CN107273924A - The Fault Analysis of Power Plants method of multi-data fusion based on fuzzy cluster analysis - Google Patents
The Fault Analysis of Power Plants method of multi-data fusion based on fuzzy cluster analysis Download PDFInfo
- Publication number
- CN107273924A CN107273924A CN201710417836.5A CN201710417836A CN107273924A CN 107273924 A CN107273924 A CN 107273924A CN 201710417836 A CN201710417836 A CN 201710417836A CN 107273924 A CN107273924 A CN 107273924A
- Authority
- CN
- China
- Prior art keywords
- unstable
- stable
- fuzzy
- cluster
- mrow
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The present invention relates to a kind of Fault Analysis of Power Plants method of the multi-data fusion based on fuzzy cluster analysis, by being standardized to the multiple sensors sample data in an equipment, and optimal classification is carried out by fuzzy clustering, classification information is merged with D S evidence theories again, the confidence value of the equipment state can be described by obtaining one, form a kind of brand-new method for diagnosing faults.D S evidence theories, fuzzy algorithmic approach, clustering methodology are effectively and rationally combined, its comprehensive respective advantage, for complicated power plant's runtime, its diagnostic result is more accurate, efficient；Connecting is strong between each algorithm, and relevance is high；Comprehensive diagnos is carried out with multi-sensor data, diagnosis process is rapider, and obtained result is more accurate；Strong applicability, it is adaptable to various to have complexity, coupling, the system of randomness, can be used for power plant systems such as thermoelectricity, nuclear powers.
Description
Technical field
The present invention relates to a kind of failure judgment method, more particularly to a kind of multi-data fusion based on fuzzy cluster analysis
Fault Analysis of Power Plants method.
Background technology
With the planning and the development of science and technology of 13 power industry new industries, the fault diagnosis skill of power equipment
Art maturation more and reliable, the direct reason of Research on Fault Diagnosis Technology and development is to improve the precision and speed of diagnosis
Degree, reduction rate of false alarm and rate of failing to report, determine correct time and the position of failure generation.
Power plant is one complicated and to security requirement very high system, has to carry out efficiently and accurately in failure early stage
Fault diagnosis, can quickly and accurately make breakdown judge, to operation maintenance personnel provide relatively timely expert's property anticipate
See, help to find that failure in early stage, and help to repair failure, it is to avoid the accident of more serious mistake is produced, and loss is dropped to most
It is low.
The diagnosing information fusion fault research started in recent years, carries out from signals layer, sign layer and diagnosis decision-making level
Exploration work.Signal level fusion spatially and temporally on the information of multisensor is merged, expect to obtain accurately
Diagnostic result, this fusion distortion is minimum, but difficulty is relatively large.Reflect holographic spectrum, the reflection time-frequency of spatial synthesis information
The methods such as the wavelet analysis of integrated information, are all the methods that fusion treatment is carried out to signal.But these current methods are provided still
Qualitative relationships mainly between analysis result and failure, except lack it is intelligent in addition to, also lack effective quantitative Diagnosis side
Method, lacks novelty, and increases Diagnostic Time, and potential safety hazard is brought to system.
The content of the invention
The problem of existing the present invention be directed to plant information fusion fault diagnosis, it is proposed that one kind is based on fuzzy cluster analysis
Multi-data fusion Fault Analysis of Power Plants method, by being standardized to the multiple sensors sample data in an equipment
Processing, and optimal classification is carried out by fuzzy clustering, then merged classification information with D-S evidence theory, obtain one
The confidence value of the equipment state can be described, a kind of brand-new method for diagnosing faults is formed.Compared to former fuzzy neural
Network, the analysis of single-sensor D-S evidence theory or expert diagnostic system, sample can have with auto-adaptive increment formula fuzzy clustering
More preferable Accuracy and high efficiency, it is adaptable to various that there is complexity, coupling, the system of randomness..
The technical scheme is that：A kind of Fault Analysis of Power Plants side of the multi-data fusion based on fuzzy cluster analysis
Method, specifically includes following steps：
1) fuzzy clustering：The first step is standardization, i.e., the training sample of experiment is standardized, sets up its feature
Value matrix；Second step is cluster, that is, selectes suitable distance mode and calculate the fuzzy display matrix for obtaining data, calculated by clustering
Method carries out cluster training to sample data；3rd step finds optimal cluster result to determine optimal cluster level, obtains steady
Determine the node output network area under operating mode and unstable period, there is respective cluster centre in each region, and stability region is gathered
Class center (ic-stable,jc-stable) and unstable region cluster centre (ic-unstable,jc-unstable)；
2) basic probability assignment is constructed by index of stability e：
1 coefficient of reliability α is provided to every kind of sampled signal, the trusting degree to sampled result is represented, it is basic meeting
Under conditions of the definition of probability assignments, construction correspondence sampled signal ζ stabilizations or unstable basic probability assignment are：
M (Θ)=1- α
M (1) represents that to unstable basic probability assignment m (2) represents that, to stable basic probability assignment, m (Θ) is represented
To uncertain basic probability assignment；
For a new sampled signal sample, the network trained with fuzzy clustering carries out clustering, obtains one
Individual new interdependent node (i, j)；
Define instability index e1With index of stability e2Formula it is as follows：
e1=[(i-ic-unstable)2+(j-jc-unstable)2]/[(i-ic-stable)2+(j-jc-stable)2+(i-ic-unstable
)2+(j-jc-unstable)2]
e2=[(i-ic-stable)2+(j-jc-stable)2]/[(i-ic-stable)2+(j-jc-stable)2+(i-ic-unstable)2+
(j-jc-unstable)2]
Wherein, e1Represent the instability index of signal；e2Represent the index of stability of signal；
3) D-S evidence theory information fusion, obtains the diagnostic result after different sampled signal fusions：
If signal condition identification framework is Θ, Θ={ θ1,θ2,θ3As proposition, corresponding " stabilization ", " unstable ", " no
It is determined that " 3 primitives, using the sampled signal after data processing as the evidence for judging job stability, if m1And m2Respectively
The same identification framework θ of correspondencekOn two kinds of signals basic probability assignment, k=1,2,3, m1And m2Jiao member be respectively u11,
u21,...,up1And u12,u22,...,uq2If,Basic probability assignment after so synthesizing
m(θk) be defined as：
The beneficial effects of the present invention are：The Fault Analysis of Power Plants of multi-data fusion of the invention based on fuzzy cluster analysis
Method, D-S evidence theory, fuzzy algorithmic approach, clustering methodology are effectively and rationally combined, its comprehensive respective advantage, right
For complicated power plant's runtime, its diagnostic result is more accurate, efficient；Connecting is strong between each algorithm, and relevance is high；
Comprehensive diagnos is carried out with multi-sensor data, diagnosis process is rapider, and obtained result is more accurate；Strong applicability, is fitted
For it is various with complexity, coupling, randomness system, can be used for power plant systems such as thermoelectricity, nuclear powers.
Brief description of the drawings
Fig. 1 is clustering flow chart of the present invention；
Fig. 2 is the fuzzy cluster analysis topological diagram after the completion of present invention analysis.
Embodiment
This method comprises the following steps：
1. fuzzy clustering：The first step is standardization, i.e., the training sample of experiment is standardized, sets up its feature
Value matrix；Second step is cluster, that is, selectes suitable distance mode and calculate the fuzzy display matrix for obtaining data, by certain
Clustering algorithm is clustered to sample data；3rd step finds optimal cluster result to determine optimal cluster level.
The modeling process for completing fuzzy clustering algorithm is to realize three below step；
If overall cluster number of samples is n, B is designated as1,B2,...,Bn；Each cluster sample is to that should have m individual quantified
Index, be designated as 1,2 ..., m.
Step one：Construct eigenvalue matrix
Data are standardized first, extreme value standardization is used herein, expression is as follows：
Wherein, x ', x, xmax,xminRefer to the numerical value after certain criterion respectively, untreated original value, it is untreated before
Maximum, minimum value.
Assuming that characteristic value collections of the matrix T for the training sample index after standardization, T matrix sizes are m × n, wherein sample
This number is n, and index number is m.
x'stFor the characteristic value of the index s of the training sample t after standardization, s=1,2 ..., m, t=1,2 ..., n.
Fuzzy similarity matrix is resettled, the method set up herein is Cosin method.
rstRepresent xsz′(xs1′,xs2′,...,xsn') and xtz′(xt1′,xt2′,...,xtn') similarity degree, xsz' and
xtz' be standardization after training sample z index s characteristic value and training sample z index t characteristic value, z=1,2 ...,
n；S=1,2 ..., m；T=1,2 ..., m.
Step 2：Using self organizing neural network in Matlab to the automatic cluster function of different mode under different operating modes
State carry out preliminary classification.
It is trained herein using the training matrix after standardization by the Clustering tool case in Matlab, obtains obtaining
Different operating modes under corresponding output node position be labeled in distinct symbols in the grid chart of one 10 × 10.
Step 3：Cluster result is analyzed, and clustering flow chart as shown in Figure 1, in the sample process stage, uses statistical method
Enter row index and carry out primary election, then about subtract with rough set attribute and postsearch screening is carried out to index, standard then is carried out to index
Change is handled, and obtains pretreated data sample, and fuzzy clustering is carried out to sample, and such as Fig. 2 obtains steady working condition and unstable work
There is respective cluster centre node output network area under condition, each network area, and now we can obtain respective
Cluster centre, stability region cluster centre (ic-stable,jc-stable) and unstable region cluster centre (ic-unstable,
jc-unstable)。
2. the acquisition of basic probability assignment
Basic probability assignment represents support of the evidence to different target pattern, and different building methods, which is related to, last to be sentenced
Other result.In order that the process of basic probability assignment is more objective and simple, elementary probability is constructed herein by index of stability
Distribution.
In order to handle the information of fusion, it is necessary to which the degree of stability to bearing quantifies.Specifically calculating index of stability e's
When, observation cluster analysis result figure can be found that data out node of the steam turbine under steady working condition and unstable period is kept
In two regions, and it is 2 different regions, can so obtains respective stabilization and unstable region clustering center
(ic-stable,jc-stable) and (ic-unstable,jc-unstable), while providing 1 coefficient of reliability α, expression pair to vibration signal
The trusting degree of result of oscillation.So under conditions of the definition of basic probability assignment is met, construction vibration signal ζ (it is stable or
It is unstable) basic probability assignment is：
M (Θ)=1- α (5)
M (1) represents that to unstable basic probability assignment m (2) represents that, to stable basic probability assignment, m (Θ) is represented
To uncertain basic probability assignment.
For a new vibration signal sample, with the network trained above, (Fig. 2 is obtained by training data sample
Cluster centre network) carry out clustering, obtain a new interdependent node (i, j).
Define instability index e1With index of stability e2Formula it is as follows：
e1=[(i-ic-unstable)2+(j-jc-unstable)2]/[(i-ic-stable)2+(j-jc-stable)2+(i-ic-unstable
)2+(j-jc-unstable)2]
e2=[(i-ic-stable)2+(j-jc-stable)2]/[(i-ic-stable)2+(j-jc-stable)2+(i-ic-unstable)2+
(j-jc-unstable)2] (6) wherein, e1Represent the instability index of signal；e2Represent the index of stability of signal.
3. D-S evidence theory information fusion
In evidence theory, it is related to the concepts such as identification framework, Basic probability assignment function, belief function, verisimilitude function,
It is defined as follows：
Define the 1 exhaustive collection for setting all probable values of variable and be combined into Θ={ θ1,θ2,...,θn, if each element in Θ is
Mutually exclusive, then Θ is called identification framework (Frame of Discernment).
Define 2 and set Θ as identification framework, his one number m (A) ∈ of correspondence is all made to one subset A (proposition) for belonging to Θ of people
[0,1], and meet：
Then claim function m：2Θ→ [0,1], (2ΘFor Θ power set) it is 2ΘOn Basic probability assignment function BPA (Basic
Probability Assignment).IfAnd m (A) ≠ 0, then A is called a m burnt member.
Define 3 and set Θ as identification framework, m：2Θ→ [0,1] be identification framework Θ on basic probability assignment, then claim by
Defined function Bel (A) is the belief function (Belief function) on identification framework Θ.Belief function
Bel (A) represents total trusting degree to A.
Define 4 and set Θ as identification framework, m：2Θ→ [0,1] be identification framework Θ on basic probability assignment, then claim by
Defined function Pl (A) is the verisimilitude function (Plausibility function) on identification framework Θ, likelihood
Function Pl (A) represents the trusting degree for not negating to A.
So [Bel (A), Pl (A)] represents A indeterminacy section, that is, trusts interval, the lower limit estimation that the event that represents occurs
The possible range estimated with the upper limit.
If m1And m2Same identification framework θ is corresponded to respectivelykOn two kinds of signals basic probability assignment.m1And m2Jiao member point
Wei not u11,u21,...,up1And u12,u22,...,uq2If,It is basic after so synthesizing
Probability assignments m (θk) may be defined as：
By taking neutral signal as an example, if it is " not true that m (Θ) zd and m (Θ) wd corresponds to same identification framework Θ propositions respectively
Two basic probability assignments calmly ", burnt member is respectively (Θ) zd1,(Θ)zd2,(Θ)zd3,(Θ)zd4(Θ) wd1,(Θ)
wd2,(Θ)wd3,(Θ)wd4, wherein m (Θ) zd=[m (Θ) zd1m(Θ)zd2m(Θ)zd3m(Θ)zd4], m (Θ) wd=[m
(Θ)wd1m(Θ)wd2m(Θ)wd3m(Θ)wd4], ifWherein p, q=
1,2,3,4；Basic probability assignment after so synthesizing may be defined as：
In multi-sensor data-fusion system, each information source provides one group of evidence and proposition, and establishes one
Individual corresponding mass function.Therefore, each sensor is that information source is equivalent to an evidence body.Differentiate same
Under framework Θ, different evidence bodies is merged into rule and into a new evidence body by D-S, and calculate the likelihood of evidence body
Degree, finally with a certain Tactic selection rule, obtains last recognition result.
Experimental analysis data are derived from SIS (the Supervisory Information System of certain 600MW monoblock
Of Plant, supervisory information system) historical data base, wherein saving at equal intervals service data of the unit more than 1 year.Therefrom divide
2 groups of steady working condition and unstable period lower axle is not taken to shake with temperature data as sample data.
This paper identification frameworks include 3 primitives：{ " stabilization ", " unstable ", " uncertain " }；And will be after data processing
Axle shake information and temperature information as the evidence for judging steam turbine bearing stability.
The characteristics of according to bear vibration being with time pulsatile change, extract the average, standard deviation, peak-to-peak value 3 of vibration signal
Individual characteristic quantity, temperature is similarly.In addition, set forth herein a frequency domain character amount, the uniformity.Wherein, " uniformity " calculation formula is：
P=R/L (11)
In formula, R is later half frequency range internal power spectrum sum in signal power Power estimation figure；L is power spectrum signal
Estimate the power spectral value sum in the first half frequency range in figure.Wherein analyze and summarize the experience by mass data, will vibrate
α (zd)=0.85, α (wd)=0.70 is taken respectively with the coefficient of reliability α of temperature signal.
(1) the training original matrix of vibration
The training original matrix of temperatureDraw
The characteristic value collection of training sample index be the set of matrices of one 4 × 100.
Initial data is standardized, recycles fuzzy cluster analysis to carry out just the stability state of different mode
Step classification, planned network is constituted by 2 layers:Input layer and analysis layer.Input layer has 4 units, and extracted 4 are corresponded to respectively
Individual characteristic quantity, analysis layer is the neuron square formation of 1 two-dimentional 10 × 10.
Totally 100 groups of data sample for analysis, first 60 groups are the characterizing magnitudes under steady working condition, and latter 40 groups are unstable
Characterizing magnitudes under operating mode.Using every group of 4 extracted characteristic quantity as input, classified using fuzzy cluster analysis, passed through
Cross after 10000 calculating, the corresponding output node position of the different steady working condition finally obtained is labeled in 1 with different symbols
Open on 10 × 10 grid chart (Fig. 2).
The difference of characteristic quantity, same to cross fuzzy cluster analysis in stabilization and unstable period in vibration and temperature signal
They are gathered in the region of two different 2 (such as Fig. 2).For the characteristic value input under steady working condition, analysis classification knot
Fruit reaction is in the upper right side in region, and the characteristic value for unstable period is inputted, a left side of the analysis classification results reaction in region
Downside.
The characteristic value collection of training sample index is that T is 100 samples, the training sample index of the sample of 4 indexs
4 rows, 100 row are shown as in characteristic value collection, T matrixes.
Thus cluster centre is obtained, that is, vibrates stability region Centroid (2.572,36.791), unstable region is vibrated
Centroid (1.137,17.334), temperature stabilization regional center node (2.241,41.442), temperature unstable region center
Node (1.325,22.593).
(2) it is 4 groups of test sample characteristic quantities of two kinds of signals such as table 1.
Table 1
According to table 1, for a new sample data, handled by above-mentioned data normalization, with what is trained above
Network carries out clustering to it, obtains a new node (i, j), is calculated respectively respectively according to formula (4), formula (5) and formula (6)
The basic probability assignment value of evidence body, it can be deduced that as follows：
Uncertainty probability distribution m (Θ) zd=[0.15 0.15 0.15 0.15] of vibration
Uncertainty probability distribution m (Θ) wd=[0.30 0.30 0.30 0.30] of temperature
The basic probability assignment of vibration signal
The basic probability assignment of temperature signal
Each evidence basic probability assignment value as shown in table 2 and independent role diagnostic result, i.e., give evidence list herein
Diagnostic result under solely acting on.
Table 2
(3) data fusion is being carried out according to formula (10) with D-S evidence theories, is obtaining vibration signal and temperature signal fusion
Diagnostic result after effect, it is as follows：
Uncertainty probability distribution m (Θ) fusion=[0.051 0.057 0.062 0.055] after fusion
Probability of stability m (2) funsion=[0.864 0.739 0.235 0.113] after fusion
Unstable probability m (1) fusion=[0.083 0.206 0.697 0.831] after fusion
The comprehensive diagnos result that each evidence fusion is acted on as shown in table 3：
Table 3
From table 2 and table 3 as can be seen that according to diagnosis decision rule, except first group during each evidence independent role before fusion
Diagnostic result is correctly outer, and other 3 groups can not all provide clear and definite conclusion, and the vibration of instruction sheet one or temperature signal are stable for bearing
Property diagnosis credibility it is relatively low and uncertain higher so that operations staff can not make to the safety and stability situation of bearing
Effectively judge.And the diagnostic result after the fusion of each evidence body with practical stability state consistency, and uncertainty greatly reduces,
So that its credibility is significantly improved.
Claims (1)
1. a kind of Fault Analysis of Power Plants method of the multi-data fusion based on fuzzy cluster analysis, it is characterised in that specifically include
Following steps：
1) fuzzy clustering：The first step is standardization, i.e., the training sample of experiment is standardized, sets up its characteristic value square
Battle array；Second step is cluster, that is, selectes suitable distance mode and calculate the fuzzy display matrix for obtaining data, pass through clustering algorithm pair
Sample data carries out cluster training；3rd step finds optimal cluster result, obtains stablizing work to determine optimal cluster level
Node output network area under condition and unstable period, each region has in respective cluster centre, stability region cluster
The heart (ic-stable,jc-stable) and unstable region cluster centre (ic-unstable,jc-unstable)；
2) basic probability assignment is constructed by index of stability e：
1 coefficient of reliability α is provided to every kind of sampled signal, the trusting degree to sampled result is represented, is meeting elementary probability
Under conditions of the definition of distribution, construction correspondence sampled signal ζ stabilizations or unstable basic probability assignment are：
<mrow>
<mi>m</mi>
<mrow>
<mo>(</mo>
<mi>&zeta;</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<msub>
<mi>e</mi>
<mi>&zeta;</mi>
</msub>
<mrow>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>&zeta;</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mn>2</mn>
</msubsup>
<msub>
<mi>e</mi>
<mi>&zeta;</mi>
</msub>
</mrow>
</mfrac>
<mi>&alpha;</mi>
</mrow>
M (Θ)=1- α
M (1) represents that to unstable basic probability assignment m (2) represents that, to stable basic probability assignment, m (Θ) is represented to not
The basic probability assignment of determination；
For a new sampled signal sample, the network trained with fuzzy clustering carries out clustering, obtains one newly
Interdependent node (i, j)；
Define instability index e1With index of stability e2Formula it is as follows：
e1=[(i-ic-unstable)2+(j-jc-unstable)2]/[(i-ic-stable)2+(j-jc-stable)2+(i-ic-unstable)2+(j-
jc-unstable)2]
e2=[(i-ic-stable)2+(j-jc-stable)2]/[(i-ic-stable)2+(j-jc-stable)2+(i-ic-unstable)2+(j-
jc-unstable)2]
Wherein, e1Represent the instability index of signal；e2Represent the index of stability of signal；
3) D-S evidence theory information fusion, obtains the diagnostic result after different sampled signal fusions：
If signal condition identification framework is Θ, Θ={ θ1,θ2,θ3As proposition, { " stabilization ", " unstable " be not " true for correspondence
3 primitives calmly " }, using the sampled signal after data processing as the evidence for judging job stability, if m1And m2It is right respectively
Answer same identification framework θkOn two kinds of signals basic probability assignment, k=1,2,3, m1And m2Jiao member be respectively u11,
u21,...,up1And u12,u22,...,uq2If,Basic probability assignment after so synthesizing
m(θk) be defined as：
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710417836.5A CN107273924B (en) | 2017-06-06 | 2017-06-06 | Multi-data fusion power plant fault diagnosis method based on fuzzy clustering analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710417836.5A CN107273924B (en) | 2017-06-06 | 2017-06-06 | Multi-data fusion power plant fault diagnosis method based on fuzzy clustering analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107273924A true CN107273924A (en) | 2017-10-20 |
| CN107273924B CN107273924B (en) | 2020-05-15 |
Family
ID=60064496
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710417836.5A Active CN107273924B (en) | 2017-06-06 | 2017-06-06 | Multi-data fusion power plant fault diagnosis method based on fuzzy clustering analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107273924B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108416309A (en) * | 2018-03-14 | 2018-08-17 | 揭阳职业技术学院 | A kind of intelligence sensor multiple faults transducing signal reconstructing method |
| CN108446458A (en) * | 2018-03-01 | 2018-08-24 | 西北工业大学 | A kind of Weighted Fusion method for diagnosing faults based on DS evidence theories |
| CN108710756A (en) * | 2018-05-18 | 2018-10-26 | 上海电力学院 | The method for diagnosing faults of lower multicharacteristic information Weighted Fusion is analyzed based on spectral clustering |
| CN108761263A (en) * | 2018-05-24 | 2018-11-06 | 深圳大图科创技术开发有限公司 | A kind of fault diagnosis system based on evidence theory |
| CN110146120A (en) * | 2019-05-19 | 2019-08-20 | 浙江大学 | A kind of sensor fault diagnosis method and system |
| CN111193742A (en) * | 2019-12-31 | 2020-05-22 | 广东电网有限责任公司 | D-S evidence theory-based power communication network anomaly detection method |
| CN112528554A (en) * | 2020-11-17 | 2021-03-19 | 中国运载火箭技术研究院 | Data fusion method and system suitable for multi-launch multi-source rocket test data |
| CN113159138A (en) * | 2021-04-01 | 2021-07-23 | 北京市燃气集团有限责任公司 | Gas boiler fault diagnosis method and device based on data fusion |
| CN117349598A (en) * | 2023-12-04 | 2024-01-05 | 小舟科技有限公司 | Electroencephalogram signal processing method and device, equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101726688A (en) * | 2009-09-17 | 2010-06-09 | 浙江大学 | Method for diagnosing multi-data-source information fusion-based power system fault |
| US20130332011A1 (en) * | 2012-06-06 | 2013-12-12 | Harris Corporation | Wireless engine monitoring system and associated engine wireless sensor network |
| CN103886316A (en) * | 2014-02-20 | 2014-06-25 | 东南大学 | Combustion monitoring and diagnosis method based on feature extraction and fuzzy C-means cluster |
| CN104537220A (en) * | 2014-12-18 | 2015-04-22 | 华北电力大学（保定） | Fault diagnosis method based on principal component analysis and D-S evidence theory |
| CN104655948A (en) * | 2013-11-21 | 2015-05-27 | 国家电网公司 | Novel multistage fault diagnosis method for power transformer |
| CN105427200A (en) * | 2015-11-05 | 2016-03-23 | 上海电力学院 | Power plant fault diagnosis method based on fuzzy theory and data fusion |
-
2017
- 2017-06-06 CN CN201710417836.5A patent/CN107273924B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101726688A (en) * | 2009-09-17 | 2010-06-09 | 浙江大学 | Method for diagnosing multi-data-source information fusion-based power system fault |
| US20130332011A1 (en) * | 2012-06-06 | 2013-12-12 | Harris Corporation | Wireless engine monitoring system and associated engine wireless sensor network |
| CN104655948A (en) * | 2013-11-21 | 2015-05-27 | 国家电网公司 | Novel multistage fault diagnosis method for power transformer |
| CN103886316A (en) * | 2014-02-20 | 2014-06-25 | 东南大学 | Combustion monitoring and diagnosis method based on feature extraction and fuzzy C-means cluster |
| CN104537220A (en) * | 2014-12-18 | 2015-04-22 | 华北电力大学（保定） | Fault diagnosis method based on principal component analysis and D-S evidence theory |
| CN105427200A (en) * | 2015-11-05 | 2016-03-23 | 上海电力学院 | Power plant fault diagnosis method based on fuzzy theory and data fusion |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108446458A (en) * | 2018-03-01 | 2018-08-24 | 西北工业大学 | A kind of Weighted Fusion method for diagnosing faults based on DS evidence theories |
| CN108446458B (en) * | 2018-03-01 | 2019-10-22 | 西北工业大学 | A kind of Weighted Fusion rotor method for diagnosing faults based on DS evidence theory |
| CN108416309A (en) * | 2018-03-14 | 2018-08-17 | 揭阳职业技术学院 | A kind of intelligence sensor multiple faults transducing signal reconstructing method |
| CN108416309B (en) * | 2018-03-14 | 2022-02-18 | 揭阳职业技术学院 | Multi-fault sensing signal reconstruction method for intelligent sensor |
| CN108710756A (en) * | 2018-05-18 | 2018-10-26 | 上海电力学院 | The method for diagnosing faults of lower multicharacteristic information Weighted Fusion is analyzed based on spectral clustering |
| CN108761263A (en) * | 2018-05-24 | 2018-11-06 | 深圳大图科创技术开发有限公司 | A kind of fault diagnosis system based on evidence theory |
| CN110146120B (en) * | 2019-05-19 | 2021-08-03 | 浙江大学 | Sensor fault diagnosis method and system |
| CN110146120A (en) * | 2019-05-19 | 2019-08-20 | 浙江大学 | A kind of sensor fault diagnosis method and system |
| CN111193742A (en) * | 2019-12-31 | 2020-05-22 | 广东电网有限责任公司 | D-S evidence theory-based power communication network anomaly detection method |
| CN112528554A (en) * | 2020-11-17 | 2021-03-19 | 中国运载火箭技术研究院 | Data fusion method and system suitable for multi-launch multi-source rocket test data |
| CN113159138A (en) * | 2021-04-01 | 2021-07-23 | 北京市燃气集团有限责任公司 | Gas boiler fault diagnosis method and device based on data fusion |
| CN117349598A (en) * | 2023-12-04 | 2024-01-05 | 小舟科技有限公司 | Electroencephalogram signal processing method and device, equipment and storage medium |
| CN117349598B (en) * | 2023-12-04 | 2024-03-08 | 小舟科技有限公司 | Electroencephalogram signal processing method and device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107273924B (en) | 2020-05-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107273924A (en) | The Fault Analysis of Power Plants method of multi-data fusion based on fuzzy cluster analysis | |
| Garg | A hybrid GA-GSA algorithm for optimizing the performance of an industrial system by utilizing uncertain data | |
| US7941701B2 (en) | Fuzzy classification approach to fault pattern matching | |
| KR101825881B1 (en) | Method of managing a manufacturing process and system using the same | |
| CN101957889B (en) | Selective wear-based equipment optimal maintenance time prediction method | |
| WO2011034805A1 (en) | Supervised fault learning using rule-generated samples for machine condition monitoring | |
| CN106250288A (en) | Root alarm analysis and identification method based on data mining | |
| JP2007249997A (en) | Method and system for monitoring industrial process | |
| CN103678881B (en) | Composite fault diagnosis method based on combination of artificial immunity and evidence theory | |
| CN112183643B (en) | Hard rock tension-shear fracture identification method and device based on acoustic emission | |
| Charbonnier et al. | A weighted dissimilarity index to isolate faults during alarm floods | |
| CN106950945A (en) | A kind of fault detection method based on dimension changeable type independent component analysis model | |
| CN106796157A (en) | Abnormal method in detection distribution network, particularly drinking water distribution network | |
| CN106682835A (en) | Data-driven complex electromechanical system service quality state evaluation method | |
| CN109034076A (en) | A kind of automatic clustering method and automatic cluster system of mechanical fault signals | |
| Kuravsky et al. | New approaches for assessing the activities of operators of complex technical systems | |
| CN110851422A (en) | Data anomaly monitoring model construction method based on machine learning | |
| CN117172414A (en) | Building curtain construction management system based on BIM technology | |
| Kuravsky et al. | Detecting abnormal activities of operators of complex technical systems and their causes basing on wavelet representations | |
| CN117196159A (en) | Intelligent water service partition metering system based on Internet big data analysis | |
| CN109345060A (en) | A kind of product quality characteristics Error Tracing & analysis method based on multi-source perception | |
| CN116771576A (en) | Comprehensive fault diagnosis method for hydroelectric generating set | |
| CN110337640A (en) | Method and system for problem alert polymerization | |
| CN112836719B (en) | Indicator diagram similarity detection method integrating two classifications and triplets | |
| CN111126477A (en) | Learning and reasoning method of hybrid Bayesian network |
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 |