, obtains the characteristic vector of each passage；Single channel disaggregated model is established, the characteristic vector based on single channel disaggregated model and each passage, obtains each passage Data classification results.
- 2. the method for system for monitoring loosening part trigger data intelligent classification according to claim 1, it is characterised in that institute Stating method also includes：The data classification results of each passage are merged, form the prediction classification results of original trigger data.
- 3. the method for system for monitoring loosening part trigger data intelligent classification according to claim 1, it is characterised in that logical Cross waveshape feature abstraction to map to original each channel data in characteristic vector space, each dimension of characteristic vector is corresponding In a feature of original each channel data.
- 4. the method for system for monitoring loosening part trigger data intelligent classification according to claim 1, it is characterised in that pine The classification of the original trigger data of de- part monitoring system includes：Control rod, which hits signal, Typical Components come off hits signal, is System self-test signal, pulse spike, water conservancy fluctuation signal, external interference signals.
- 5. the method for system for monitoring loosening part trigger data intelligent classification according to claim 1, it is characterised in that right In loosening part trigger signal, extraction triggers peak-to-peak amplitude, rise time, time delay, the dominant frequency of the basic passage of section response, And peak-to-peak amplitude, rise time, time delay, the time difference affair character amount of minimum time difference passage and time small time difference passage.
- 6. the method for system for monitoring loosening part trigger data intelligent classification according to claim 1, it is characterised in that logical Cross machine learning algorithm and produce the single channel disaggregated model that can establish loosening part trigger data.
- 7. the method for system for monitoring loosening part trigger data intelligent classification according to claim 6, it is characterised in that adopt Model training is carried out with random forests algorithm；Random forest learns to generate more decision trees using integrated study mode, and by institute The prediction result for having decision tree is merged, and forms final result.
- 8. the method for system for monitoring loosening part trigger data intelligent classification according to claim 1, it is characterised in that pine The original trigger data of de- part monitoring system is that Wave data is made up of 13 channel datas, and each passage includes 5000 data Point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710843795.6A CN107544337B (en) | 2017-09-19 | 2017-09-19 | Method for intelligently classifying triggering data of loose part monitoring system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710843795.6A CN107544337B (en) | 2017-09-19 | 2017-09-19 | Method for intelligently classifying triggering data of loose part monitoring system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107544337A true CN107544337A (en) | 2018-01-05 |
| CN107544337B CN107544337B (en) | 2020-10-30 |
Family
ID=60963985
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710843795.6A Active CN107544337B (en) | 2017-09-19 | 2017-09-19 | Method for intelligently classifying triggering data of loose part monitoring system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107544337B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109543093A (en) * | 2018-09-29 | 2019-03-29 | 中国电子科技集团公司电子科学研究院 | A kind of integral intelligent information assistant's platform |
| CN109543894A (en) * | 2018-11-15 | 2019-03-29 | 深圳中广核工程设计有限公司 | A kind of nuclear power station loosening part ex ante forecasting system and prediction technique |
Citations (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS644610B2 (en) * | 1980-10-08 | 1989-01-26 | Chugoku Denryoku Kk | |
| JPH01321397A (en) * | 1988-06-23 | 1989-12-27 | Babcock Hitachi Kk | Monitor for loose parts |
| CN101685679A (en) * | 2008-09-27 | 2010-03-31 | 中国核动力研究设计院 | System for monitoring loosening part of nuclear reactor and coolant system |
| CN101718581A (en) * | 2009-11-13 | 2010-06-02 | 浙江大学 | Alarming method of nuclear power station loose part based on support vector machine |
| CN101750197A (en) * | 2008-12-11 | 2010-06-23 | 中国核动力研究设计院 | Loose part diagnosis method for nuclear power station |
| JP2012057690A (en) * | 2010-09-07 | 2012-03-22 | Japan Steel Works Ltd:The | Bolt tensioner and bolt tensioner adjusting device |
| CN103093840A (en) * | 2013-01-25 | 2013-05-08 | 杭州电子科技大学 | Reactor loose part alarm method |
| CN103137224A (en) * | 2013-01-25 | 2013-06-05 | 杭州电子科技大学 | Nuclear power station loose part quality estimation method based on wavelet energy spectrum |
| CN103137225A (en) * | 2013-01-25 | 2013-06-05 | 杭州电子科技大学 | Nuclear power station loose part positioning method based on wavelet transform and hilbert transform |
| CN104376881A (en) * | 2014-11-20 | 2015-02-25 | 西安交通大学 | Nuclear power plant loose part positioning method based on Hilbert transform and data screening |
| CN104408303A (en) * | 2014-11-20 | 2015-03-11 | 西安交通大学 | Laser pulse mass spectrometry (LPMS) mass estimating method based on data matching |
| CN204886920U (en) * | 2015-08-22 | 2015-12-16 | 上海核工程研究设计院 | Not hard up part of nuclear power station signal conditioning circuit for monitoring system |
| CN105225433A (en) * | 2015-08-22 | 2016-01-06 | 上海核工程研究设计院 | A kind of nuclear power station loose parts monitor and alarm system and method |
| CN105427904A (en) * | 2015-11-13 | 2016-03-23 | 陕西卫峰核电子有限公司 | Loose part positioning method based on nuclear power station |
| JP2016057206A (en) * | 2014-09-10 | 2016-04-21 | 三菱重工業株式会社 | Loose parts monitoring device |
| CN105702305A (en) * | 2016-02-03 | 2016-06-22 | 杭州电子科技大学 | Multi-target three-dimensional dynamic tracking method of loose parts of nuclear power plant |
| CN205982317U (en) * | 2016-06-22 | 2017-02-22 | 江苏核电有限公司 | Nuclear power station pine takes off part acceleration sensor installation fixing device |
| CN106874896A (en) * | 2017-03-31 | 2017-06-20 | 阳江核电有限公司 | A kind of one loop of nuclear power station part releases the assisted learning method and system of diagnostic system signal characteristic identification |
-
2017
- 2017-09-19 CN CN201710843795.6A patent/CN107544337B/en active Active
Patent Citations (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS644610B2 (en) * | 1980-10-08 | 1989-01-26 | Chugoku Denryoku Kk | |
| JPH01321397A (en) * | 1988-06-23 | 1989-12-27 | Babcock Hitachi Kk | Monitor for loose parts |
| CN101685679A (en) * | 2008-09-27 | 2010-03-31 | 中国核动力研究设计院 | System for monitoring loosening part of nuclear reactor and coolant system |
| CN101750197A (en) * | 2008-12-11 | 2010-06-23 | 中国核动力研究设计院 | Loose part diagnosis method for nuclear power station |
| CN101718581A (en) * | 2009-11-13 | 2010-06-02 | 浙江大学 | Alarming method of nuclear power station loose part based on support vector machine |
| JP2012057690A (en) * | 2010-09-07 | 2012-03-22 | Japan Steel Works Ltd:The | Bolt tensioner and bolt tensioner adjusting device |
| CN103137225A (en) * | 2013-01-25 | 2013-06-05 | 杭州电子科技大学 | Nuclear power station loose part positioning method based on wavelet transform and hilbert transform |
| CN103137224A (en) * | 2013-01-25 | 2013-06-05 | 杭州电子科技大学 | Nuclear power station loose part quality estimation method based on wavelet energy spectrum |
| CN103093840A (en) * | 2013-01-25 | 2013-05-08 | 杭州电子科技大学 | Reactor loose part alarm method |
| JP2016057206A (en) * | 2014-09-10 | 2016-04-21 | 三菱重工業株式会社 | Loose parts monitoring device |
| CN104376881A (en) * | 2014-11-20 | 2015-02-25 | 西安交通大学 | Nuclear power plant loose part positioning method based on Hilbert transform and data screening |
| CN104408303A (en) * | 2014-11-20 | 2015-03-11 | 西安交通大学 | Laser pulse mass spectrometry (LPMS) mass estimating method based on data matching |
| CN204886920U (en) * | 2015-08-22 | 2015-12-16 | 上海核工程研究设计院 | Not hard up part of nuclear power station signal conditioning circuit for monitoring system |
| CN105225433A (en) * | 2015-08-22 | 2016-01-06 | 上海核工程研究设计院 | A kind of nuclear power station loose parts monitor and alarm system and method |
| CN105427904A (en) * | 2015-11-13 | 2016-03-23 | 陕西卫峰核电子有限公司 | Loose part positioning method based on nuclear power station |
| CN105702305A (en) * | 2016-02-03 | 2016-06-22 | 杭州电子科技大学 | Multi-target three-dimensional dynamic tracking method of loose parts of nuclear power plant |
| CN205982317U (en) * | 2016-06-22 | 2017-02-22 | 江苏核电有限公司 | Nuclear power station pine takes off part acceleration sensor installation fixing device |
| CN106874896A (en) * | 2017-03-31 | 2017-06-20 | 阳江核电有限公司 | A kind of one loop of nuclear power station part releases the assisted learning method and system of diagnostic system signal characteristic identification |
Non-Patent Citations (3)
| Title |
|---|
| 何元峰: "核电站松动部件的质量估计方法", 《知网》 * |
| 杨波: "核电厂松动部件冲击仿真及信噪分离技术研究", 《万方数据》 * |
| 郑华文: "基于支持向量机的核电站松动件质量估计方法", 《浙江大学学报》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109543093A (en) * | 2018-09-29 | 2019-03-29 | 中国电子科技集团公司电子科学研究院 | A kind of integral intelligent information assistant's platform |
| CN109543894A (en) * | 2018-11-15 | 2019-03-29 | 深圳中广核工程设计有限公司 | A kind of nuclear power station loosening part ex ante forecasting system and prediction technique |
| CN109543894B (en) * | 2018-11-15 | 2020-11-24 | 深圳中广核工程设计有限公司 | System and method for predicting loose parts of nuclear power station in advance |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107544337B (en) | 2020-10-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107885999B (en) | Vulnerability detection method and system based on deep learning | |
| CN102629904B (en) | Detection and determination method of network navy | |
| CN109753801B (en) | Intelligent terminal malicious software dynamic detection method based on system call | |
| CN102779249B (en) | Malware detection methods and scanning engine | |
| CN110232280A (en) | A kind of software security flaw detection method based on tree construction convolutional neural networks | |
| CN107133518B (en) | Source code based on parameter and information flow is gone beyond one's commission detection method and device | |
| CN109462575A (en) | A kind of webshell detection method and device | |
| CN104112026A (en) | Short message text classifying method and system | |
| CN107391598A (en) | One kind threatens information automatic generation method and system | |
| CN106971254A (en) | A kind of service monitoring system and method | |
| CN114090402A (en) | User abnormal access behavior detection method based on isolated forest | |
| CN107544337A (en) | A kind of method of system for monitoring loosening part trigger data intelligent classification | |
| CN110263539A (en) | A kind of Android malicious application detection method and system based on concurrent integration study | |
| CN109697361A (en) | A kind of wooden horse classification method based on Trojan characteristics | |
| CN111598179A (en) | Power monitoring system user abnormal behavior analysis method, storage medium and equipment | |
| CN110866318A (en) | Dangerous rock instability mode discrimination method and device based on deep learning technology | |
| CN108763096A (en) | Software Defects Predict Methods based on depth belief network algorithm support vector machines | |
| CN112364631B (en) | Chinese grammar error detection method and system based on hierarchical multitask learning | |
| CN106681980B (en) | A kind of refuse messages analysis method and device | |
| CN114385775A (en) | Sensitive word recognition method based on big data | |
| CN111488501A (en) | E-commerce statistical system based on cloud platform | |
| CN101247279A (en) | Internet content safety detecting system | |
| CN110162472A (en) | A kind of method for generating test case based on fuzzing test | |
| CN111967003A (en) | Automatic wind control rule generation system and method based on black box model and decision tree | |
| CN115186095B (en) | Juvenile text recognition method and device |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |