Research Article
Automatic segmentation and prognostic method of a turbofan engine using manifold learning and spectral clustering algorithms
Advances in Mechanical Engineering 2017, Vol. 9(9) 1–11 Ó The Author(s) 2017 DOI: 10.1177/1687814017722712 journals.sagepub.com/home/ade
Guangqi Qiu, Si Huang and Yu Chen
Abstract In most of the previous fault diagnostic literatures, the fault modes and states are pre-determined (i.e. the model structure (topology) is a priori known). However, in practical situation, the monitoring data, especially for the entire life-cycle data, nothing is known about the nature and the origin of the degradation (i.e. the model structure is unknown). Moreover, there is no consensus, how to determine the optimal model structure. In this condition, the different model structures may lead to different fault diagnosis/prognosis results. To address the optimal structure–selection problem, this article presents an automatic segmentation method based on Laplacian eigenmaps manifold learning and adaptive spectral clustering algorithms. Given an entire lifetime data of turbofan engine, we attempt to automatically segment the data into a sequence of contiguous regions corresponding to the degradation states. Furthermore, intrinsic dimensionality estimation, nonlinear dimension reduction, and the optimal number of degradation state estimation have been implemented. Automatic segmentation is applied for degradation state segmentation of non-label life-cycle data, and the output can be considered as the available information for developing fault diagnosis/prognosis. The experimental verification results indicate that the proposed automatic segmentation method is highly efficient and feasible for automatically determining the optimal model structure. Keywords Turbofan engine, automatic segmentation, fault diagnosis, fault prognosis, Laplacian eigenmaps, adaptive spectral clustering
Date received: 8 November 2016; accepted: 2 July 2017 Academic Editor: Yangmin Li
Introduction The turbofan engine providing thrust to the aircraft is one of the most vital systems to aviation system, and large amounts of researches have been carried out in the area of turbofan engine fault diagnosis/prognosis.1–5 The performance of an aircraft’s engine deteriorates when it is operated because its components physically degrade. To comprehensively describe the failure evolution, the degradation states before failure should be paid more attention. So far, many available fault diagnostic techniques, such as kernel principal component analysis (PCA),6
Kalman filters,7–9 sliding mode observer,10 support vector data description,11 Bayesian network,12 artificial neural networks,13 and hidden Markov model14, have been successfully applied in many industries. In most of these approaches,15,16 it is assumed that the model
School of Mechanical and Automotive Engineering, South China University of Technology, Guangzhou, China Corresponding author: Si Huang, School of Mechanical and Automotive Engineering, South China University of Technology, Guangzhou 510640, China. Email:
[email protected]
Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/ open-access-at-sage).
2
Advances in Mechanical Engineering
Figure 1. Basic segmentation problem of lifetime data.
structure is pre-determined, ignoring the basic segmentation problem: given an entire lifetime data, we wish to partition our data into contiguous regions corresponding to the degradation states (see Figure 1). The segmentation problem is taken into account in a few of these approaches with unsupervised clustering algorithms,17,18 but the degradation state number is predetermined as well. There are three possible reasons that explain why the degradation states division is rarely considered in the previous literatures: (1) the degradation states division process itself is time-consuming, (2) finding a reasonable degradation state structure which requires large amount of data, is usually not available in real world, and (3) the curses of dimensionality in high-dimensional data are still the major issue that challenges the degradation states division. Actually, the degradation states division is an important step for fault diagnosis and prognosis. If the number of degradation state is too small, the diagnosis/ prognosis results are affected. On the contrary, the CPU time is expensive. As a matter of fact, the life-cycle data of turbofan engine are the nonlinear and non-stationary time series, which contain abundant potential failure information. However, with the exponential increase of performance monitoring data, the curses of dimensionality are also the main issue that affects the accuracy of fault diagnosis/prognosis. Fortunately, manifold learning algorithm is a perfect tool for data mining that discovers the structure of high-dimensional data and provides better understanding of data.19 Recently, a variety of nonlinear dimensionality reduction techniques have been proposed which aim to address the limitations of traditional techniques such as PCA.20 The core idea of manifold learning algorithm is to find a nonlinear low-dimensional embedding of high-dimensional data without losing much information. Additionally, the dimension of the embedding is a key parameter for manifold projection methods: if the dimension is too small, important data features are ‘‘collapsed’’ in the same dimension. On the contrary, the projections become noisy and unstable.21 There is no consensus, however, on how this dimension should be determined. Automatic segmentation methods have been employed in image processing and speech
recognition22–24 but few about fault diagnosis and prognosis. Forecasting the future states of a complex system is a complicated challenge that is encountered in many industrial applications covered in the community of prognostics and health management. Practically, states can be either continuous or discrete: continuous states generally represent the value of a signal, while discrete states generally depict functioning modes reflecting the current degradation.25–27 In this article, we introduce an automatic segmentation method which aims to address the optimal structure–selection problem and the uncertainty of degradation process. First, we attempt to automatically segment the life-cycle data into a sequence of contiguous regions corresponding to the degradation states. Furthermore, intrinsic dimensionality estimation, nonlinear dimension reduction, and the optimal number of degradation state estimation have been implemented in the proposed method. The procedure of this method is presented in the next section. Section ‘‘Automatic segmentation algorithm’’ discusses the corresponding automatic segmentation algorithms in detail, including Laplacian eigenmaps (LE) manifold learning and adaptive spectral clustering methods. The flowchart of fault diagnosis and prognosis method is given in section ‘‘Fault diagnosis and prognosis.’’ In section ‘‘Experimental validation,’’ the run to failure experiment of turbofan engines is performed to verify the proposed method in this article. Finally, conclusions are presented in section ‘‘Conclusion.’’
The procedure of automatic segmentation method There are four steps for machine fault diagnosis and prognosis: (1) data acquisition, (2) data process, (3) fault diagnosis, and (4) fault prognosis. To avoid the curses of dimensionality and improve the performance of classification, the high-dimensional data can be efficiently summarized in a space of a much lower dimension without losing much information. There are many approaches for dimensionality reduction, such as Sammon mapping,28 kernel PCA,29 isomap,30 locally linear embedding,31 LE,32 maximum variance unfolding,33 and t-distributed stochastic
Qiu et al.
3
Figure 2. The procedure of automatic segmentation and fault diagnosis.
neighbor embedding.34 Van Der Maaten et al.35 presented a systematic comparison of 13 existing common dimensionality reduction techniques by four general properties: (1) the parametric nature of the mapping between the high-dimensional and the low-dimensional space, (2) the main free parameters that have to be optimized, (3) the computational complexity of the main computational part of the technique, and (4) the memory complexity of the technique. In practice, one of the main limitations is the computational complexity of these approaches, and this limits to apply in large-scale data.35–37 In terms of the computational and memory complexity, LE is one of the most outstanding techniques. Additionally, constructing a reliable estimator of intrinsic dimension and understanding its statistical properties will clearly facilitate further applications of manifold projection methods and improve their performance. So far many researchers have made great contributions on intrinsic dimensionality estimation, and maximum likelihood estimation (MLE),21 correlation dimension, nearest neighbor evaluation, packing numbers, and geodesic minimum spanning tree are main available techniques.38,39
In summary, the main problems for affecting the performance of fault diagnosis/prognosis are intrinsic dimensionality estimation, curses of dimensionality, and degradation process uncertainty. In this section, an automatic segmentation method is presented, which can be used to automatically find the low-dimensional embedding and determine the degradation state number and state label. The corresponding automatic segmentation algorithms, fault diagnostic and prognostic approaches are described in sections ‘‘Automatic segmentation algorithm’’ and ‘‘Fault diagnosis and prognosis,’’ respectively. Figure 2 shows the procedure of the automatic segmentation and fault diagnosis. The life-cycle data are collected by the sensors installed on the equipment, and feature extraction or selection method is employed to select the system degradation indicators. The measured data may be redundant; effective feature extraction and selection is a step for accurate diagnosis and prognosis. To eliminate the degradation process uncertainty and determine the degradation states, an automatic segmentation method is employed, including two phases: (1) phase 1: dimensionality reduction and (2) phase 2:
4 adaptive spectral clustering. In phase 1, LE manifold learning technique is employed to find a lowdimensional embedding via high-dimensional data. In other words, the effectiveness of fault diagnostic and prognostic are increased by this stage. In this article, the fusion data are as the input of degradation state segmentation and fault diagnostic. For fault prognostic, several degradation indicators are often converted into a single degradation indicator in previous papers, so that we choose the first dimensional fusion data as the prognostic parameter. In phase 2, an adaptive spectral clustering algorithm is presented to determine the degradation state number and output the degradation state label. With such a strategy, the life-cycle data are automatically divided into contiguous regions corresponding to the degradation states. Finally, the output of the automatic segmentation method can be considered as the available information for developing fault diagnostic.
Advances in Mechanical Engineering
Phase 1: manifold learning dimensionality reduction For a fixed point x, assume f (x) = const in a small sphere Sx (R) of radius R around x, and the observation is considered as a non-stationary random process fN (t, x), 0 t Rg N (t, x) =
fXi 2 Sx (t)g
ð1Þ
i=1
where N (t, x) is the number of points in the small sphere Sx (R). Approximating this non-stationary random process (fixed n) by a Poisson process, then k ’ f (x)V (d)(Tk (x))d n
ð2Þ
where Tk (x) is the k-nearest neighbor Euclidean distance of x. Let t being a fixed value, the rate l(t) of the process N (t, x) can be written as l(t) ’ f (x)V (d)dtd1
Automatic segmentation algorithm (P) (P) Let YP = fx(P) 1 , x2 , . . . , xn g be the entire life-cycle data of the equipment from the multi-sensors time series. Where xi and P represent the data point at time i and the dimension of multi-sensors data, respectively. Our objective is to find a set of points y1 , y2 , . . . , yn in (d) (d) Yd = fy(d) 1 , y2 , . . . yn g (d P) which represents the (P) (P) observations x1 , x2 , . . . , xn in YP = fx(P) 1 , x2 , . . . , xn g, and to partition the low-dimensional embedding Yd into r contiguous regions corresponding to the degradation states. The automatic segmentation algorithm is summarized as follows:
n X
ð3Þ
where V (d) = pd=2 ½G((d=2 + 1)1 is the volume of the unit sphere in Yd . Let u = log ( f (x)), the log-likelihood function is obtained as ðR L(d, u) =
ðR
log l(t)dN(t) l(t)dt
0
which yields ÐR l(t)dt). 0
ð4Þ
0
d^R (x) = arg max ( dR (x)
ÐR 0
log l(t)dN(t)
Algorithm: Automatic segmentation Input: Life-cycle data YP, neighborhood range k = k1, ..., k2, the neighbor number k3 and the parameter t. Phase 1: Manifold learning dimensionality reduction 1. Compute matrix of log nearest neighbor distances log (Tk (xi)). 2. Constructing the maximum likelihood function by equation (4) 3. Compute the maximum likelihood estimate dk by equation (6). 4. Compute the intrinsic dimensionality number d by equation (7). 5. Constructing the similarity graphs by the k-nearest neighbor graph, where the neighbor number is k2. 6. Compute the affinity matrix is vij = exp*(2Aij2/t) if i6¼j and vii = 0 by equation (8). 7. Compute eigenvalues and eigenvectors by equation (9). 8. Find the low-dimension embedding Yd = {y1, y2, ..., yd}. Phase 2: Adaptive spectral clustering 1. Constructing the similarity graphs by the mutual k-nearest neighbor graph, where the neighbor number is k3, and the input data is the low-dimension embedding Yd. 2. Compute the affinity matrix vij = exp*(2Aij2/t) if i6¼j, and vii = 0. 3. Compute the normalized laplacian L. 4. Compute the eigenvalues li and arranging them with the descending order l1l2.ln. 5. Compute the eigengap gi = li2li + 1, and determine the cluster number by equation (10). 6. Form the matrix U2Yr, including the first r eigenvectors u1, u2, ..., ur. 7. The normalization matrix U# can be obtained by equation (11). 8. Partition n samples of U# into r contiguous regions corresponding to the degradation states by k-means clustering algorithm. Output: Intrinsic dimensionality d, degradation state number r, the degradation state with label C1, C2, ..., Cr.
Qiu et al.
5
The MLE must satisfy ∂L=∂u = 0 and ∂L=∂d = 0, thus the MLE of d is
dR (x) =
NX (R, x) 1 R log N (R, x) j = 1 Tj (x)
!1 ð5Þ
In practice, it may be more convenient to fix the number of neighbors k rather than the radius of the sphere R. Then equation (5) becomes k1 1 X Tk (x) log dk (x) = k 1 j=1 Tj (x)
!1 ð6Þ
The intrinsic dimensionality d is obtained by setting the range of the neighborhood k = k1 , . . . , k2 dk =
n 1X dk (x), n i=1
d=
k2 X 1 dk k2 k 1 + 1 k = k
ð7Þ
1
Let G = (V , E) be an undirected graph with vertex set V = fv1 , v2 , . . . , vn g. Where each vertex vi in this graph represents a data point xi and the number of V is n. The respective edge weights E for vertices (vi , vj ) are determined by the non-negative weight wij 0, where wij = 0 means that there is no connection between vertices (vi , vj ). Thus, the distance matrix of vertices (vi , vj ) is solved by A2ij = jjxi xj jj2 , and the corresponding affinity matrix wij can be defined as wij =
exp ( A2ij =t) xj 2 Tk (xi ) 0 otherwise
ð8Þ
where the parameter t is the thermonuclear width. In this article, the k-nearest neighbor graph was chosen to construct the similarity graphs for dimensionality reduction, and the neighborhood number was k2 . The mutual k-nearest neighbor graph was chosen to spectral clustering, and the neighborhood number was k3 . Assume the constructed graph G is connected. Compute eigenvalues and eigenvectors for the generalized eigenvector problem Ly = lDy
Phase 2: adaptive spectral clustering Adaptive spectral clustering and LE manifold learning are both based on LE algorithm, the difference between them is calculated by the generalized eigenvector problem. LE manifold learning finds a lower dimensional embedding by a pre-defined value, but adaptive spectral clustering determines the degradation state number by the first maximal eigengap. For all clustering algorithms, how to automatically determine the number of clusters is a general problem. Kong et al.40 designed the first maximal eigengap for determining the number of clusters. According to the spectral graph theory, in the ideal case of r completely disconnected clusters, the normalized Laplacian matrix contains the eigenvalue as 1 with multiplicity r and then, there is a strict gap to the (r + 1)th eigenvalue that ld + 1 1. Thus, the eigengap gi = li li + 1 can be designed to automatically determine the number of clusters. The practical cases can be considered as the perturbation form of the ideal cases. In this condition, the Laplacian matrix L is not diagonal block. There are r number of eigenvalues are large, such that l1 , l2 , . . . , lr , but the value of lr + 1 is relatively small. The eigengap gr = lr lr + 1 is relatively large. Therefore, the clustering number is determined by the first maximal eigengap sequence n o r = arg min gi gj jj\i .0 & pgi gi + 1 .0
ð10Þ
i
Note that the input data of adaptive spectral clustering are the low-dimensional embedding Yd . Let U 2 Yn 3 r be the matrix including the first r eigenvectors u1 , u2 , . . . , ur . Normalized the matrix U , the normalization matrix U 0 is obtained by u0ij = P
uij 2 r uir
1=2 ,
i = 1, 2, . . . , n, j = 1, 2, . . . , r ð11Þ
Therefore, the k-means clustering algorithm is employed to cluster n samples of U 0 into r clusters C1 , C2 , . . . , Cr .
ð9Þ
where D is a diagonal matrix, and its entries are column (or row, P since W is symmetric) sums of W , Dii = nj wji . L = D W is the Laplacian matrix. Laplacian is a symmetric, positive semi-definite matrix which can be thought of as an operator on functions defined on vertices of G. Let y0 , y1 , . . . , yn1 be the solutions of equation (9), order according to their eigenvalues with y0 having the smallest eigenvalue (in fact 0). The life-cycle data (P) (P) YP = fx(P) 1 , x2 , . . . , xn g under the embedding into the is given by lower dimensional space Yd fy1 (i), . . . , yd (i)g.
Fault diagnosis and prognosis Fault diagnostics and prognostics have received increased attention due their potential to provide early warning of system failures, forecast maintenance as needed, and reduce life-cycle costs. Support vector machine (SVM)41,42 has been proven to have an excellent generalization capability and has been successfully applied in machinery fault diagnosis. In this article, SVM was chosen to develop fault diagnosis, and Cox proportional hazards model (PHM) was to develop fault prognosis.
6
Advances in Mechanical Engineering
Figure 3. The flowchart of fault diagnosis and prognosis.
Figure 3 shows the flowchart of fault diagnosis and prognosis method. The steps of fault diagnosis and prognosis are summarized as follows: Step 1. Data acquisition: the monitoring data are collected by sensors installed on the equipment. The historical data are as the training data, and the realtime data are as the testing data. Step 2. Feature extraction and selection: the monitoring data may be redundant, so that the appropriate degradation indicators are selected in this stage. Step 3. Data fusion: to improve the effectiveness of fault diagnosis and prognosis, the degradation indicators are processed by LE algorithms. The fusion data are as the input of degradation states segmentation and fault diagnosis, and the first dimensional fusion data as the prognostic parameter. Step 4. Degradation states segmentation: adaptive spectral clustering is employed to segment the unlabeled monitoring data. Step 5. Fault diagnosis: SVM is employed to develop fault diagnosis; the corresponding algorithm is described in section ‘‘Fault diagnosis based on SVM.’’ Step 6. Fault prognosis: Cox PHM is employed to develop fault prognosis; the corresponding algorithm is described in section ‘‘Fault prognosis based on Cox PHM.’’
Fault diagnosis based on SVM For linearly separable part, there is an optimal hyperplane wy + b = 0 that minimizes jjwjj2 . Consider a problem of binary classification where training data are given as (yi , zi ), y 2 Rd , zi 2 f + 1, 1g, i = 1, 2, . . . , n. The problem of finding a separating hyperplane can be
transformed into a quadratic programming problem by Lagrange optimization method 8 n n P P > > ai 12 ai aj zi zj (yi , yj ) > max w(a) = > < i=1 i, j = 1 n P > zi a i s:t: > > > i=1 : ai 0, i = 1, 2, . . . , n
ð12Þ
where n, d, zi , ai denote sample number, sample dimension, sample classification, Lagrange’s multiplier, respectively. + 1 and –1 are the classification labels. The optimal classification function is (
f (y) = signf(w y) + b g = sign
n X
) ai zi (yi y) + b
i=1
ð13Þ where ai and b denote the optimal Lagrange coefficient and classification threshold, respectively. The positive and negative of the classification function are the classification labels. For linear non-separable part, the main idea is to map the original d-dimensional space into a d#-dimensional space (d 0 .d), where the points can possibly be linearly separated. The only operation required in the transformed space is the inner product f(yi )T f(yj ), which is defined with the kernel function (K) between yi and yj . The common kernel functions are linear kernel function (Linear), polynomial kernel function (Polynomial), the radial basis function (RBF), and a hyperbolic tangent sigmoid kernel function. The objective function is w(a) =
n X i=1
ai
n 1 X ai aj zi zj K(yi , yj ) 2 i, j = 1
ð14Þ
Qiu et al.
7
The optimal classification function is ( f (y) = sign
n X
) ai zi K(yi y) + b
ð15Þ
i=1
Fault prognosis based on Cox PHM Let Zi = fZi1 , Zi2 , . . . , Zip g be the realized values of the covariates for subject i. The hazard function for the Cox PHM has the form h(t, Zi ) = h0 (t) exp (b1 Zi1 + b2 Zi2 + + bp Zip ) = h0 (t) exp (bZi )
ð16Þ
where h0 (t), b, and Zi are the baseline hazard function, regression coefficient, and covariate vector, respectively. The cumulative hazard rate is ðt H(t, Zi ) =
ð17Þ
h(t, Zi )dt 0
ð18Þ
R(t, Zi ) = exp ( H(t, Zi ))
Treating the subjects’ events as if they are statistically independent, the joint probability density of loglikelihood function is
ln L(Zi ; b) =
i, Ci = 1
0 @Zi b ln
jRULactual RULestimated j MAPE = mean RULactual
ð22Þ
where RULactual is the actual remaining service life.
Experimental validation
The reliability function is
Y
Figure 4. Actual lifetime of each engine in the training set.
X
1 exp (Zi b)A
ð19Þ
j:Tj .Ti
where Ti denotes the observed time (either censoring time or event time) for subject i. Ci is the indicator that the time corresponds to an event (i.e. if Ci = 1, the event occurred and if Ci = 0, the time is a censoring time). The estimate of regression coefficient b can be obtained by maximum likelihood method. The overall service lifetime can be obtained by R(t, Zi ) = exp ( H(t, Zi )).RULcon
ð20Þ
Degradation state segmentation
where RULcon is the reliability threshold. The remaining service life can be obtained by RULestimated = TOF Tcurrent
To validate the effectiveness of the proposed method, the run-to-failure tests on turbofan engines were performed. The experimental data were downloaded from the Prognostics Data Repository (http://ti.arc.nasa.gov/ project/prognostic-data-repository).43 The data sets consisted of multiple multivariate time series, reflecting the natural degradation of turbofan engines. The dataset 1, namely, train_FD001, test_FD001, and RUL_FD001, was considered in this article. In the training set, the fault grows in magnitude until system failure. In the test set, the time series ends some time prior to system failure. RUL_FD001 provides a vector of true remaining useful life (RUL) values for the test data. The dataset contained 24 element vectors consisted of 3 operational settings and 21 sensor measurements.44 The total lifetime for each engine in the training set is shown in Figure 4. It can be seen from Figure 4 that the lifetime varies from 128 to 362 cycles (mean = 206.3 cycles, standard deviation = 46.34 cycles).
ð21Þ
where TOF and Tcurrent are the overall service lifetime and current time, respectively. To evaluate the prognostic performance, the mean absolute percent error (MAPE) is given as
To improve the effectiveness of fault diagnosis and prognosis and reflect the natural degradation of turbofan engines, the respective features are selected. For fault diagnosis, the 24 features are as the degradation indicators. By automatic segmentation in section ‘‘Automatic segmentation algorithm,’’ the dimension of the life-cycle data (24-dimension) was reduced into 6-dimension. Where, the neighborhood number in
8
Advances in Mechanical Engineering worst performance of the system. The performance of the system is gradual degradation in this order. The output of the automatic segmentation method can be considered as the available information for developing fault diagnosis.
Fault diagnosis
Figure 5. Visualization of the reduction results by LE manifold learning.
Phase 1 was in the range of 6–12, and the neighborhood number k2 and parameter t were 12 and 1, respectively. The first three-dimension of the reduction results is selected for visualization, which is shown in Figure 5. Note that here the output of LE manifold learning is four-dimensional data. By Phase 2 in section ‘‘Automatic segmentation algorithm’’ with the output of Phase 1, the state number and class label were obtained. Where, the neighborhood number k3 and parameter t in Phase 2 were 15 and 1, respectively. Figure 6(a) shows the estimation results of degradation states number, and the corresponding state is shown in Figure 6(b). From Figure 6, it can be seen that the turbofan engine will go through r = 4 degradation states from time zero to failure. The degradation state 1 ! r indicates the system performance is a gradual deterioration process. Where state 1 denotes the perfect performance of the system, state r denotes the
For fault diagnosis, the engines No. 1–No. 60 are taken as the train samples and the engines No. 81–No. 100 are taken as the test samples. Note that here the samples are random distribution, and the optimal parameters of SVM are found by grid search algorithm, and the cluster labels are considered as the true labels of turbofan engine. The diagnosis results by SVM are compared with the true labels, which are shown in Figure 7. The diagnosis accuracy is 99.55%. The respective corresponding degradation states of No. 94 engine and No. 82 engine over the operational time are shown in Figure 8. It can be seen that two engines start with different initial operational states owing to different degrees of initial wear and manufacturing variation which is unknown to the user. This wear or variation is considered normal, that is, it is not considered a fault condition. No. 94 engine goes through four degradation states, while No. 82 engine only three degradation states. The results indicate that No. 82 engine is operated at degradation state 2, and No. 94 engine has the higher reliability in the initial operational state than No. 82 engine. The reason is that products exist the difference in material and manufacturing process.
Fault prognosis For fault prognosis, train_FD001 are taken as the training samples, test_FD001 are taken as the testing
Figure 6. Results of automatic segmentation: (a) estimation of degradation states number and (b) results of adaptive spectral clustering.
Qiu et al.
Figure 7. SVM diagnosis results.
Figure 8. Degradation states of No. 94 engine and No. 82 engine.
samples. The literature45,46 suggested the features f7, 8, 12, 16, 17, 20g led to the best prognostic performance. In this article, we also chose the features f7, 8, 12, 16, 17, 20g as the degradation indicators for fault prognosis. By Cox PHM in section ‘‘Fault prognosis based on Cox PHM,’’ the Cox PHM RUL estimation result of RULcon = 0.9 and RULcon = 0.95 are shown in Figures 9 and 10, respectively. Where the estimated value of b = 37.6364 and the covariate vector are the first dimensional fusion data. It can be seen from Figures 9 and 10, the MAPE values of RULcon = 0.9 and RULcon = 0.95 are 50.2934 and 56.809, respectively. To better evaluate the prognostic performance, Cox PHM RUL estimation method was compared with Weibull estimation. Figure 11 shows the Weibull estimated results. Where the estimated values of a and h are 4.4087 and 225.0258, respectively, and the MAPE value of Weibull estimation is 98.5072. It is concluded that Cox PHM is more suitable to
9
Figure 9. Cox PHM estimated results of RULcon = 0.9: (a) Cox PHM RUL estimates and (b) Cox PHM RUL estimation error: MAPE = 50.2934.
Figure 10. Cox PHM estimated results of RULcon = 0.95: (a) Cox PHM RUL estimates and (b) Cox PHM RUL estimation error: MAPE = 50.809.
Figure 11. Weibull estimated results: (a) Weibull RUL estimates and (b) Weibull RUL estimation error: MAPE = 98.5072.
10
Advances in Mechanical Engineering
predict the remaining service life than Weibull estimation and has achieved a good performance. Where the Weibull failure rate is a t a1 l(t) = h h
ð23Þ
where a and h are shape parameter and scale parameter, respectively. The mean remaining service life of Weibull estimation is estimated by 1 RULestimated = S(t)
+‘ ð
data. Therefore, the proposed method can be seamlessly applied to any mechanical equipment, and the output of the automatic segmentation method can be considered as the available information for developing fault diagnosis/prognosis. Declaration of conflicting interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding S(t)dt
ð24Þ
t
where t is the current time, and S() is the survival function. From the above results, the automatic segmentation has been proven to be a high-efficiency and feasibility method, which can automatically find the lowdimensional embedding and determine the structure of the degradation model. Thus, the data uncertainty can be eliminated, which is of great significance for fault diagnosis and prognosis. Additionally, the main challenge in the automatic segmentation method is the available life-cycle data. Finding a reasonable model which requires large amount of data is usually not available in real world. The overall accuracy of fault diagnosis/prognosis will be seriously affected by the insufficient data. Moreover, selection of the monotonous features representing the degradation progression is a prerequisite for effective fault prognostics. However, the performance of the automatic segmentation method itself is sufficiently accurate.
Conclusion In this article, an automatic segmentation was performed, which was validated through three artificial data sets and the run-to-failure experiments of turbofan engine. The results suggest that the automatic segmentation method has effectively overcome the curse of dimensionality and eliminated the degradation process uncertainty. It is a high-efficiency and feasibility method for model selection of high-dimensional data. The proposed approach has three main benefits. First, only the performance monitoring data are required but not the prior knowledge about the equipment or the operating conditions. Second, the method can be employed in single or multiple operating conditions. Third, the method not only can deal with low-dimensional data (directly Phase 2) and highdimensional data but also the non-convex and convex
The author(s) received no financial support for the research, authorship, and/or publication of this article.
References 1. Litt JS. An optimal orthogonal decomposition method for Kalman filter-based turbofan engine thrust estimation. J Eng Gas Turb Power 2008; 130: 011601. 2. Tayarani-Bathaie SS, Vanini ZS and Khorasani K. Dynamic neural network-based fault diagnosis of gas turbine engines. Neurocomputing 2014; 125: 153–165. 3. Ramasso E and Saxena A. Review and analysis of algorithmic approaches developed for prognostics on CMAPSS dataset. In: Proceedings of the annual conference of the prognostics and health management society, Fort Worth, TX, USA, September 2014. 4. Donat W, Choi K, An W, et al. Data visualization, data reduction and classifier fusion for intelligent fault diagnosis in gas turbine engines. J Eng Gas Turb Power 2008; 130: 041602. 5. Ramasso E and Gouriveau R. Prognostics in switching systems: evidential Markovian classification of real-time neuro-fuzzy predictions. In: Proceedings of the prognostics and health management conference, 2010 (PHM’10), Macao, China, 12–14 January 2010, pp.1–10. New York: IEEE. 6. Feng D, Xiao M, Liu Y, et al. A kernel principal component analysis–based degradation model and remaining useful life estimation for the turbofan engine. Adv Mech Eng. Epub ahead of print 20 May 2016. DOI: 10.1177/ 1687814016650169. 7. Dewallef P and Borguet S. A methodology to improve the robustness of gas turbine engine performance monitoring against sensor faults. J Eng Gas Turb Power 2013; 135: 051601. 8. Dewallef P, Romessis C, Le´onard O, et al. Combining classification techniques with Kalman filters for aircraft engine diagnostics. J Eng Gas Turb Power 2006; 128: 281–287. 9. Chang X, Huang J, Lu F, et al. Gas-path health estimation for an aircraft engine based on a sliding mode observer. Energies 2016; 9: 598. 10. Lu F, Zheng W, Huang J, et al. Life cycle performance estimation and in-flight health monitoring for gas turbine engine. J Dyn Syst Meas Control 2016; 138: 091009.
Qiu et al. 11. Dreiseitl S and Ohno-Machado L. Logistic regression and artificial neural network classification models: a methodology review. J Biomed Inform 2002; 35: 352–359. 12. Romessis C and Mathioudakis K. Bayesian network approach for gas path fault diagnosis. J Eng Gas Turb Power 2006; 128: 64–72. 13. Ogaji SO and Singh R. Advanced engine diagnostics using artificial neural networks. Appl Soft Comput 2003; 3: 259–271. 14. Ramasso E. Contribution of belief functions to hidden Markov models with an application to fault diagnosis. In: Proceedings of the IEEE international workshop on machine learning for signal processing (MLSP 2009), Grenoble, 1–4 September 2009, pp.1–6. New York: IEEE. 15. Yu SZ. Hidden semi-Markov models. Artif Intell 2010; 174: 215–243. 16. Peng Y and Dong M. A prognosis method using agedependent hidden semi-Markov model for equipment health prediction. Mech Syst Signal Pr 2011; 25: 237–252. 17. Kim K, Ball C and Nwadiogbu E. Fault diagnosis in turbine engines using unsupervised neural networks technique. In: Proceedings of the SPIE defense and security, Orlando, FL, April 2004, pp.150–158. Bellingham, WA: International Society for Optics and Photonics. 18. Moghaddass R and Zuo MJ. A parameter estimation method for a condition-monitored device under multistate deterioration. Reliab Eng Syst Safe 2012; 106: 94–103. 19. Kouropteva O, Okun O and Pietika¨inen M. Incremental locally linear embedding. Pattern Recogn 2005; 38: 1764–1767. 20. Van der Maaten LJP, Postma EO and Van Den HJ. Dimensionality reduction: a comparative review. Technical report, Maastricht University, Maastricht, May 2007. 21. Levina E and Bickel PJ. Maximum likelihood estimation of intrinsic dimension. In: Proceedings of the advances in neural information processing systems, Vancouver, BC, Canada, 13–18 December 2004, pp.777–784. New York: Association for Computing Machinery. 22. He R, Qin B and Liu T. A novel approach to update summarization using evolutionary manifold-ranking and spectral clustering. Expert Syst Appl 2012; 39: 2375–2384. 23. Prastawa M, Gilmore JH, Lin W, et al. Automatic segmentation of MR images of the developing newborn brain. Med Image Anal 2005; 9: 457–466. 24. Creutz M. Induction of the morphology of natural language: unsupervised morpheme segmentation with application to automatic speech recognition. Helsinki University of Technology, 2006, http://lib.tkk.fi/Diss/2006/ isbn9512282119/isbn9512282119.pdf 25. Ramasso E and Denoeux T. Making use of partial knowledge about hidden states in HMMs: an approach based on belief functions. IEEE T Fuzzy Syst 2014; 22: 395–405. 26. Ramasso E and Gouriveau R. Remaining useful life estimation by classification of predictions based on a neurofuzzy system and theory of belief functions. IEEE T Reliab 2014; 63: 555–566. 27. Ramasso E, Rombaut M and Zerhouni N. Joint prediction of continuous and discrete states in time-series based on belief functions. IEEE T Cyb 2013; 43: 37–50.
11 28. De Ridder D and Duin RP. Sammon’s mapping using neural networks: a comparison. Pattern Recogn Lett 1997; 18: 1307–1316. 29. Cao LJ, Chua KS, Chong WK, et al. A comparison of PCA, KPCA and ICA for dimensionality reduction in support vector machine. Neurocomputing 2003; 55: 321–336. 30. Samko O, Marshall AD and Rosin PL. Selection of the optimal parameter value for the Isomap algorithm. Pattern Recogn Lett 2006; 27: 968–979. 31. Roweis ST and Saul LK. Nonlinear dimensionality reduction by locally linear embedding. Science 2000; 290: 2323–2326. 32. Belkin M and Niyogi P. Laplacian eigenmaps for dimensionality reduction and data representation. Neural Comput 2003; 15: 1373–1396. 33. Weinberger KQ and Saul LK. An introduction to nonlinear dimensionality reduction by maximum variance unfolding. In: Proceedings of the 21st national conference on artificial intelligence (AAAI’06), Boston, MA, 16–20 July 2006, vol. 6, pp.1683–1686. New York: Association for Computing Machinery. 34. Maaten LVD and Hinton G. Visualizing data using tSNE. J Mach Learn Res 2008; 9: 2579–2605. 35. Van Der Maaten L, Postma E and Van den Herik J. Dimensionality reduction: a comparative. J Mach Learn Res 2009; 10: 66–71. 36. Van der Maaten L. Learning a parametric embedding by preserving local structure. RBM 2009; 500: 26. 37. Van Der Maaten L. Accelerating t-SNE using tree-based algorithms. J Mach Learn Res 2014; 15: 3221–3245. 38. Camastra F. Data dimensionality estimation methods: a survey. Pattern Recogn 2003; 36: 2945–2954. 39. Carter KM, Raich R and Hero AO III. On local intrinsic dimension estimation and its applications. IEEE T Signal Proces 2010; 58: 650–663. 40. Kong W, Sun C, Hu S, et al. Automatic spectral clustering and its application. In: Proceedings of the 2010 international conference on intelligent computation technology and automation (ICICTA), Changsha, China, 11–12 May 2010, vol. 1, pp.841–845. New York: IEEE. 41. Hsu CW, Chang CC and Lin CJ. A practical guide to support vector classification, 2003, http://www.csie.ntu. edu.tw/;cjlin/papers/guide/guide.pdf 42. Chang CC and Lin CJ. LIBSVM: a library for support vector machines. ACM Trans Intell Syst Tech 2011; 2: 27. 43. Saxena A, Goebel K, Simon D, et al. Damage propagation modeling for aircraft engine run-to-failure simulation. In: Proceedings of the international conference on prognostics and health management (PHM 2008), Denver, CO, 6–9 October 2008, pp.1–9. New York: IEEE. 44. Liu K and Huang S. Integration of data fusion methodology and degradation modeling process to improve prognostics. IEEE T Autom Sci Eng 2016; 13: 344–354. 45. Soumik S, Jin X and Asok R. Data-driven fault detection in aircraft engines with noisy sensor measurements. J Eng Gas Turb Power 2011; 133: 783–789. 46. Ramasso E. Investigating computational geometry for failure prognostics. PHM Soc 2014; 5: 1–18.
