We propose an approach for failure prognosis based on the estimation of the Remaining Useful Life (RUL) of a system in a situation in which monitoring signals providing information about its degradation evolution are not measured and no operating model of the system is available. These conditions are of practical interest for industrial applications such as mechanical (e.g. rolling bearing) or electrical (e.g. wind turbine) devices or equipment-critical components (e.g. batteries) in which the addition of sensors to the system is not feasible (e.g. space limitations for sensors, cost, etc.). The approach is based on an estimation of the system degradation using residual generation (where the difference between the system and the observer outputs is processed) and Hidden Markov Models with discrete observations. The prediction of the system RUL is given by the Markov property concerning the mean time before absorption. The approach comprises two phases: a training phase to model the degradation behavior and an “on-line” use phase to estimate the remaining life of the system. Two case studies were conducted for RUL prediction to verify the effectiveness of the proposed approach.
SikorskaJZHodkiewiczMMaL.Prognostic modelling options for remaining useful life estimation by industry. Mech Syst Signal Process2011; 25(5): 1803–1836.
2.
DragomirOEGouriveauRDragomirF, et al. Review of prognostic problem in condition-based maintenance. In: European control conference, Budapest, Hungary, pp.1585–1592, 2009.
3.
ChehadeASongCLiuK, et al. A data-level fusion approach for degradation modeling and prognostic analysis under multiple failure modes. J Qual Technol2018; 50(2): 150–165.
4.
JardineAKSLinDBanjevicD.A review on machinery diagnostics and prognostics implementing condition-based maintenance. Mech Syst Signal Process2006; 20(7): 1483–1510.
5.
LeeJWuFZhaoW, et al. Prognostics and health management design for rotary machinery systems: reviews, methodology and applications. Mech Syst Signal Process2014; 42(1–2): 314–334.
6.
SiXSWangWHuCH, et al. Remaining useful life estimation. Eur J Oper Res2011; 213(1): 1–14.
7.
WangHKLiYFLiuY, et al. Remaining useful life estimation under degradation and shock damage. Proc IMechE, Part O: J Risk and Reliability2015; 229(3): 200–208.
8.
WangY-FTsengS-TLindqvistBH, et al. End of performance prediction of lithium-ion batteries. J Qual Technol2019; 51: 198–213.
9.
SutharssanTMontalvaoDChenYK, et al. A review on prognostics and health monitoring of proton exchange membrane fuel cell. Renew Sustain Energ Rev2017; 75: 440–450.
10.
HengAZhangSTanACC, et al. Rotating machinery prognostics: state of the art, challenges and opportunities. Mech Syst Signal Process2009; 23(3): 724–739.
11.
ZioE.Diagnostics and prognostics of engineering systems: methods and techniques. Hershey, PA: IGI Global, 2012.
12.
QiuJSethBBLiangSY, et al. Damage mechanics approach for bearing lifetime prognostics. Mech Syst Signal Process2002; 16(5): 817–829.
13.
LuoJPattipatiKRQiaoL, et al. Model-based prognostic techniques applied to a suspension system. IEEE Trans Syst Man Cybern Syst2008; 38(5): 1156–1168.
14.
ClarottiCLannoyAOdinS, et al. Detection of equipment aging and determination of the efficiency of a corrective measure. Reliab Eng Syst Saf2004; 84(1): 57–64.
15.
JardineABanjevicD. Interpretation of inspection data emanating from equipment condition monitoring tools: method and software. In: ArmijoYM (ed.) Mathematical and statistical methods in reliability. Singapore: World Scientific Publishing Company, 2005, pp. 263–278.
16.
DePoldHRGassFD. The application of expert systems and neural networks to gas turbine prognostics and diagnostics. J Eng Gas Turbine Power1999; 121: 607–612.
17.
CremonaMALiuBHuY, et al. Predicting railway wheel wear under uncertainty of wear coefficient, using universal kriging. Reliab Eng Syst Saf2016; 154: 49–59.
18.
ZemouriRRacoceanuDZerhouniN.Recurrent radial basis function network for time-series prediction. Eng Appl Artif Intell2003; 16(5–6): 453–463.
19.
SonJZhouSSankavaramC, et al. Remaining useful life prediction based on noisy condition monitoring signals using constrained Kalman filter. Reliab Eng Syst Saf2016; 152: 38–50.
20.
PechtMJaaiR.A prognostics and health management roadmap for information and electronics-rich systems. Microelectron Reliab2010; 50(3): 317–323.
21.
HuCYounBDWangP, et al. Ensemble of data-driven prognostic algorithms for robust prediction of remaining useful life. Reliab Eng Syst Saf2012; 103: 120–135.
22.
RezaeianjouybariBShangY.Deep learning for prognostics and health management: state of the art, challenges, and opportunities. Measurement2020; 163: 107929.
23.
JavedKGouriveauRZerhouniN, et al. A feature extraction procedure based on trigonometric functions and cumulative descriptors to enhance prognostics modeling. In: IEEE conference on prognostics and health management, Gaithersburg, MD, USA, 2013.
24.
CelayaJRWysockiPVashchenkoV, et al. Accelerated aging system for prognostics of power semiconductor devices. In: IEEE Autotestcon, Orlando, FL, 2010.
25.
FardNLiC.Optimal simple step stress accelerated life test design for reliability prediction. J Stat Plan Inference2009; 139(5): 1799–1808.
26.
HamadaMS.Bayesian analysis of step-stress accelerated life tests and its use in planning. Qual Eng2015; 27(3): 276–282.
27.
OcakHLoparoKADiscenzoFM.Online tracking of bearing wear using wavelet packet decomposition and probabilistic modeling: a method for bearing prognostics. J Sound Vib2007; 302(4–5): 951–961.
28.
NguyenDNDieulleLGrallA.Remaining useful lifetime prognosis of controlled systems: a case of stochastically deteriorating actuator. Math Probl Eng2015; 2015: 1–16.
29.
LindeYBuzoAGrayR.An algorithm for vector quantizer design. IEEE Trans Commun1980; 28(1): 84–95.
30.
KotsiantisSKanellopoulosD.Discretization techniques: a recent survey. GESTS Int Trans Comput Sci Eng2006; 32(1): 47–58.
31.
ViterbiA.Error bounds for convolutional codes and an asymptotically optimum decoding algorithm. IEEE Trans Inf Theory1967; 13: 260–269.
32.
GrinsteadCMSnellJL. Markov chains: introduction to probability. Providence, RI: American Mathematical Society, 1997.
33.
VenkatasubramanianVRengaswamyRYinK, et al. A review of process fault detection and diagnosis Part I: quantitative model-based methods. Comput Chem Eng2003; 27(3): 293–311.
34.
FrankPMDingX.Survey of robust residual generation and evaluation methods in observer-based fault detection systems. J Process Control1997; 7(6): 403–424.
35.
ChenJPattonRJ. Robust model-based fault diagnosis for dynamic systems. In: CaiK-Y (ed.) The International series on Asian studies in computer and information science. New York; Philadelphia: Springer US, 1999, p. 356.
36.
JohanssonABaskMNorlanderT.Dynamic threshold generators for robust fault detection in linear systems with parameter uncertainty. Automatica2006; 42(7): 1095–1106.
37.
KhanAQDingSX.Threshold computation for fault detection in a class of discrete-time nonlinear systems. Int J Adapt Control Signal Process2011; 25: 407–429.
38.
ChowEWillskyA.Analytical redundancy and the design of robust failure detection systems. IEEE Trans Automat Contr1984; 29: 603–614.
39.
GertlerJ.Fault detection and isolation using parity relations. Control Eng Pract1997; 5(5): 653–661.
40.
StaroswieckiMComtet-VargaG.Analytical redundancy relations for fault detection and isolation in algebraic dynamic systems. Automatica2001; 37: 687–699.
41.
RambeauxFHamelinFSauterD.Optimal thresholding for robust fault detection of uncertain systems. Int J Robust Nonlinear Control2000; 10: 1155–1173.
42.
HenryDZolghadriA.Norm-based design of robust FDI schemes for uncertain systems under feedback control: comparison of two approaches. Control Eng Pract2006; 14(9): 1081–1097.
43.
GrenailleSHenryDZolghadriA.A method for designing fault diagnosis filters for LPV polytopic systems. J Contr Sci Eng2008; 2008: 231697.
44.
SamantarayAKMedjaherKOuld BouamamaB, et al. Diagnostic bond graphs for online fault detection and isolation. Simul Model Pract Theory2006; 14(3): 237–262.
45.
DjeziriMAMerzoukiRBouamamaBO, et al. Robust fault diagnosis by using bond graph approach. IEEE/ASME Trans Mechatron2007; 12(6): 599–611.
46.
DakilMSimonCBoukhobzaT.Reliability and availability analysis of the structural observability of bilinear systems: a graph-theoretical approach. Proc IMechE, Part O: J Risk and Reliability2014; 228(3): 218–229.
47.
BlankeMKinnaertMLunzeJ, et al. Diagnosis and fault-tolerant control. 2nd ed.Berlin; New York: Springer, 2006.
48.
BaumLE.An equality and associated maximization technique in statistical estimation for probabilistic functions of Markov processes. Inequalities1972; 3: 1–8.
49.
DoMN.Digital signal processing mini-project: an automatic speaker recognition system. Lausanne, Switzerland: Audio Visual Communication Laboratory, Swiss Federal Institute of Technology, 1994.
50.
ErtuncHMLoparoKAOcakH.Tool wear condition monitoring in drilling operations using hidden Markov models (HMMs). Int J Mach Tools Manuf2001; 41(9): 1363–1384.
51.
LeeJMKimSJHwangY, et al. Diagnosis of mechanical fault signals using continuous hidden Markov model. J Sound Vib2004; 276(3–5): 1065–1080.
52.
LiZWuZHeY, et al. Hidden Markov model-based fault diagnostics method in speed-up and speed-down process for rotating machinery. Mech Syst Signal Process2005; 19(2): 329–339.
53.
NtalampirasSSoupionisYGiannopoulosG.A fault diagnosis system for interdependent critical infrastructures based on HMMs. Reliab Eng Syst Saf2015; 138: 73–81.
54.
ChinnamRBBaruahP.Autonomous diagnostics and prognostics through competitive learning driven HMM-based clustering. In: Proceedings of the international joint conference on neural networks, Portland, OR, 2003, vol. 4, pp.2466–2471. IEEE.
55.
ZhangXXuRKwanC, et al. An integrated approach to bearing fault diagnostics and prognostics. In: Proceedings of the 2005, American control conference, Portland, OR, 2005, vol. 1, pp.2750–2755. IEEE.
56.
SoualhiARazikHClercG, et al. Prognosis of bearing failures using hidden Markov models and the adaptive neuro-fuzzy inference system. IEEE Trans Ind Electron2014; 61(6): 2864–2874.
57.
FortAMugnainiMVignoliV.Hidden Markov models approach used for life parameters estimations. Reliab Eng Syst Saf2015; 136: 85–91.
58.
LingZHRichmondKYamagishiJ, et al. Integrating articulatory features into HMM-based parametric speech synthesis. IEEE Trans Audio Speech Lang Processing2009; 17: 1171–1185.
59.
CeleuxGDurandJB.Selecting hidden Markov model state number with cross-validated likelihood. Comput Stat2008; 23(4): 541–564.
60.
McLachlanGPeelD.Assessing the number of components in mixture models. Finite Mixture Models2005; 1: 175–220.
61.
DongWLiuSFangZ, et al. A model based on hidden graphic evaluation and review technique network to evaluate reliability and lifetime of multi-state systems. Proc IMechE, Part O: J Risk and Reliability2019; 233(3): 369–378.
62.
RabinerLR.A tutorial on hidden Markov models and selected applications in speech recognition. Proc IEEE1989; 77(2): 257–286.
63.
VachtsevanosGJVachtsevanosGJ. Intelligent fault diagnosis and prognosis for engineering systems. Hoboken, NJ: Wiley, 2006.
64.
KurfessTBillingtonSLiangS. Advanced diagnostic and prognostic techniques for rolling element bearings. In: WangLGaoRX (eds.) Condition monitoring and control for intelligent manufacturing. London: Springer, 2006, pp.137–165.
65.
MedjaherKTobon-MejiaDAZerhouniN.Remaining useful life estimation of critical components with application to bearings. IEEE Trans Reliab2012; 61(2): 292–302.
66.
BachirSTnaniSTrigeassouJC, et al. Diagnosis by parameter estimation of stator and rotor faults occurring in induction machines. IEEE Trans Ind Electron2006; 53(3): 963–973.
67.
BimaleK.Modern power electronics and AC drives. Upper Saddle River: Prentice-Hall, 2001.
68.
RajashekaraKKawamuraAMatsuseK.Sensorless control of AC motor drives: speed and position sensorless operation. New York, NY: IEEE Press, 1996.
69.
NguyenDNDieulleLGrallA.Feedback control system with stochastically deteriorating actuator: remaining useful life assessment. In: World congress, Cape Town, South Africa, 2014, vol. 1, pp.3244–3249.
70.
AggabTKratzFVrignatP, et al. Remaining useful life prediction method using an observer and statistical inference estimation methods. In: Annual conference of the PHM society (PHM17), St Petersburg, FL, 2017.
71.
KhelassiATheilliolDWeberP. Control design for over-actuated systems based on reliability indicators. In: UKACC international conference on control, Coventry, UK, 2010.
72.
Gonzalez-ContrerasBMTheilliolDSauterD. Actuator fault tolerant controller synthesis based on second order information. In: European control conference, Kos, Greece, 2007.
73.
AyadiAHajjiSSmaouiM, et al. Dynamic high-gain observer to estimate pneumatic actuator temperatures. J Dyn Syst Meas Control2016; 138(2): 1–7.
74.
Gucik-DerignyDOutbibROuladsineM.A comparative study of unknown-input observers for prognosis applied to an electromechanical system. IEEE Trans Reliab2016; 65(2): 704–717.
75.
DaigleMJGoebelK.A model-based prognostics approach applied to pneumatic valves. Int J Progn Health Manage2011; 2(2): 84–99.
76.
WelteT.Deterioration and maintenance models for components in hydropower plants. Doctoral dissertation, Department of Production and Quality Engineering, Norwegian University of Science and Technology, Trondheim, 2008.
77.
VrignatPAvilaMDucultyF, et al. Maintenance policy: degradation laws versus hidden Markov model availability indicator. Proc IMechE, Part O: J Risk and Reliability2012; 226(2): 137–155.
