CN110110363A - A kind of satellite momentum wheel fault diagnosis modeling method and system based on HC-TP-K2 algorithm - Google Patents
A kind of satellite momentum wheel fault diagnosis modeling method and system based on HC-TP-K2 algorithm Download PDFInfo
- Publication number
- CN110110363A CN110110363A CN201910248691.XA CN201910248691A CN110110363A CN 110110363 A CN110110363 A CN 110110363A CN 201910248691 A CN201910248691 A CN 201910248691A CN 110110363 A CN110110363 A CN 110110363A
- Authority
- CN
- China
- Prior art keywords
- model
- algorithm
- node
- network model
- momenttum wheel
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Probability & Statistics with Applications (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention discloses a kind of satellite momentum wheel fault diagnosis modeling methods and system based on HC-TP-K2 algorithm.In order to solve the problems, such as the node sequence in K2 algorithm, the present invention proposes a kind of K2 algorithm based on hill-climbing algorithm and topological sorting, data are learnt first with hill-climbing algorithm in the algorithm to obtain initial network structure, the topological sorting of node is obtained using topological sorting algorithm later, finally the node sequence is applied in K2 algorithm, to improve accuracy of the K2 algorithm in the modeling of satellite momentum wheel fault diagnosis.
Description
Technical field
The present invention relates to research method of the Bayesian network in the foundation of satellite momentum wheel fault diagnosis model, more particularly to
The fields such as Space Vehicle Health management and computer technology, more particularly, it relates to which a kind of be based on HC-TP-K2 algorithm
Satellite momentum wheel fault diagnosis modeling method and system.
Background technique
Momenttum wheel is the critical component of satellite attitude control system, and traditional satellite momentum wheel fault diagnosis technology is as observed
Device method, expert system approach etc. can not preferably solve uncertainty present in satellite momentum wheel failure.Bayesian network is as one
Kind probability net, there is apparent advantage, the mould that bayesian network structure learning is established in the reasoning of uncertain problem
Type will affect the accuracy of final the reasoning results.Bayesian Network Structure study is a np hard problem, and K2 algorithm is common shellfish
This web frame learning algorithm of leaf, but K2 algorithm there are problems that depending on model node sequence unduly.
Bayesian network structure learning K2 algorithm brief introduction
K2 algorithm is classical Algorithm for Bayesian Networks Structure Learning, and initially use CH scoring is used as scoring functions, later
There is scholar that the score functions such as BIC scoring and BDeu scoring are introduced into K2 algorithm, it is high the purpose is to find score function
Model.Due to computation complexity, K2 algorithm finds optimal models under the present conditions, it uses variables reordering and section
Maximum father node number is put to reduce the search space of algorithm.
K2 algorithm is since one comprising all nodes but not the non-directed graph on side, and in search process, algorithm successively compares
Compared with the node in variables reordering.In variables reordering, present node be only possible to be subsequent node father node, therefore looking for father
During node, algorithm need to only traverse the network structure that the node before coming present node is constituted, and present node it
Node afterwards then without the concern for.The father node number of any variable in algorithm setting network structure is no more than given simultaneously
Maximum father node number, when the scoring maximum for the network structure that each variable and its father node are constituted, final network knot
The scoring of structure also will be maximum.
Hill-climbing algorithm brief introduction
Hill-climbing algorithm (HC) is also common classic algorithm in bayesian network structure learning, and target is to find out scoring
Highest model.The algorithm is searched for since an initial model first, and initial model is generally set to boundless model.Later
In each step of search, partial modification is carried out to "current" model with searching operators first, obtains a series of candidate family;Then
The scoring of each candidate family is calculated, and best candidate model is compared with "current" model;If best candidate model is commented
Divide greatly, is then continued searching with it for next "current" model;Otherwise, it just stops search, and returns to "current" model.
The searching operators of hill-climbing algorithm have 3, when being edged respectively, subtracting while and reverse.Edged is in network structure
Middle addition a line, subtracting side is to subtract a line, and turning side then is that the direction of a line is overturn.Edged and turn making for side operator
With there is a premise, i.e., directed cycle cannot be formed in a network.
Summary of the invention
The purpose of invention is to propose a kind of satellite momentum wheel fault diagnosis modeling method based on HC-TP-K2 algorithm, with gram
The deficiencies of calculating present in existing momenttum wheel fault diagnosis modeling method is complicated, and model is not flexible is taken, the failure of momenttum wheel is made
Diagnostic result is more acurrate.
The present invention solves its technical problem, the used satellite momentum wheel fault diagnosis modeling based on HC-TP-K2 algorithm
Method the following steps are included:
(1) Bayesian network model of momenttum wheel failure is established based on HC-TP-K2 algorithm combination momenttum wheel fault data;
(2) Bayesian network model established based on step (1) learns to obtain in model using Bayesian network parameters
The parameter of each network node;
(3) it is based on the evident information of step (2) model and acquisition obtained with node parameter, using Bayes
Network reasoning algorithm carries out probability calculation to that may cause a variety of causes that momenttum wheel breaks down;
(4) according to step (3) calculated probability, the maximum node of wherein posterior probability is found out, as final diagnosis
As a result.
Further, in the satellite momentum wheel fault diagnosis modeling method of the invention based on HC-TP-K2 algorithm, institute
State step (1) the following steps are included:
(1-1) extracts the variable in momenttum wheel fault data as the node in network model to be built;
(1-2) generates initial network model θ comprising all nodes but boundless0, and as current network model θ ←
θ0;
(1-3) is scored to obtain oldScore using BIC score function to current network model;
Operation operator carries out partial modification to current network model when (1-4) is utilized respectively edged, subtracts while or reverse, and obtains
Each candidate family θ ';
(1-5) scores to each candidate family θ ' using BIC score function, finds the highest candidate family that wherein scores
θ*;
(1-6) is if the scoring newScore of candidate family θ * is greater than the scoring oldScore of "current" model θ, by candidate mould
Type θ * is as "current" model, even θ ← θ *, oldScore ← newScore, and return step (1-4);Otherwise continue step (1-
7);
(1-7) carries out depth-first traversal to current network model, obtains the topological sorting of each node in network model;
Node topology sequence is input in K2 algorithm by (1-8), obtains optimal momenttum wheel failure Bayesian network model
Final Bayesian network model as momenttum wheel failure.
Further, in the satellite momentum wheel fault diagnosis modeling method of the invention based on HC-TP-K2 algorithm, institute
It states in step (1-1):
If momenttum wheel fault data be non-discrete data, need to first to fault data carry out sliding-model control after carry out again after
Continuous processing.
According to another aspect of the present invention, the present invention is to solve its technical problem, is additionally provided based on HC-TP-K2 algorithm
Satellite momentum wheel fault diagnosis modeling, comprise the following modules:
Model building module, for establishing the shellfish of momenttum wheel failure based on HC-TP-K2 algorithm combination momenttum wheel fault data
This network model of leaf;
Node parameter obtains module, and the Bayesian network model for being established based on model building module utilizes pattra leaves
The study of this network parameter obtains the parameter of each network node in model;
Probability evaluation entity, for obtaining module model obtained with node parameter based on node parameter and obtaining
The evident information obtained is carried out generally using Bayesian Network Inference algorithm to that may cause a variety of causes that momenttum wheel breaks down
Rate calculates;
As a result determining module, for it is maximum to find out wherein posterior probability according to the calculated probability of probability evaluation entity
Node, as final diagnostic result.
Further, in the satellite momentum wheel fault diagnosis modeling of the invention based on HC-TP-K2 algorithm, institute
Stating model building module includes with lower unit:
Node acquiring unit, for extracting the variable in momenttum wheel fault data as the section in network model to be built
Point;
Model initialization unit includes all nodes but boundless initial network model θ for generating0, and as
Current network model θ ← θ0;
First scoring unit, for being scored to obtain oldScore to current network model using BIC score function;
Candidate family establishes unit, and operation operator is to current network model when for being utilized respectively edged, subtracting while or reverse
Partial modification is carried out, each candidate family θ ' is obtained;
Second scoring unit, scored using BIC score function each candidate family θ ', find wherein score it is highest
Candidate family θ *;
Model modification unit, if being greater than the scoring of "current" model θ for the scoring newScore of candidate family θ *
OldScore, then using candidate family θ * as "current" model, even θ ← θ *, oldScore ← newScore, and return to candidate mould
Type establishes unit;Otherwise continue step topological sorting acquiring unit;
Topological sorting acquiring unit obtains each in network model for carrying out depth-first traversal to current network model
The topological sorting of node;
Model determination unit obtains optimal momenttum wheel failure shellfish for node topology sequence to be input in K2 algorithm
This network model of leaf, the final Bayesian network model as momenttum wheel failure.
Further, in the satellite momentum wheel fault diagnosis modeling of the invention based on HC-TP-K2 algorithm, institute
It states in node acquiring unit:
If momenttum wheel fault data be non-discrete data, need to first to fault data carry out sliding-model control after carry out again after
Continuous processing.
Implement the satellite momentum wheel fault diagnosis modeling method and system of the invention based on HC-TP-K2 algorithm, have with
Down the utility model has the advantages that in order to solve the problems, such as that the node sequence in K2 algorithm, the present invention propose a kind of based on hill-climbing algorithm and topological sorting
K2 algorithm, data are learnt first with hill-climbing algorithm to obtain initial network structure in the algorithm, later using opening up
The topological sorting that sort algorithm obtains node is flutterred, finally the node sequence is applied in K2 algorithm, to improve K2 algorithm in satellite
Accuracy in the modeling of momenttum wheel fault diagnosis.
Detailed description of the invention
Present invention will be further explained below with reference to the attached drawings and examples, in attached drawing:
Fig. 1 is the stream of satellite momentum wheel fault diagnosis modeling method one embodiment of the invention based on HC-TP-K2 algorithm
Cheng Tu;
Fig. 2 is the flow chart of HC-TP-K2 algorithm of the invention;
Fig. 3 is the momenttum wheel bayesian network structure figure established by different level.
Specific embodiment
For a clearer understanding of the technical characteristics, objects and effects of the present invention, now control attached drawing is described in detail
A specific embodiment of the invention.
Fig. 1 is the stream of satellite momentum wheel fault diagnosis modeling method one embodiment of the invention based on HC-TP-K2 algorithm
Cheng Tu.In the present embodiment, the satellite momentum wheel fault diagnosis modeling method of the invention based on HC-TP-K2 algorithm include with
Lower step:
(1) Bayesian network model of momenttum wheel failure is established based on HC-TP-K2 algorithm combination momenttum wheel fault data;
(2) Bayesian network model established based on step (1) learns to obtain in model using Bayesian network parameters
The parameter of each network node;
(3) it is based on the evident information of step (2) model and acquisition obtained with node parameter, using Bayes
Network reasoning algorithm carries out probability calculation to that may cause a variety of causes that momenttum wheel breaks down;
(4) according to step (3) calculated probability, the maximum node of wherein posterior probability is found out, as final diagnosis
As a result.
The flow chart of HC-TP-K2 algorithm of the invention with reference to Fig. 2, Fig. 2, the step (1) the following steps are included:
(1-1) extracts the variable in momenttum wheel fault data as the node in network model to be built;
(1-2) generates initial network model θ comprising all nodes but boundless0, and as current network model θ ←
θ0;
(1-3) is scored to obtain oldScore using BIC score function to current network model;
Operation operator carries out partial modification to current network model when (1-4) is utilized respectively edged, subtracts while or reverse, and obtains
Each candidate family θ ';
(1-5) scores to each candidate family θ ' using BIC score function, finds the highest candidate family that wherein scores
θ*;
(1-6) is if the scoring newScore of candidate family θ * is greater than the scoring oldScore of "current" model θ, by candidate mould
Type θ * is as "current" model, even θ ← θ *, oldScore ← newScore, and return step (1-4);Otherwise continue step (1-
7);
(1-7) carries out depth-first traversal to current network model, obtains the topological sorting of each node in network model;
Node topology sequence is input in K2 algorithm by (1-8), obtains optimal momenttum wheel failure Bayesian network model
Final Bayesian network model as momenttum wheel failure.
Wherein, in step (1-1), if momenttum wheel fault data is non-discrete data, fault data need to be carried out first
Subsequent processing is carried out after sliding-model control again.
Below to algorithm proposed by the present invention carry out experimental verification, experiment the following steps are included:
(1) according to momenttum wheel physical structure and fault mode, using the momenttum wheel fault model established by different level as experiment
Contrast standard;
(2) HC-TP-K2 algorithm, Random-K2 algorithm and hill-climbing algorithm are utilized respectively and establishes momenttum wheel failure Bayesian network
Network model;
(3) network model that above-mentioned three kinds of algorithms are established is compared with master pattern.
By momenttum wheel FMEA, the fault mode of momenttum wheel itself, i.e. momenttum wheel stalling, control precision deficiency, function are obtained
It consumes excessive;Then the fault mode for obtaining remaining component obtains the fail close between each components and component by logic analysis
System.The momenttum wheel failure criterion model established according to experimental procedure (1) is as shown in Figure 3.
Table 1 gives the value condition of each corresponding identifier of node and each node in model.Wherein, X node generation
Whether table momenttum wheel breaks down, and tri- nodes of A, B, C respectively represent three kinds of common failure patterns of momenttum wheel, and D~I etc. six
The node on behalf fault mode of four components of momenttum wheel, 12 nodes such as J~U then represent may cause momenttum wheel
The reason of failure, i.e., the fault mode of each components in momenttum wheel.
1 network model node specification of table
Under 5 groups of data of Experimental comparison, hill-climbing algorithm, HC-TP-K2 algorithm and the random K2 algorithm for generating node sequence
The correctness of (hereinafter referred to as Random-K2 algorithm) learning outcome.It is quasi- as structure learning algorithm result using following desired values
The evaluation criterion of true property, the value is bigger, represents algorithm and learns network structure out closer to above-mentioned standard network, the value is most
Greatly 1.
Wherein, CE is that algorithm learns correct side out compared with standard network architecture, and ME is missing side, and IE is to reverse side,
RE is extra side.In order to avoid the contingency of experimental result, every group of data take the average value of 5 experimental results to be used as and most terminate
Fruit.
The comparison of 2 experimental result of table
According to the experimental result of table 2 it is found that the learning effect of HC-TP-K2 algorithm is better than Random-K2 algorithm and HC
Hill-climbing algorithm is obtained node in conjunction with topological sorting and sorted in advance by algorithm, and the sequence is applied to can be improved in K2 algorithm
Correctness of the K2 algorithm in the foundation of momenttum wheel fault diagnosis model.
The embodiment of the present invention is described with above attached drawing, but the invention is not limited to above-mentioned specific
Embodiment, the above mentioned embodiment is only schematical, rather than restrictive, those skilled in the art
Under the inspiration of the present invention, without breaking away from the scope protected by the purposes and claims of the present invention, it can also make very much
Form, all of these belong to the protection of the present invention.
Claims (6)
1. a kind of satellite momentum wheel fault diagnosis modeling method based on HC-TP-K2 algorithm, which is characterized in that including following step
It is rapid:
(1) Bayesian network model of momenttum wheel failure is established based on HC-TP-K2 algorithm combination momenttum wheel fault data;
(2) Bayesian network model established based on step (1) learns to obtain each net in model using Bayesian network parameters
The parameter of network node;
(3) it is based on the evident information of step (2) model and acquisition obtained with node parameter, using Bayesian network
Reasoning algorithm carries out probability calculation to that may cause a variety of causes that momenttum wheel breaks down;
(4) according to step (3) calculated probability, the maximum node of wherein posterior probability is found out, as final diagnostic result.
2. the satellite momentum wheel fault diagnosis modeling method according to claim 1 based on HC-TP-K2 algorithm, feature
Be, the step (1) the following steps are included:
(1-1) extracts the variable in momenttum wheel fault data as the node in network model to be built;
(1-2) generates θ ° of initial network model comprising all nodes but boundless, and as θ ← θ ° of current network model;
(1-3) is scored to obtain oldScore using BIC score function to current network model;
Operation operator carries out partial modification to current network model when (1-4) is utilized respectively edged, subtracts while or reverse, and obtains each time
Modeling type θ ';
(1-5) scores to each candidate family θ ' using BIC score function, finds the highest candidate family θ that wherein scores*;
(1-6) is if candidate family θ*Scoring newScore be greater than "current" model θ scoring oldScore, then by candidate family θ*
As "current" model, even θ ← θ*, oldScore ← newScore, and return step (1-4);Otherwise continue step (1-7);
(1-7) carries out depth-first traversal to current network model, obtains the topological sorting of each node in network model;
Node topology sequence is input in K2 algorithm by (1-8), obtains optimal momenttum wheel failure Bayesian network model conduct
The final Bayesian network model of momenttum wheel failure.
3. the satellite momentum wheel fault diagnosis modeling method according to claim 2 based on HC-TP-K2 algorithm, feature
It is, in the step (1-1):
If momenttum wheel fault data is non-discrete data, subsequent place is carried out again after sliding-model control need to being carried out to fault data first
Reason.
4. a kind of satellite momentum wheel fault diagnosis modeling based on HC-TP-K2 algorithm, which is characterized in that including with lower die
Block:
Model building module, for establishing the Bayes of momenttum wheel failure based on HC-TP-K2 algorithm combination momenttum wheel fault data
Network model;
Node parameter obtains module, and the Bayesian network model for being established based on model building module utilizes Bayesian network
Network parameter learning obtains the parameter of each network node in model;
Probability evaluation entity, for obtaining the module model and acquisition obtained with node parameter based on node parameter
Evident information carries out probability meter to that may cause a variety of causes that momenttum wheel breaks down using Bayesian Network Inference algorithm
It calculates;
As a result determining module, for finding out the maximum node of wherein posterior probability according to the calculated probability of probability evaluation entity,
As final diagnostic result.
5. the satellite momentum wheel fault diagnosis modeling according to claim 4 based on HC-TP-K2 algorithm, feature
It is, the model building module includes with lower unit:
Node acquiring unit, for extracting the variable in momenttum wheel fault data as the node in network model to be built;
Model initialization unit includes all nodes but θ ° boundless of initial network model for generating, and as current
θ ← θ ° of network model;
First scoring unit, for being scored to obtain oldScore to current network model using BIC score function;
Candidate family establishes unit, and operation operator carries out current network model when for being utilized respectively edged, subtracting while or reverse
Partial modification obtains each candidate family θ ';
Second scoring unit, is scored to each candidate family θ ' using BIC score function, finds the highest candidate that wherein scores
Model θ*;
Model modification unit, if being used for candidate family θ*Scoring newScore be greater than "current" model θ scoring oldScore, then
By candidate family θ*As "current" model, even θ ← θ*, oldScore ← newScore, and return to candidate family and establish unit;
Otherwise continue step topological sorting acquiring unit;
Topological sorting acquiring unit obtains each node in network model for carrying out depth-first traversal to current network model
Topological sorting;
Model determination unit obtains optimal momenttum wheel failure Bayes for node topology sequence to be input in K2 algorithm
Network model, the final Bayesian network model as momenttum wheel failure.
6. the satellite momentum wheel fault diagnosis modeling according to claim 5 based on HC-TP-K2 algorithm, feature
It is, in the node acquiring unit:
If momenttum wheel fault data is non-discrete data, subsequent place is carried out again after sliding-model control need to being carried out to fault data first
Reason.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910248691.XA CN110110363A (en) | 2019-03-29 | 2019-03-29 | A kind of satellite momentum wheel fault diagnosis modeling method and system based on HC-TP-K2 algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910248691.XA CN110110363A (en) | 2019-03-29 | 2019-03-29 | A kind of satellite momentum wheel fault diagnosis modeling method and system based on HC-TP-K2 algorithm |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110110363A true CN110110363A (en) | 2019-08-09 |
Family
ID=67484715
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910248691.XA Pending CN110110363A (en) | 2019-03-29 | 2019-03-29 | A kind of satellite momentum wheel fault diagnosis modeling method and system based on HC-TP-K2 algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110110363A (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102288398A (en) * | 2011-05-16 | 2011-12-21 | 南京航空航天大学 | Momentum wheel fault detection device and method based on support vector machine |
CN103245911A (en) * | 2013-05-03 | 2013-08-14 | 云南电力试验研究院(集团)有限公司电力研究院 | Breaker fault diagnosis method based on Bayesian network |
CN108363310A (en) * | 2018-01-15 | 2018-08-03 | 北京航空航天大学 | Artificial intelligence program person writes the weight clustering decision-making of digital aircraft source code |
US20180276322A1 (en) * | 2017-03-24 | 2018-09-27 | Kabushiki Kaisha Toshiba | Failure diagnosis apparatus, monitoring apparatus, failure diagnosis method and recording medium |
CN109033710A (en) * | 2018-08-30 | 2018-12-18 | 电子科技大学 | A kind of momenttum wheel reliability estimation method based on more performance degradations |
-
2019
- 2019-03-29 CN CN201910248691.XA patent/CN110110363A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102288398A (en) * | 2011-05-16 | 2011-12-21 | 南京航空航天大学 | Momentum wheel fault detection device and method based on support vector machine |
CN103245911A (en) * | 2013-05-03 | 2013-08-14 | 云南电力试验研究院(集团)有限公司电力研究院 | Breaker fault diagnosis method based on Bayesian network |
US20180276322A1 (en) * | 2017-03-24 | 2018-09-27 | Kabushiki Kaisha Toshiba | Failure diagnosis apparatus, monitoring apparatus, failure diagnosis method and recording medium |
CN108363310A (en) * | 2018-01-15 | 2018-08-03 | 北京航空航天大学 | Artificial intelligence program person writes the weight clustering decision-making of digital aircraft source code |
CN109033710A (en) * | 2018-08-30 | 2018-12-18 | 电子科技大学 | A kind of momenttum wheel reliability estimation method based on more performance degradations |
Non-Patent Citations (2)
Title |
---|
厉海涛: "基于贝叶斯网络的动量轮可靠性建模与分析", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
吴永广等: "K2&HC结构学习算法", 《计算机与数字工程》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Tao et al. | An object detection system based on YOLO in traffic scene | |
Kong et al. | Ron: Reverse connection with objectness prior networks for object detection | |
CN106168799B (en) | A method of batteries of electric automobile predictive maintenance is carried out based on big data machine learning | |
CN106022380A (en) | Individual identity identification method based on deep learning | |
Djenouri et al. | Fast and accurate deep learning framework for secure fault diagnosis in the industrial internet of things | |
CN106533759B (en) | A kind of link prediction method based on path entropy in multitiered network | |
CN112149721B (en) | Target detection method for reducing labeling requirements based on active learning | |
CN111488946A (en) | Radar servo system fault diagnosis method based on information fusion | |
CN104268629A (en) | Complex network community detecting method based on prior information and network inherent information | |
CN106780058A (en) | The group dividing method and device of dynamic network | |
CN111311702A (en) | Image generation and identification module and method based on BlockGAN | |
CN111598435A (en) | Quality trend prediction method based on adaptive feature selection and improved thought evolution algorithm | |
CN113780432B (en) | Intelligent detection method for operation and maintenance abnormity of network information system based on reinforcement learning | |
CN114897144A (en) | Complex value time sequence signal prediction method based on complex value neural network | |
CN109766905A (en) | Target cluster dividing method based on Self-Organizing Feature Maps | |
CN110110363A (en) | A kind of satellite momentum wheel fault diagnosis modeling method and system based on HC-TP-K2 algorithm | |
CN109492677A (en) | Time-varying network link prediction method based on bayesian theory | |
WO2016122636A1 (en) | Updates to a prediction model using statistical analysis groups | |
Sarkar | Learning to form large groups of salient image features | |
CN116975742A (en) | Partial discharge pattern recognition method, apparatus, device, and storage medium | |
Berton et al. | The Impact of Network Sampling on Relational Classification. | |
CN108846439A (en) | Bayesian network parameters learning method based on small data set under the conditions of a kind of no priori knowledge | |
Gubanov et al. | Language games in investigation of social networks: Finding communities and influential agents | |
Shi et al. | Fault diagnosis of pneumatic actuator using adaptive network-based fuzzy inference system models and a learning vector quantization neural network | |
Zhou | Research on the Complexity Characteristics of Convolutional Neural Networks |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190809 |