CN106599445B - The method of geometry of rotating machinery health status prediction based on single branch truncated data - Google Patents
The method of geometry of rotating machinery health status prediction based on single branch truncated data Download PDFInfo
- Publication number
- CN106599445B CN106599445B CN201611132579.2A CN201611132579A CN106599445B CN 106599445 B CN106599445 B CN 106599445B CN 201611132579 A CN201611132579 A CN 201611132579A CN 106599445 B CN106599445 B CN 106599445B
- Authority
- CN
- China
- Prior art keywords
- truncated data
- truncated
- time
- rotating machinery
- 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
Classifications
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/17—Mechanical parametric or variational design
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Geometry (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The method of geometry for the rotating machinery health status prediction based on single branch truncated data that the invention discloses a kind of, including:By being monitored in real time to rotating machinery vibrating, numerous truncation number strong points are obtained;The dimension-reduction treatment based on time continuity is carried out by numerous truncation number strong points high dimensional feature to numerous truncation number strong points, obtains the truncated data health manifold space with multiple truncated data point features based on Annual distribution;According to the time relationship between multiple truncated data point features in truncated data health manifold space, link process is carried out to multiple truncated data point features, forms the health status evolutional path of time-based truncated data;It is fitted processing by the health status evolutional path to truncated data, obtains the life value of the truncated data of rotating machinery vibrating.The present invention solves the technical issues of ' neighbour structure ' is destroyed with ' overall structure ';Thus preferably truncated data can be utilized to predict rotating machinery health status.
Description
Technical field
The present invention relates to electric powder prediction, especially a kind of rotating machinery health status based on single branch truncated data is pre-
The method of geometry of survey.
Background technology
With social progress and development, people do not only need to know the current state of monitored target, are also required to simultaneously
Understand the state in monitored target future, and then carry out follow-up work, effectively to control risk.However, the essence of prediction is
Seek the dynamic evolution rule in things degenerative process.In recent years, when a large amount of fail data is available, many data-driven sides
Method achieves preferable prediction effect, such as:Smart network, Hidden Markov Model.In addition, some methods attempt to solve it is dilute
Dredge the forecasting problem under the conditions of data or truncated data.However, these research work are often based upon the more considerable history of quantity
Sample data (sparse or block).
Under the premise of essential change does not occur for performance degradation rule, blocking for individual subject data of prediction does not interfere with it
The similitude of its individual manifold structure and sample totality manifold structure.However, when all samples totally be all truncated when (i.e. without times
The completely cut off data mode of what full longevity data), the healthy manifold space of research object is constructed, and ensure its knot as far as possible
The integrality of structure is in turn the difficult point and matter of utmost importance for having to solve for predicting.
Analytic explanation is carried out with the full longevity data instance being made of the 520 of rolling bearing characteristic value point.Utilize manifold
Learning method blocks degree structure healthy manifold space according to different.With being continuously increased for bearing high dimensional feature data volume, structure
The healthy manifold space structure built out is closer to full longevity data entirety manifold structure.Naturally, for being truncated to only have 200
The manifold structure of high dimensional feature data point and final architectural difference are very big.
The present invention using in the manifold space built neighborhood geodesic curve distance and accumulation geodesic curve distance it is strong as quantitative analysis
Health manifold space neutral can fail the Measure Indexes of prediction.Fig. 2 show part different data block under status condition (200,
300,400,500,510,515,520) geodesic curve distance and accumulation geodesic curve is apart from coordinate diagram.From manifold space health status
In the audio-visual picture of Measure Indexes, we can obtain conclusion as before, while provide neighborhood geodesic curve distance and accumulation
Quantitive error analysis between geodesic curve distance.Firstly the need of pointing out, quantitive error analysis here is all with full longevity data sample
The manifold of foundation neighborhood geodesic curve distance spatially and accumulation geodesic curve distance are that reference standard is calculated.It is calculated
Correlated results is as shown in table 1 (by taking neighborhood geodesic curve distance as an example, the result of calculation for accumulating geodesic curve distance is similar).N is in table 1
Truncated data amount M is the data volume that error calculation is carried out within the scope of N.(have light blue with the data instance corresponding to (500,400)
The data of color shade), it indicates when building manifold space with 500 truncated datas, the neighborhood geodesic curve of preceding 400 data points
Absolute error and relative error between distance and preceding 400 neighborhoods geodesic curve distance of full longevity data.
From the data in the table, for any one numerical result for showing meaning, whether relative error or exhausted
To error, error amount is all reducing from top to bottom, illustrates (to block changing for situation with being continuously increased for truncated data sample size
It is kind), the manifold space structure similarity that obtained manifold space structure is built with full longevity data is constantly promoted.
1 truncated data manifold structure error analysis of table
Wherein, AE is absolute error, and RE is relative error
As the above analysis, different data is blocked under state, and the structure of constructed manifold is different, obtains corresponding neighborhood
Geodesic curve distance and accumulation geodesic distance are different.It is then desired to be directed to truncated data state, design and development is a kind of effective strong
Health manifold construction method keeps the manifold space structure obtained through it and full longevity data manifold space structure difference as small as possible, i.e.,
How so that the manifold acquired under the conditions of truncated data is exactly the subset for obtaining manifold under life-cycle data qualification, for rear
Continuous truncated data prediction.
(such as different manifold construction methods:Manifold learning arithmetic), core concept is in global optimization condition
Under, as keep certain geometrical relationship or geometrical property constant as possible, such as:ISOMAP keeps geodesic distance between points not
Become, LLE holding local geometric linear relationships, LTSA keeps local tangent space projection coordinate relationship constant etc..In this way, when being cut
When the new sampled point having no progeny, the manifold that the addition of these new data points can construct former truncated data brings following problem:
1)' neighbour structure ' destroys:The addition of new data point set so that ' neighborhood ' of initial truncated data collection structure occurs
Essence variation, so causes, being the data point of ' neighborhood ' relationship in original truncated data may no longer keep original
' neighborhood ' relationship, this certainly will lead to the reallocation of local geometric relationship, and then change the structure that original blocks manifold;
2)' overall structure ' is destroyed:Block the addition of rear new data point set so that new change occurs for ' neighborhood ' geometrical relationship
Change (even if ' neighborhood ' relationship remains unchanged before), meanwhile, it is this new under the constraints of global optimizing
Geometrical relationship certainty adds global optimizing process, this optimization process necessarily makes the manifold obtained by former truncated data empty
Between in data point (embedded point) change, in turn, change the space structure of former truncated data manifold on the whole.
Invention content
The method for the rotating machinery health status prediction based on single branch truncated data that the object of the present invention is to provide a kind of, root
According to the data point in the time relationship structure manifold space of each data point in truncated data, it is broken to solve above-mentioned ' neighbour structure '
The technical issues of bad and ' overall structure ' is destroyed.
The present invention it is a kind of based on single branch truncated data rotating machinery health status prediction method of geometry include:
By being monitored in real time to rotating machinery vibrating, numerous truncation number strong points are obtained;
It is carried out based on time continuity by numerous truncation number strong points high dimensional feature to numerous truncation number strong points
Dimension-reduction treatment obtains the truncated data health manifold space with multiple truncated data point features based on Annual distribution;
According to the time relationship between multiple truncated data point features in the truncated data health manifold space, to described
Multiple truncated data point features carry out link process, form the health status evolutional path of time-based truncated data;
It is fitted processing by the health status evolutional path to the truncated data, obtains the rotating machinery vibrating
Truncated data life value.
Wherein, by carrying out feature extraction processing to truncated data point, truncation number strong point high dimensional feature is obtained.
Wherein, the high dimensional feature by numerous truncation number strong points carries out the dimensionality reduction based on time continuity
Processing, obtaining the truncated data health manifold space with multiple truncated data point features based on Annual distribution includes:
Numerous truncated data point features are calculated using TRIM algorithms, by numerous truncated data point features
It is reduced to multiple truncated data point features;
According to time relationship, Annual distribution processing is carried out to the multiple truncated data point feature, forms the truncation number
According to healthy manifold space;
Wherein, the TRIM algorithms are to continue related Isometric Maps algorithm the time.
Wherein, in the healthy manifold space according to the truncated data between multiple truncated data point features when
Between relationship, to the multiple truncated data point carry out link process include:
Neighborhood survey is carried out respectively to all two adjacent truncated data point features in the healthy manifold space of truncated data
Ground linear distance NGDs processing obtains all two adjacent truncated data point features based on time relationship in healthy manifold space
NGDs;
The NGDs of two all adjacent truncated data point features is subjected to accumulated process, is formed in healthy manifold space
Accumulate geodesic curve distance CGDs, to obtain time-based truncated data truncated data health status evolutional path.
Wherein, described that place is fitted by the health status evolutional path of the truncated data to the truncated data
Reason, the life value for obtaining the truncated data of the rotating machinery vibrating include:
The extension based on future time is carried out to the CGDs to handle, obtain for prediction and carry using Gaussian function
The extension CGDs values of future time.
Wherein, the health stream by the truncated data carries out process of fitting treatment, obtains the rotating machinery and shakes
The life value of dynamic truncated data further includes:
The extension CGDs values are compared with preset CGDs threshold values;
According to comparison result, the life value of the truncated data of the rotating machinery vibrating is determined.
Wherein, described according to comparison result, determine that the truncated data life value of the rotating machinery vibrating includes:
If the extension CGDs values are more than the CGDs threshold values, future of the CGDs values equal to the CGDs threshold values will be extended
Time determines the life value of the truncated data of the rotating machinery vibrating.
' truncated data ' of the present invention is meant:All Condition Monitoring Datas when end-of-life/failure occurs are not arrived.
The method have the benefit that solve ' neighbour structure ' existing for classical manifold space construction process with
The technical issues of ' overall structure ' is destroyed, so as to preferably truncated data be utilized to predict rotating machinery health status.
Description of the drawings
Fig. 1 is the schematic diagram of the method for the rotating machinery health status prediction based on single branch truncated data of the present invention;
Fig. 2 is the schematic diagram of neighborhood geodesic curve distance and accumulation geodesic curve distance that different data is blocked under state;
Fig. 3 is the geodesic schematic diagram on swiss roll;
Fig. 4 is the schematic diagram of the time domain index extracted by bear vibration data;
Fig. 5 is the schematic diagram of the energy indexes extracted by bear vibration data;
Fig. 6 a are the schematic diagrames of the state evolution manifold under the conditions of the different data being calculated through TRIM;
Fig. 6 b be based on neighborhood geodesic curve distance under TRIM method difference truncated data status conditions and accumulation geodesic curve away from
From schematic diagram;
Fig. 7 be based on TRIM methods obtain difference block in state health manifold space accumulate geodesic curve apart from affine pass
System.
Fig. 8 is the fitting for accumulating geodesic curve distance Curve in healthy manifold space based on the single branch truncated datas of Gaussian 6
The schematic diagram of prediction.
Specific implementation mode
The method that Fig. 1 shows a kind of rotating machinery health status prediction based on single branch truncated data of the present invention, should
Method includes:
The present invention it is a kind of based on single branch truncated data rotating machinery health status prediction method of geometry include:
By being monitored in real time to rotating machinery vibrating, numerous truncation number strong points are obtained;
It is carried out based on time continuity by numerous truncation number strong points high dimensional feature to numerous truncation number strong points
Dimension-reduction treatment obtains the truncated data health manifold space with multiple truncated data point features based on Annual distribution;
According to the time relationship between multiple truncated data point features in the truncated data health manifold space, to described
Multiple truncated data point features carry out link process, form the health status evolutional path of time-based truncated data;
It is fitted processing by the health status evolutional path to the truncated data, obtains the rotating machinery vibrating
Truncated data life value.
Wherein, by carrying out feature extraction processing to truncated data point, truncation number strong point high dimensional feature is obtained.
Wherein, the high dimensional feature by numerous truncation number strong points carries out the dimensionality reduction based on time continuity
Processing, obtaining the truncated data health manifold space with multiple truncated data point features based on Annual distribution includes:It utilizes
TRIM algorithms calculate numerous truncated data point features, and numerous truncated data point features are reduced to multiple sections
Disconnected data point feature;According to time relationship, Annual distribution processing is carried out to the multiple truncated data point feature, forms described cut
The healthy manifold space of disconnected data;Wherein, the TRIM algorithms are to continue related Isometric Maps algorithm the time.
Wherein, it is closed according to the time between multiple truncated data point features in the healthy manifold space of the truncated data
System, carrying out link process to the multiple truncated data point feature includes:
Neighborhood survey is carried out respectively to all two adjacent truncated data point features in the healthy manifold space of truncated data
Ground linear distance NGDs processing (that is, according to time relationship, calculates separately between all two adjacent truncated data point features most
Short distance), obtain the NGDs of all two adjacent truncated data point features based on time relationship in healthy manifold space
(that is, foring the line of the shortest distance between multiple two neighboring truncated data point features);
The NGDs at two all adjacent truncation number strong points is subjected to accumulated process (that is, by multiple two neighboring truncation numbers
It is attached processing according to the line of shortest distance between points), accumulation geodesic curve distance CGDs is formed in healthy manifold space, to
The health status evolutional path of time-based truncated data is obtained (that is, being closed based on the time in healthy manifold space
All truncated data point features are connected into the line segment of truncated data feature by system).
Wherein, described that place is fitted by the health status evolutional path of the truncated data to the truncated data
Reason, the life value for obtaining the truncated data of the rotating machinery vibrating include:The CGDs is based on using Gaussian function
The extension of future time is handled, and obtains extension CGDs values for prediction and with future time.
Wherein, the health stream by the truncated data carries out process of fitting treatment, obtains the rotating machinery and shakes
The life value of dynamic truncated data further includes:The extension CGDs values are compared with preset CGDs threshold values;According to comparing
As a result, determining the life value of the truncated data of the rotating machinery vibrating.Wherein, described according to comparison result, determine described in
The truncated data life value of rotating machinery vibrating includes:If the extension CGDs values are more than the CGDs threshold values, will extend
CGDs values determine the life value of the truncated data of the rotating machinery vibrating equal to the future time of the CGDs threshold values.
The thinking and particular technique point of the solution technical problem of the present invention are carried out specifically with reference to Fig. 2 to Fig. 8
It is bright.
4.1.1 the thinking solved the problems, such as
1) whether truncated data collection or full longevity data set, manifold space construction process of the same race are all to maintain data point
Between certain geometrical relationship or characteristic it is constant, this geometrical relationship or characteristic are present in naturally in former high-dimensional feature space.
Thus when ' neighborhood ' relationship is fixed, the geometrical property between this point kept under the conditions of truncated data and full longevity data qualification
Under geometrical property between obtained these points should differ an affine transformation;
2) existing manifold space construction process is mainly used in the fields such as pattern-recognition, classification, assessment and image procossing.
Manifold structure in these areas is all the integral transformation of all data point sets, few the problem of increasing point set newly occur.However
When there is new point set to add, these points are also considered the generalized point without any particularity, and are added to and are totally divided again
Class --- this process is invalid to predicting;
3) for the forecasting problem in health management system arranged, the structure of manifold has following property:
● it predicts to be permanently present inevitable contact with the time;
● by time series theory it is found that prediction result only (office related to a certain range of partial data
Domain);
● the introducing of new data point will not have an impact historical data and data structure;
It follows that the property of certainty and prediction theory itself based on healthy manifold space structure, it can be by drawing
Angle of incidence parameter and fixed manifold of higher dimension space ' neighborhood ' range, and then determine and cure its ' neighborhood ' structure, it is above-mentioned to solve
The problems such as ' neighbour structure ' is destroyed with ' overall structure ', important foundation is provided for performance degradation prognosis under truncation condition.
4.1.2 the healthy manifold developing algorithm towards completely cut off data
The present invention will combine the above-mentioned thinking solved the problems, such as, improves classical manifold learning arithmetic ISOMAP, will tie up simultaneously
The global characteristics for keeping truncated data can be failed the needs of prediction with meeting healthy manifold space neutral under the conditions of truncated data.
Through being improved to ISOMAP algorithms, time consistent correlation Isometric Maps (Time-continuous is obtained
Relevant isometric mapping, TRIM) algorithm description is as follows:
On the basis of the theory of ISOMAP methods is with algorithm in 3.3.1, ISOMAP algorithms are improved, obtaining the time holds
Algorithm description is such as continuous correlation Isometric Maps (Time-continuous relevant isometric mapping, TRIM)
Under:
(1) m- Neighborhood Graph (t-G) structure when
In view of the time continuity that time series and state are degenerated, i-ththThe when m- Neighborhood Graph of point, t-G are defined as
Point ithAnd ithIt is preceding close to ithThe figure that is constituted of K point.Wherein K determines that any limit of .t-G can indicate by time series relationship
For d (xi,xj).
(2) time correlation path minimum distance calculation between any two points
In t-G, i-ththThe most short of time correlation path between point and any other point j can be described as:Work as i, j is shared
Same one side, and j is a point in t-G, i.e. j is ithPreceding K Neighbor Points for the moment, dt-G(xi,xj)=d (xi,xj);It is other
In the case of:dt-G(xi,xj)=∞ are accordingly, for arbitrary point m=1,2 ..., i-K, and i-ththThe distance between point and point m can tables
It is shown as:In turn, distance matrix is represented by Dt-G={ dt-G(xi,xj)}.
(3)Dt-GD dimensions embedded calculate
It calculates
Enable λ1,...,λdFor the preceding d maximum value of matrix H characteristic value, and μ1,...,μdFor its corresponding feature vector.Then
Dt-GD dimension insertion be represented byWherein, UT=[μ1,...,μd].
4.2 other relevant methods
4.2.1 the geometry measurement that health status is degenerated
(1) manifold geodesic curve spatially and geodesic distance
Art of mathematics, especially Differential Geometry field, geodesic curve (Geodesic) are one of " straight line " in " curved space "
As change concept.If the communication is to measure the Levi-Civita communications from shifting onto by Li Man, geodesic distance
(Geodesic distance) shortest distance of distance between any two points in manifold space.
Fig. 3 illustrates the geodesic curve (a) on two-dimentional " swiss roll " for any two points p and q, relative to traditional European
Distance, geodesic distance can more accurately reflect the inherent similitude between 2 points;(b) former in " swiss roll " plane of expansion
First geodesic curve is correspondingly mapped as straight line.
The approximate calculation of geodesic distance on flexure plane is important computational geometry problem, the field being related to include from
Calculate graphics, medical image, geophysics, to robot motion planning and navigation etc..The side of graph theory is used in this research
Method approximate calculation geometric distance, this method on Science were proposed and were delivered in 2000 by Tenenbaum et al..
(1) healthy manifold (that is, health status evolutional path of truncated data)
Assuming that:The health of Mechatronic Systems is in continuous downturn state.
Gather defined in function space:
M=f (t, X) | and X ∈ Θ }, X=(x1,x2,…,xn) (2.11)
Wherein, Θ is theorem in Euclid space RnOpener, i.e.,:By assuming it is found that establishing one-to-one correspondence between M and Θ
Relationship.This correspondence to induce the topological structure with Θ homeomorphisms on M.Consider M on any point F=f (t, X) and
A neighborhood U of FF, define UFTo RnMapping phi:φ (F)=X.In this way,All such neighborhood UFConstitute M one opens Covering, as soon as φ is a differomorphism mapping, this establishes a differential structrue on M, and forms Differential Manifold, i.e.,:Health Manifold of states, referred to as:Healthy manifold (Health Manifold).Wherein, X claims the natural coordinates of the manifold, can also determine certainly
The other coordinates of justice.
(2) neighborhood geodesic curve distance NGDs
Geodesic curve and geodesic distance have reacted healthy manifold spatially the distance between any two points relationship, embedding for higher-dimension
Enter manifold space neutral energy forecasting problem, neighborhood geodesic curve distance indicates in healthy manifold space that object is in evolution degenerative process
In any state point and its neighborhood (determined by the order dependent radius of neighbourhood of predicted time) between state point apart from degree of a relation amount,
To describe healthy manifold space local geometry.
(3) accumulation geodesic curve distance CGDs
It accumulates geodesic curve distance and indicates healthy manifold space any state point and (the prediction Object Evolution decline of original state point
State point in the process) between geodesic curve distance, be in manifold of higher dimension space free position deviate original state measurement.
4.2.2 radiation transformation and radiation relationship structure
By Such analysis it is found that healthy manifold structure of the electromechanical object under individual difference and the combined influence of variable working condition exists
It remains certain similitude in a way, and answers the healthy intrinsic dimension of manifold having the same and spatial context dimension.Cause
This, needs by means of the similitude between healthy manifold structure, is changed with operating mode using corresponding technological means reduction individual difference
The manifold structure difference of introducing, performance degradation prognosis under the conditions of to solve the problems, such as variable working condition in healthy manifold space.
Due to affine transformation has ' line segment is become line segment, and keep line segment point than constant ' and ' keep conllinear three
The simple ratio of point is constant ' etc. properties so that affine transformation, which has, keeps the space ability constant with respect to geometrical property.Due to
2 points of health status is identical on arbitrary ideal life line on health status space-time diagram, b points as shown in Figure 5 and d points.Therefore, real
In the engineering of border, for fixed failure relative threshold thre_, the lifeline of each object in world map Origin And Destination it is strong
Health state coordinate value is identical, ob and od as shown in Figure 5.Keeping each electromechanical object geometry in its healthy manifold space special herein
Property it is geostationary under the premise of, the affine transformation in manifold space is carried out, with the accumulation geodesic curve of healthy manifold spatially apart from degree
Health status is measured, the unified beginning and end to identical health status of each lifeline in space-time diagram, and then is predicted for the later stage
Basis is provided.Specific algorithm is described as follows:
(1) it (is determined by individual difference and operating mode by accumulation geodesic distance information, the decline rate information of each manifold spatially
Calmly), initial setting decline threshold value thre_ information etc., establishes preliminary affine scaling relationship rate0;
(2) setting optimization threshold value, such as:0.001;
(3) it in a certain range of preliminary affine scaling relationship rate0, by numerical method, calculates and optimizes affine
Proportionate relationship;
(4) in health status space-time diagram sample lifeline affine transformation.
Geodesic fitting and prediction are accumulated in 4.3 healthy manifold spaces
The healthy manifold obtained by TRIM methods spatially, calculate accumulation geodesic curve distance Curve, and then carry out accumulation
The fitting and prediction of geodesic curve distance Curve.And the process is be achieved one of method of geometry of truncated data forecasting problem
Necessary links.
Associate cumulation geodesic curve of the present invention apart from the characteristics of, using in MATLAB analysis tools (CFTOOL)
' Gaussian 6 ' realizes the fitting to accumulating geodesic curve distance Curve, and then obtains the service life of truncated data object.
The method of geometry of truncated data object performance degradation prognosis in 4.4 healthy manifold spaces
The present invention is with rolling bearing data, the performance degradation of truncated data properties object in analytic explanation health manifold space
Prediction technique and process.
4.4.1 experimental test device and data description (summary)
4.4.2 performance degradation feature and data truncation status analysis under completely cut off data qualification
As previously mentioned, performance degradation feature extraction is premise and the basis of all prediction works.In view of truncated data sheet
For body there is the incompleteness of data information, failing the screening of information for this categorical data will be more stringent, so as to can be
It is as much as possible under the conditions of limited data resource to reflect performance degradation characteristic.
Fig. 4 show 4 time domain indexes extracted from the full longevity vibration data of bearing, including:RMS (root-mean-square value), peak value
The factor, kurtosis, nargin.Fig. 6 show 8 energy indexes that the full longevity vibration data of bearing is extracted through wavelet package transforms.
By Fig. 4 and Fig. 5 it is found that RMS and 8 energy value of bearing preferably reflects its performance degradation state, and conduct
The high dimensional feature (9) that the present invention is extracted.Meanwhile by diagram it is found that bearing performance decline course in, starting stage property
The speed that can be failed is slow, and kept stable whithin a period of time.Enter rapid decline after reaching certain recession level
Process, and rapid failure.Therefore, it in order to solve the problems, such as the performance degradation prognosis under completely cut off data qualification, still needs to a certain amount of
Performance degradation status data, to seek the essential Evolution of sustainable prediction.
4.4.3 the completely cut off data manifold of higher dimension space structure based on TRIM
The content of present invention will extract the high dimensional feature (9 feature compositions) of bear vibration data in upper section, utilize the side TRIM
Method builds the healthy manifold space of truncated data point set, to obtain the health obtained through TRIM methods under the conditions of different truncated datas
Manifold neighborhood geodesic curve distance spatially and accumulation geodesic curve distance.Based on this, completes TRIM methods and be applied to truncated data
Manifold space builds compliance test result.
Test analysis is carried out with the original full longevity data instance being made of 520 data points.Neighborhood k=8, spatial context dimension
D=3 is spent, using TRIM be successively 450,510,515 to the state of blocking and the complete healthy manifold of longevity data progress is built, and is calculated each
The neighborhood geodesic curve distance and accumulation geodesic curve distance of truncated data state, as shown in figures 6 a and 6b.
By taking truncated data state is 450 and 510 as an example, calculates healthy manifold of the two through TRIM structures and spatially accumulate survey
Affine scaling relationship between ground linear distance.It is computed, affine transformation coefficient k _ affine=0.8399.As Fig. 7 is shown as ' 510 '
A truncation number strong point is directly obtained through TRIM methods are built and are calculated accumulation geodesic distance curve and after affine transformation
Affine accumulation geodesic curve distance Curve (with ' 450 ' truncated datas establish accumulation geodesic curve distance Curve establish affine pass
System).
By Fig. 7 and its data analysis it is found that TRIM, which effectively maintains original difference, blocks status data and full longevity data
Geometric properties in high-dimensional feature space, this provides necessary precondition for the development of subsequent prediction work.
4.4.4 the completely cut off data object performance degradation prognosis based on TRIM
Using the 20 groups of truncated datas and original full longevity data randomly selected in table 2, respective health is built using TRIM
Manifold space, and neighborhood geodesic curve distance and accumulation geodesic curve distance are extracted in healthy manifold space, it is based on ' Gaussian
6 ' complete the fitting and prediction of accumulation geodesic curve distances, and then realize ' truncated data object performance degradation in healthy manifold space
The method of geometry of prediction ' test and validation.With original full longevity data instance to the completely cut off data object based on TRIM
It can decline prediction progress analytic explanation (original full longevity data are through artificial truncation).
The manifold space of truncated data is built using TRIM, and calculates accumulation geodesic curve distance, using Gaussian6 to cutting
Disconnected data accumulation geodesic curve distance Curve is fitted and predicts, it is as follows to obtain correlated results:
Fit found when optimization terminated:
General model Gauss6:
F (x)=
a1*exp(-((x-b1)/c1)^2)+a2*exp(-((x-b2)/c2)^2)+
a3*exp(-((x-b3)/c3)^2)+a4*exp(-((x-b4)/c4)^2)+
a5*exp(-((x-b5)/c5)^2)+a6*exp(-((x-b6)/c6)^2)
Coefficients (with 95%confidence bounds):
A1=-3.721e+005 (- 1.455e+007,1.381e+007)
B1=566.5 (186.5,946.5)
C1=30.49 (- 81.69,142.7)
A2=1.666e+009 (- 3.795e+011,3.828e+011)
B2=723.6 (- 3941,5388)
C2=62.36 (- 651.2,775.9)
A3=-347.6 (- 706.7,11.47)
B3=287 (259.8,314.1)
C3=94.16 (57.29,131)
A4=-266.5 (- 512.1, -20.89)
B4=427.1 (412.6,441.7)
C4=32.97 (14.97,50.96)
A5=30.92 (- 34.55,96.39)
B5=404.2 (393.1,415.4)
C5=9.422 (- 12.09,30.94)
A6=5297 (4348,6246)
B6=505.6 (432.1,579.1)
C6=247.2 (212.9,281.5)
Goodness of fit:
SSE:1.424e+006
R-square:0.9993
Adjusted R-square:0.9993
RMSE:53.25
By Fig. 8 and fitting precision analysis it is found that 6 functions of Gaussian can be fitted truncated data health manifold well
Accumulation geodesic distance curve spatially.Further, to number in table 2 be the 1st~30 in 20 groups of differences randomly selecting cut
Disconnected status data carries out identical operating process, and is respectively blocked the life prediction value under state.
By above-mentioned curve matching precision and truncated data manifold of higher dimension space neutral energy metric parameter (accumulation geodesic curve away from
From) prediction it is found that 6 curves of Gaussian can effectively realize curve matching, and in certain fiducial interval range, can be obtained
Each corresponding life value of truncated data.And then on the basis of TRIM builds manifold of higher dimension space, solves truncated data performance
Fail forecasting problem.
On the basis of the method for geometry that subsequent content of the present invention will be predicted in the truncated data proposed, in conjunction with intelligent product
Estimator method (data with existing can be efficiently used) is limited, the probability expression (survival probability) of performance degradation prognosis is provided, and is predicted
The service life of truncated data object.
TABLE II
K step predictions under different phase status condition
Although describing the invention in detail above, but the invention is not restricted to this, those skilled in the art of the present technique
It can be carry out various modifications with principle according to the present invention.Therefore, all to be changed according to made by the principle of the invention, all it should be understood as
Fall into protection scope of the present invention.
Claims (6)
1. a kind of method of geometry of the rotating machinery health status prediction based on single branch truncated data, including:
By being monitored in real time to rotating machinery vibrating, numerous truncation number strong points are obtained;
The dimensionality reduction based on time continuity is carried out by numerous truncation number strong points high dimensional feature to numerous truncation number strong points
Processing obtains the truncated data health manifold space with multiple truncated data point features based on Annual distribution;
According to the time relationship between multiple truncated data point features in the truncated data health manifold space, to the multiple
Truncated data point feature carries out link process, forms the health status evolutional path of time-based truncated data;
It is fitted processing by the health status evolutional path to the truncated data, obtains cutting for the rotating machinery vibrating
The life value of disconnected data;
Wherein, the time in the healthy manifold space according to the truncated data between multiple truncated data point features closes
System, carrying out link process to the multiple truncated data point includes:
Neighborhood geodesic curve is carried out respectively to all two adjacent truncated data point features in the healthy manifold space of truncated data
Distance NGDs processing obtains all two adjacent truncated data point features based on time relationship in healthy manifold space
NGDs;
The NGDs of two all adjacent truncated data point features is subjected to accumulated process, forms accumulation in healthy manifold space
Geodesic curve distance CGDs, to obtain time-based truncated data truncated data health status evolutional path.
2. according to the method described in claim 1, wherein, by carrying out feature extraction processing to truncated data point, being blocked
Data point high dimensional feature.
3. according to the method described in claim 2, wherein, the high dimensional feature by numerous truncation number strong points into
Dimension-reduction treatment of the row based on time continuity, obtains the truncated data with multiple truncated data point features based on Annual distribution
Healthy manifold space includes:
Numerous truncated data point features are calculated using TRIM algorithms, by numerous truncation number strong points high dimensional feature
It is reduced to multiple truncated data point features;
According to time relationship, Annual distribution processing is carried out to the multiple truncated data point feature, forms the truncated data
Healthy manifold space;
Wherein, the TRIM algorithms are to continue related Isometric Maps algorithm the time.
4. according to the method described in claim 1, wherein, the healthy shape of the truncated data by the truncated data
State evolutional path is fitted processing, and the life value for obtaining the truncated data of the rotating machinery vibrating includes:
The extension based on future time is carried out to the CGDs to handle, obtain for prediction and carry future using Gaussian function
The extension CGDs values of time.
5. according to the method described in claim 4, wherein, the health status evolutional path by the truncated data
It is fitted processing, the life value for obtaining the truncated data of the rotating machinery vibrating further includes:
The extension CGDs values are compared with preset CGDs threshold values;
According to comparison result, the life value of the truncated data of the rotating machinery vibrating is determined.
6. it is described according to comparison result according to the method described in claim 5, wherein, determine the rotating machinery vibrating
Truncated data life value includes:
If the extension CGDs values are more than the CGDs threshold values, future time of the CGDs values equal to the CGDs threshold values will be extended
Determine the life value of the truncated data of the rotating machinery vibrating.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611132579.2A CN106599445B (en) | 2016-12-09 | 2016-12-09 | The method of geometry of rotating machinery health status prediction based on single branch truncated data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611132579.2A CN106599445B (en) | 2016-12-09 | 2016-12-09 | The method of geometry of rotating machinery health status prediction based on single branch truncated data |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106599445A CN106599445A (en) | 2017-04-26 |
CN106599445B true CN106599445B (en) | 2018-08-21 |
Family
ID=58598572
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611132579.2A Active CN106599445B (en) | 2016-12-09 | 2016-12-09 | The method of geometry of rotating machinery health status prediction based on single branch truncated data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106599445B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110175425B (en) * | 2019-05-31 | 2023-02-21 | 重庆大学 | Prediction method of residual life of gear based on MMALSTM |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102944435A (en) * | 2012-10-25 | 2013-02-27 | 北京航空航天大学 | Health assessment and fault diagnosis method for rotating machinery based on fisher discriminant analysis and mahalanobis distance |
CN103324834A (en) * | 2013-04-07 | 2013-09-25 | 北京航空航天大学 | Service life prediction method for electromechanical system and critical components under completely truncated data condition |
CN105303028A (en) * | 2015-08-20 | 2016-02-03 | 扬州大学 | Intelligent medical diagnosis classification method based on supervised isometric mapping |
-
2016
- 2016-12-09 CN CN201611132579.2A patent/CN106599445B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102944435A (en) * | 2012-10-25 | 2013-02-27 | 北京航空航天大学 | Health assessment and fault diagnosis method for rotating machinery based on fisher discriminant analysis and mahalanobis distance |
CN103324834A (en) * | 2013-04-07 | 2013-09-25 | 北京航空航天大学 | Service life prediction method for electromechanical system and critical components under completely truncated data condition |
CN105303028A (en) * | 2015-08-20 | 2016-02-03 | 扬州大学 | Intelligent medical diagnosis classification method based on supervised isometric mapping |
Non-Patent Citations (3)
Title |
---|
"机电系统故障预测技术的现状与分析";陶来发 等;《控制工程》;20110731;第18卷(第4期);第636-639页 * |
"船舶主推进系统故障预测与健康管理设计";马剑 等;《南京航空航天大学学报》;20110731;第43卷(第S期);第119-124页 * |
《Bearing diagnostics:A method based on differential geometry》;Ye Tian等;《Mechanical Systems and Signal Processing》;20160517;第377-391 * |
Also Published As
Publication number | Publication date |
---|---|
CN106599445A (en) | 2017-04-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Sameen et al. | Application of convolutional neural networks featuring Bayesian optimization for landslide susceptibility assessment | |
Sedighizadeh et al. | Particle swarm optimization methods, taxonomy and applications | |
CN105787501B (en) | Power transmission line corridor region automatically selects the vegetation classification method of feature | |
Eseye et al. | Adaptive predictor subset selection strategy for enhanced forecasting of distributed PV power generation | |
Rahman et al. | Prediction of land cover change based on CA-ANN model to assess its local impacts on Bagerhat, southwestern coastal Bangladesh | |
Yu | Disaster prediction model based on support vector machine for regression and improved differential evolution | |
CN106599445B (en) | The method of geometry of rotating machinery health status prediction based on single branch truncated data | |
Maduako et al. | Deep learning for component fault detection in electricity transmission lines | |
CN116954233A (en) | Automatic matching method for inspection task and route | |
CN115423810B (en) | Blade icing form analysis method for wind generating set | |
Ghalati et al. | LCM logistic regression modelling of land-use changes in Kouhmare Sorkhi, Fars province. | |
CN106548259B (en) | The method of geometry that rotating machinery health status is predicted under completely cut off data qualification | |
CN115980885A (en) | Rainfall forecast deviation correction method based on ensemble forecast | |
Smith et al. | EarthPT: a foundation model for Earth Observation | |
CN115759395A (en) | Training of photovoltaic detection model, detection method of photovoltaic power generation and related device | |
CN115359054A (en) | Power equipment defect detection method based on pseudo defect space generation | |
Zhan | Research on path planning method of humanoid robot based on improved genetic algorithm | |
Okwuashi et al. | Predicting future land use change using support vector machine based GIS cellular automata: A case of Lagos, Nigeria | |
Aqbelaghi et al. | Environmental approach in modelling of urban growth: Tehran City, Iran | |
Kalantar et al. | A comparison between three conditioning factors dataset for landslide prediction in the sajadrood catchment of Iran | |
Valero et al. | Analysis of different testing parameters in self-organizing maps for short-term load demand forecasting in Spain | |
de Mattos Neto et al. | An intelligent perturbative approach for the time series forecasting problem | |
Hui et al. | Power equipment segmentation of 3D point clouds based on geodesic distance with K-means clustering | |
Wang et al. | Data-Driven Outage Restoration Time Prediction via Transfer Learning with Cluster Ensembles | |
Zhang et al. | Genetic algorithm optimized neutral network classification and overlapping pixel change detection based on remote sensing for urban sprawl: a case study in Jiading district of Shanghai, China |
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 |