CN103678886B - A kind of satellite Bayesian network health based on ground test data determines method - Google Patents
A kind of satellite Bayesian network health based on ground test data determines method Download PDFInfo
- Publication number
- CN103678886B CN103678886B CN201310611919.XA CN201310611919A CN103678886B CN 103678886 B CN103678886 B CN 103678886B CN 201310611919 A CN201310611919 A CN 201310611919A CN 103678886 B CN103678886 B CN 103678886B
- Authority
- CN
- China
- Prior art keywords
- satellite
- data
- node
- bayesian network
- test data
- 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
Landscapes
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a kind of satellite Bayesian network health based on ground test data and determine method.The invention provides satellite health status comprehensive quantification based on ground test data and determine method, from historical data angle, excavate the probability between satellite and on-board equipment and relevant parameter, fully demonstrate the feature of data under actual condition, the implicit association relation of parameter and system holistic health level has been excavated, true reflection system health level with the conditional probability of multiparameter.The present invention can relatively accurately realize the health state evaluation in satellite test stage, it is achieved satellite is quantified as unified Health Category by multi objective telemetry parameter, and the ground test analysis for satellite provides efficient method.
Description
Technical field
The present invention relates to a kind of satellite health and determine method, particularly relate to a kind of satellite pattra leaves based on ground test data
This network health determines method, belongs to satellite ground stage prognostic and health management technical field.
Background technology
The ground test data of satellite all have non-for aspects such as the condition discrimination before satellite launch, emission decision formulations
The most important effect.Will appreciate that in conjunction with the ground test data before satellite launch, ground test personnel and design of satellites personnel
Overall health before satellite launch, thus take corresponding ground test safeguard further, to ensure satellite launch
The kilter in rear active phase and in orbit stage.During satellite ground test, test data are to embody satellite to transport at present
The important evidence that whether normal row state is.Satellite health status based on ground test data determine method, mainly utilize satellite
Ground test data, the injection failure condition of combined ground test phase, by the reasoning of network, provide object to be determined
General health grade.
In current satellite ground test process, currently mainly use interpretation method, by multidimensional remote measurement number
Determine the health status of on-board equipment according to the interpretation of bound, rate of change and variation tendency, this method is to satellite and subsystem
The detection exception in ground test stage, fault etc. serve important function.But owing to satellite ground amount of test data is big, data
Dimension is high, not yet carries out overall quantization according to multi-Dimensional parameters, it is impossible to reflection satellite holistic health state.The most undue dependence is artificial
Heuristics, it is difficult to abnormal phenomena is made the reaction of decision-making fast and accurately.
Accordingly, it would be desirable to find a kind of solution, it is possible to for above not enough, in conjunction with satellite ground testing parameter data,
History ground test data are carried out complex reasoning with Current terrestrial test data, provides Current terrestrial test phase in real time and defend
Star Health Category aggregate level.
Summary of the invention
Present invention solves the technical problem that and be: overcome the deficiencies in the prior art, it is provided that be a kind of based on ground test data
Satellite Bayesian network health determines method, it is possible to relatively accurately determine the whole star of satellite and the health in equipment ground testing stage
State, the ground test analysis for satellite provides efficient means.
The technical scheme is that a kind of satellite Bayesian network health based on ground test data determines method,
Comprise the steps:
(1) Bayesian network is built: build the Bayesian network for determining satellite health status to be assessed, wherein pattra leaves
The leaf node of this network is whole star health status, and intermediate node is on-board equipment, and root node is the ground that on-board equipment is corresponding
Test parameter；
(2) training sample is determined: from the ground test historical data of satellite to be assessed, extract the ground that each root node is corresponding
Face test parameter historical data, carries out error code and singular point is rejected, null value is filled and mean value computation pretreatment；To therein continuously
Variable carries out discretization, determines the corresponding relation from connection attribute to Category Attributes, if will the codomain of numerical attribute be divided into
Dry subinterval, each interval corresponding centrifugal pump；Determine in the middle of under the conditions of root node difference value according to historical evaluation result
The value of leaf node under the conditions of the value of node, and intermediate node difference value；Through pretreatment, the root node of discretization
Data and the intermediate node of correspondence and leaf node data constitute satellite Bayesian network health and determine the training sample of model；
(3) utilize the Bayesian network constructed by training sample training step (1) that step (2) obtains, obtain condition general
Rate: by root node and the prior probability of each intermediate node in the constructed Bayesian network of training sample statistics, and enter one accordingly
Step is calculated each father node and takes the conditional probability of child node under the conditions of different value, i.e. obtains under the conditions of each root node takes different value
The conditional probability of intermediate node, each intermediate node take the conditional probability of leaf node under the conditions of different value；
(4) satellite health status is determined: rejected through error code and singular point by satellite to be assessed, null value is filled and average meter
In the Bayesian network constructed by Current terrestrial test parameter input step (1) of each root node obtained after calculating pretreatment, shellfish
This network of leaf calculates the health status exporting whole star according to the conditional probability that step (3) obtains.
Described step (2) is extracted the ground test parameters history data that each root node is corresponding, carries out error code and singular point
The implementation method that pretreatment is filled in rejecting, null value is as follows:
Singular point rejects the main serial mean using and calculating ground test parameters history data corresponding to each root node
And variance yields, determine, by the way of meansigma methods three times of variance yields of plus-minus, the number that in test data, the data of 95% should be in
According to scope, thus eliminate singular point data not in this range, directly this data point is set to sky；Calculate reject error code and
Minute average of data is tested after singular point；For null value point data, utilize the method for linear interpolation that null value point is carried out benefit value,
Thus forming a complete ordered series of numbers, it is assumed that the missing data section of sequence of test data is Xt,Xt+1,L,Xt+M-1, then linear interpolation
Method data mend some formula:
Wherein, i=0,1,2, L, M-1,2≤t≤M-1, M are disappearance test data length, and M is the nature more than or equal to 3
Number, X is sequence of test data to be detected, Xt,Xt+1,L,Xt+M-1For disappearance test data, Xt-1、Xt+MIt is respectively disappearance test number
Previous data, the later telemetry at disappearance test data segment end according to section beginning.
Under the conditions of in described step (3), each father node of calculating takes different value, the method for the conditional probability of child node is as follows:
Wherein, Y={y1,y2,...,yN, X={x1,x2,...,xK, 1≤i≤K, 1≤j≤N, K, N are natural number；X、
Y is the node in constructed Bayesian network, and Y is the child node of X, and X is the father node of Y；P(Y=yj,X=xi) it is Y=yj、X
=xiPrior probability, P (Y=yj) it is Y=yjPrior probability；P(Y=yj|X=xi) it is X=xiUnder the conditions of Y=yjConditional probability.
What the present invention compared with prior art had the advantages that the present invention provides based on ground test data defends
Star health status comprehensive quantification determines method, and from historical data angle, that has excavated between system under evaluation relevant parameter is general
Rate, has fully demonstrated the feature of data under actual condition, has excavated parameter and system holistic health with the conditional probability of multiparameter
The implicit association relation of level, true reflection system health level.The present invention can relatively accurately realize on the whole star of satellite, star
The health status in equipment ground testing stage determines, it is achieved the objects such as the whole star of satellite, on-board equipment are by multi objective telemetry parameter amount
Turning to unified Health Category, the ground test analysis for satellite provides efficient method.
Accompanying drawing explanation
Fig. 1 is the inventive method flow chart；
Fig. 2 is to utilize what the inventive method set up to include accumulator battery and the Bayesian network of two on-board equipments of momenttum wheel
Network；
Fig. 3 is that root node only includes monomer voltage and the accumulator battery Bayesian network of A group voltage.
Detailed description of the invention
In order to solve problems of the prior art, the present invention provides a kind of satellite pattra leaves based on ground test data
This network health appraisal procedure, the core concept of described method is: owing to satellite is in the ground test stage, the working condition of setting
Different, satellite actual health level difference, causes its test data variation to differ, be difficult to comprehensively determine health level.By inciting somebody to action
System under evaluation, work condition state and test parameter are set up corresponding Bayes and are determined network, utilize historical data to carry out network instruction
After white silk, it is achieved utilize current data to determine satellite health status.
The principle of the method for the invention is: the method for the present invention realizes satellite for the history telemetry utilizing satellite
The detection of abnormality and the problem of extraction, first, by means of satellite fields ground stage design principle knowledge and ground test
Engineering experience, is designed for the Bayesian Network Topology Structures of satellite ground test health evaluating；Secondly, analyzing and processing history ground
Face test data, set up Bayesian network training sample in conjunction with expertise experience and historical data；Again, according to training sample
Calculate the condition between Bayesian network each node prior probability and conditional probability, i.e. the training each Hierarchy nodes of Bayesian network general
Rate；Finally, current test data to be assessed are utilized to calculate corresponding on-board equipment and whole star health status.
Specifically include structure Bayesian network, process historical test data and determine training sample, utilize training sample to instruct
Practice Bayesian network to obtain prior probability and conditional probability, determine satellite health status；
(1) build Bayesian network: according to satellite whole star structure composition and each parameter index, build be used for determining to be assessed
The Bayesian network of satellite health status, determines the hierarchical relationship between whole star, on-board equipment, ground test parameter；Bayes
Network is a directed acyclic graph, and it is by the node representing variable and connects the directed edge of these nodes and constitutes, and directed edge is by father
Node points to child node, and the node without child node is leaf node, and the node without father node is root node；The leaf of Bayesian network
Child node is whole star health status, and the intermediate node i.e. father node of leaf node is on-board equipment, the i.e. on-board equipment of root node
Father node is this equipment ground testing parameter, and ground test parameter is without father node, for root node；
(2) process historical data and determine training sample: extracting each from the ground test historical data of satellite to be assessed
The ground test parameters history data that root node is corresponding, carry out error code and singular point are rejected, null value is filled and mean value computation is located in advance
Reason；Continuous variable therein is carried out discretization, determines the corresponding relation from connection attribute to Category Attributes, will numerical attribute
Codomain be divided into some subintervals, each interval corresponding centrifugal pump；Determine that often group ground is surveyed according to historical evaluation result
On-board equipment and the Health Category of whole star under examination Parameter Conditions, i.e. determine taking of intermediate node under the conditions of root node difference value
Value, and the value of leaf node under the conditions of intermediate node difference value；Through pretreatment, the root node data of discretization and right
The intermediate node answered and leaf node data constitute satellite Bayesian network health and determine the training sample of model；
(3) utilize the Bayesian network constructed by training sample training step (1), obtain prior probability and conditional probability:
Utilize the Bayesian network constructed by training sample training, by root node in training sample the built Bayesian network of statistics with each
The prior probability of intermediate node, calculates each father node the most successively and takes the conditional probability of child node under the conditions of different value, i.e. calculate
Each root node takes the conditional probability of intermediate node under the conditions of different value, each intermediate node takes the bar of leaf node under the conditions of different value
Part probability, the conditional probability of whole star health status under the conditions of finally drawing on-board equipment Health Category；
(4) satellite health status is determined: rejected through error code and singular point by satellite to be assessed, null value is filled and average meter
In the Bayesian network constructed by Current terrestrial test parameter input step (1) of each root node obtained after calculating pretreatment, shellfish
This network of leaf calculates the health status exporting whole star according to the conditional probability that step (3) obtains.
Embodiment:
Realize process the following detailed description of the present invention, for simplifying the analysis, the most only consider accumulator battery and momenttum wheel
The impact on whole star health status of two on-board equipments, sets up the Bayesian network for determining satellite health status, such as Fig. 2 institute
Show.And Bayesian network probability calculation and Bayesian inference processes are described as a example by monomer voltage and two test points of A group voltage,
Comprise the Bayesian network of monomer voltage and A two root nodes of group voltage as shown in Figure 3.
(1) Bayesian network for determining satellite health status to be assessed is set up.
Set up for determining the Bayesian network of satellite health status as in figure 2 it is shown, whole starlike state is Bayesian network
The leaf node of network, its father node is on-board equipment i.e. accumulator and momenttum wheel；The father node of momenttum wheel is its ground test ginseng
Number, i.e. momenttum wheel bearing temperature, momenttum wheel current of electric, momenttum wheel motor current and momenttum wheel rotating speed, these parameter points are root
Node；The father node of accumulator is its ground test parameter, i.e. A group temperature, B group temperature, monomer voltage, charging current and A group
Voltage, these Parameter nodes are root node.
(2) sample set of training Bayesian network is determined.
A, the ground test parameters history data prediction corresponding to root node
The ground test parameters history number that each root node is corresponding is extracted from the ground test historical data of satellite to be assessed
According to, carry out error code and singular point is rejected, null value is filled and mean value computation pretreatment, the average per minute fetched data here；
Singular point rejects main serial mean and the variance yields using and calculating ground test data, is added and subtracted by meansigma methods
The mode of three times of variance yields determines the scope of data that in test data, the data of 95% should be in, thus eliminates not at this
In the range of singular point data, directly this data point is set to sky；Calculate test after rejecting error code and singular point data minute
Average；For null value point data, utilize the method for linear interpolation that null value point carries out benefit value, thus form a complete number
Row, it is assumed that the missing data section of sequence of test data is Xt,Xt+1,L,Xt+M-1, then a linear interpolation data benefit point formula is:
Wherein, i=0,1,2, L, M-1,2≤t≤M-1, M are disappearance test data length, and M is the nature more than or equal to 3
Number, X is sequence of test data to be detected, Xt,Xt+1,L,Xt+M-1For disappearance test data, Xt-1、Xt+MIt is respectively disappearance test number
Previous data, the later telemetry at disappearance test data segment end according to section beginning.
Continuous data discretization in B, the ground test parameter corresponding to root node
By satellite ground test Data Detection to the supplemental characteristic of each root node, continuous variable therein is carried out discrete
Changing, the codomain of numerical attribute will be divided into some subintervals, each interval corresponding centrifugal pump, numerical discretization algorithm is just
It it is the requirement corresponding relation that can automatically determine from connection attribute to Category Attributes.Leaf node monomer voltage with accumulator battery
As a example by A group voltage, by actual data analysis being determined the boundary that each node state divides is 3 classes.When parameter value is less than
The when of minimum threshold, it is set to the 1st grade, when parameter value is higher than high threshold, its grade is set to 3rd level.For
Value in normal fluctuation range, being evenly dividing is 2 grades.Parameters in accumulator health evaluating Bayesian network
Grade classification scope is as shown in table 1.
The discrete data table of table 1 ground test data
| Discrete level | 1 | 2 | 3 |
| Monomer voltage | <0.4 | 0.4～1.0 | >1.0 |
| A group voltage | <6 | 6～14.75 | >14.75 |
C, intermediate node divide with leaf node Health Category, i.e. on-board equipment divides with whole star Health Category
As shown in table 2, accumulator and satellite are all divided into Three Estate: normal, abnormal.When health status result of calculation
During more than 1.5, determine that satellite/equipment state is normal, when health status result of calculation is less than 1.5, determine satellite/equipment state
There may be exception, need to be analyzed further, diagnose.
Table 2 accumulator divides table with whole star Health Category
| Health Category | 1 | 2 |
| Health status describes | Abnormal | Normally |
Satellite shellfish is constituted through pretreatment, the root node data of discretization and the intermediate node of correspondence and leaf node data
This network health of leaf determines the training sample of model.
(3) utilize the Bayesian network constructed by training sample training step (1), obtain prior probability and conditional probability.
A, calculating prior probability
Through sample data statistical analysis, obtain root node monomer voltage, A group voltage priori under different discrete level
Probability is as shown in table 3, and the prior probability of accumulator is as shown in table 4.
The prior probability table of table 3 ground test discretization data
| Discrete level | 1 | 2 | 3 |
| Monomer voltage | 0.478 | 0.1016 | 0.4086 |
| A group voltage | 0.2257 | 0.3645 | 0.4098 |
Table 4 accumulator prior probability table
| Health Category | 1 | 2 |
| Prior probability | <0.102 | 0.898 |
B, design conditions probability
The method that conditional probability calculates is as follows:
Wherein, Y={y1,y2,...,yN, X={x1,x2,...,xK, 1≤i≤K, 1≤j≤N, K, N are natural number；X、
Y is the node that satellite health determines in Bayesian network, and Y is the child node of X, and X is the father node of Y；P(Y=yj,X=xi) it is Y=
yj、X=xiPrior probability, P (Y=yj) it is Y=yjPrior probability；P(Y=yj|X=xi) it is X=xiUnder the conditions of Y=yjCondition general
Rate.
Through statistical computation, each node is relative to the conditional probability of its father node as shown in table 5, table 6, and wherein XDC represents electric power storage
Pond, ZX represents whole star, and ADY represents A group voltage, and DDY represents monomer voltage.
Table 5 accumulator conditional probability
| (DDY,ADY) | XDC=1 | XDC=2 |
| (1,1) | 0.811 | 0.189 |
| (1,2) | 0.372 | 0.628 |
| (1,3) | 0.122 | 0.878 |
| (2,1) | 0.431 | 0.869 |
| (2,2) | 0.129 | 0.871 |
| (2,3) | 0.233 | 0.767 |
| (3,1) | 0.011 | 0.989 |
| (3,2) | 0.061 | 0.939 |
| (3,3) | 0.824 | 0.176 |
Table 6 whole star conditional probability
| XDC | ZX=1 | XDC=2 |
| 1 | 0.361 | 0.639 |
| 2 | 0.197 | 0.803 |
(4) satellite health status is determined
Satellite to be assessed is inputted through error code and singular point rejecting, null value in the Bayesian network constructed by step (1)
The Current terrestrial test data of each root node filled and obtain after pretreatment in mean value computation, the bar drawn according to step (3)
Part probability calculation goes out the health status of whole star.
As, current monolithic voltage=0.84, A group voltage=15.13, it is known that discretization monomer voltage=2, A group voltage=3；Will
In the Bayesian network that the monomer voltage of discretization and the input of A group magnitude of voltage build, obtaining whole star health status result of calculation is
1.765, so that it is determined that present satellites is in normal condition.
Bayesian network internal calculation process is: obtained by table 5, and the probability of accumulator=1 is 0.233, accumulator=2 general
Rate is 0.767, and obtaining accumulator health status result of calculation is 1.767, more than 1.5, determines that accumulator is in normal condition；Knot
Close table 6, whole star be the probability of 1 be 0.233*0.361+0.767*0.197=0.235, whole star be the probability of 2 be 0.233*0.639
+ 0.767*0.803=0.765, obtaining whole star health status is:
0.235*1+0.765*2=1.765>1.5
Determine that the current state of satellite is normal.
The non-detailed description of the present invention is known to the skilled person technology.
Claims (3)
1. satellite Bayesian network health based on ground test data determine method, it is characterised in that include walking as follows
Rapid:
(1) Bayesian network is built: build the Bayesian network for determining satellite health status to be assessed, wherein Bayesian network
The leaf node of network is whole star health status, and intermediate node is on-board equipment, and root node is the ground test that on-board equipment is corresponding
Parameter；
(2) training sample is determined: from the ground test historical data of satellite to be assessed, extract ground corresponding to each root node survey
Examination parameters history data, carry out error code and singular point are rejected, null value is filled and mean value computation pretreatment；To continuous variable therein
Carry out discretization, determine the corresponding relation from connection attribute to Category Attributes, the codomain of numerical attribute will be divided into some sons
Interval, each interval corresponding centrifugal pump；Intermediate node under the conditions of root node difference value is determined according to historical evaluation result
Value, and the value of leaf node under the conditions of intermediate node difference value；Through pretreatment, the root node data of discretization
And the intermediate node of correspondence and leaf node data constitute satellite Bayesian network health and determine the training sample of model；
(3) utilize the Bayesian network constructed by training sample training step (1) that step (2) obtains, obtain conditional probability: be logical
Cross root node and the prior probability of each intermediate node in the constructed Bayesian network of training sample statistics, and calculate the most further
Obtain each father node and take the conditional probability of child node under the conditions of different value, i.e. obtain each root node and take middle node under the conditions of different value
The conditional probability of point, each intermediate node take the conditional probability of leaf node under the conditions of different value；
(4) satellite health status is determined: rejected through error code and singular point by satellite to be assessed, null value is filled and mean value computation is pre-
In the Bayesian network constructed by Current terrestrial test parameter input step (1) of each root node obtained after process, Bayes
Network calculates the health status exporting whole star according to the conditional probability that step (3) obtains.
A kind of satellite Bayesian network health based on ground test data the most according to claim 1 determine method, its
It is characterised by: described step (2) is extracted the ground test parameters history data that each root node is corresponding, carries out error code and singular point
The implementation method that pretreatment is filled in rejecting, null value is as follows:
Singular point rejects main serial mean and the side using and calculating ground test parameters history data corresponding to each root node
Difference, determines, by the way of meansigma methods three times of variance yields of plus-minus, the scope of data that test data should be in, thus rejects
Go out singular point data not in this range, directly this data point is set to sky；Calculate and test number after rejecting error code and singular point
According to minute average；For null value point data, utilize the method for linear interpolation that null value point carries out benefit value, thus formed one complete
Whole ordered series of numbers, it is assumed that the missing data section of sequence of test data is Xt,Xt+1,…,Xt+M-1, then linear interpolation data mend some public affairs
Formula is:
Wherein, i=0,1,2 ..., M-1,2≤t≤M-1, M are disappearance test data length, and M is the natural number more than or equal to 3, X
For sequence of test data to be detected, Xt,Xt+1,…,Xt+M-1For disappearance test data, Xt-1、Xt+MIt is respectively disappearance test data segment
The previous test data of beginning, the later test data at disappearance test data segment end.
A kind of satellite Bayesian network health based on ground test data the most according to claim 1 determine method, its
It is characterised by: under the conditions of in described step (3), each father node of calculating takes different value, the method for the conditional probability of child node is as follows:
Wherein, Y={y1,y2,...,yN, X={x1,x2,...,xK, 1≤i≤K, 1≤j≤N, K, N are natural number；X、Y
Being the node in constructed Bayesian network, Y is the child node of X, and X is the father node of Y；P (Y=yj, X=xi) it is Y=
yj, X=xiPrior probability, P (Y=yj) it is Y=yjPrior probability；P (Y=yj| X=xi) it is X=xiUnder the conditions of Y=yj's
Conditional probability.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310611919.XA CN103678886B (en) | 2013-11-26 | 2013-11-26 | A kind of satellite Bayesian network health based on ground test data determines method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310611919.XA CN103678886B (en) | 2013-11-26 | 2013-11-26 | A kind of satellite Bayesian network health based on ground test data determines method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103678886A CN103678886A (en) | 2014-03-26 |
| CN103678886B true CN103678886B (en) | 2017-01-04 |
Family
ID=50316416
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310611919.XA Active CN103678886B (en) | 2013-11-26 | 2013-11-26 | A kind of satellite Bayesian network health based on ground test data determines method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103678886B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104590594B (en) * | 2015-01-27 | 2017-05-24 | 北京空间飞行器总体设计部 | Method for testing and verifying information flow among spacecraft |
| CN105468917B (en) * | 2015-12-01 | 2019-03-08 | 北京无线电计量测试研究所 | A kind of pipeline failure prediction technique and device |
| CN105573329B (en) * | 2015-12-16 | 2018-05-18 | 上海卫星工程研究所 | Rail control data analysing method based on decision tree |
| CN110046376B (en) * | 2019-02-26 | 2022-11-22 | 中国西安卫星测控中心 | Multi-working-condition health assessment method for satellite attitude control system based on Bayesian network |
| CN110334395B (en) * | 2019-05-28 | 2020-11-27 | 中国地质大学（武汉） | JADE-based satellite momentum wheel fault diagnosis method and system for initializing EM algorithm |
| CN112668749B (en) * | 2020-11-24 | 2023-07-07 | 江苏中矿安华科技发展有限公司 | Coal mine gas early warning method based on class mark weighting extreme learning machine |
| CN113222421A (en) * | 2021-05-18 | 2021-08-06 | 哈尔滨工业大学 | Satellite task support performance evaluation method and system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101354757A (en) * | 2008-09-08 | 2009-01-28 | 中国科学院地理科学与资源研究所 | Method for predicting dynamic risk and vulnerability under fine dimension |
| CN103218689A (en) * | 2013-04-23 | 2013-07-24 | 湖南工学院 | Analyzing method and analyzing device for operator state assessment reliability |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110270671A1 (en) * | 2010-05-03 | 2011-11-03 | Hsiu-Kheurn Tang | Predicting number of selections of advertisement using hierarchical Bayesian model |
-
2013
- 2013-11-26 CN CN201310611919.XA patent/CN103678886B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101354757A (en) * | 2008-09-08 | 2009-01-28 | 中国科学院地理科学与资源研究所 | Method for predicting dynamic risk and vulnerability under fine dimension |
| CN103218689A (en) * | 2013-04-23 | 2013-07-24 | 湖南工学院 | Analyzing method and analyzing device for operator state assessment reliability |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103678886A (en) | 2014-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103678886B (en) | A kind of satellite Bayesian network health based on ground test data determines method | |
| CN112765772B (en) | Power battery residual life prediction method based on data driving | |
| Cui et al. | Deep learning-based time-varying parameter identification for system-wide load modeling | |
| CN102789545B (en) | Based on the Forecasting Methodology of the turbine engine residual life of degradation model coupling | |
| Fan et al. | A novel machine learning method based approach for Li-ion battery prognostic and health management | |
| CN104268381B (en) | A kind of satellite failure diagnostic method based on AdaBoost algorithms | |
| CN112651167A (en) | Semi-supervised rolling bearing fault diagnosis method based on graph neural network | |
| CN110458230A (en) | A kind of distribution transforming based on the fusion of more criterions is with adopting data exception discriminating method | |
| CN106599920A (en) | Aircraft bearing fault diagnosis method based on coupled hidden semi-Markov model | |
| CN110222991B (en) | Metering device fault diagnosis method based on RF-GBDT | |
| CN111680875B (en) | Unmanned aerial vehicle state risk fuzzy comprehensive evaluation method based on probability baseline model | |
| CN105300692A (en) | Bearing fault diagnosis and prediction method based on extended Kalman filtering algorithm | |
| Niu et al. | Lebesgue sampling based deep belief network for lithium-ion battery diagnosis and prognosis | |
| CN116842459B (en) | Electric energy metering fault diagnosis method and diagnosis terminal based on small sample learning | |
| CN114290960B (en) | Method and device for acquiring battery health of power battery and vehicle | |
| CN113538353B (en) | Five-phase asynchronous motor rolling bearing fault diagnosis method based on single-channel diagram data enhancement and migration training residual error network | |
| CN114444582A (en) | Mechanical equipment fault diagnosis method based on convolutional neural network and Bayesian network | |
| CN109307852A (en) | A kind of method and system of the measurement error of determining electric automobile charging pile electric energy metering device | |
| CN116245227A (en) | Daily weather drought prediction method, device, storage medium and equipment | |
| CN114252103A (en) | Fusion power station operation fault prediction method | |
| CN111126477A (en) | Learning and reasoning method of hybrid Bayesian network | |
| CN105741184A (en) | Transformer state evaluation method and apparatus | |
| CN112232570A (en) | Forward active total electric quantity prediction method and device and readable storage medium | |
| CN114383845B (en) | Bearing composite fault diagnosis method based on embedded zero sample learning model | |
| Zuiev et al. | Analysis of control processes influence on UAV equipment classification veracity |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |