Connexionist-Systems-Based Long Term Prediction Approaches for Prognostics

Prognostics and Health Management aims at estimating the remaining useful life of a system (RUL) , i.e. the remaining time before a failure occurs. It benefits thereby from an increasing interest: prognostic estimates (and related decision-making processes) enable increasing availability and safety of industrial equipment while reducing costs. However, prognostics is generally based on a prediction step which, in the context of data-driven approaches as considered in this paper, can be hard to achieve because future outcomes are in essence difficult to estimate. Also, a prognostic system must perform sufficient long term estimates, whereas many works focus on short term predictions. Following that, the aim of this paper is to formalize and discuss the connexionist-systems-based approaches to ensure multi-step ahead predictions for prognostics. Five approaches are pointed out: the Iterative, Direct, DirRec, Parallel, and MISMO approaches. Conclusions of the paper are based, on one side, on a literature review; and on the other side, on simulations among 111 time series prediction problems, and among a real engine fault prognostics application. These experiments are performed using the exTS (evolving extended Takagi-Sugeno system). As for comparison purpose, three types of performances measures are used: prediction accuracy, complexity (computational time), and implementation requirements. Results show that all three criteria are never optimized at the same time (same experiment), and best practices for prognostics application are finally pointed out.

[1]  Stephen A. Billings,et al.  Dual-orthogonal radial basis function networks for nonlinear time series prediction , 1998, Neural Networks.

[2]  Antti Sorjamaa,et al.  Multiple-output modeling for multi-step-ahead time series forecasting , 2010, Neurocomputing.

[3]  Wilson Wang,et al.  An adaptive predictor for dynamic system forecasting , 2007 .

[4]  Amir F. Atiya,et al.  Multi-step-ahead prediction using dynamic recurrent neural networks , 1999, IJCNN'99. International Joint Conference on Neural Networks. Proceedings (Cat. No.99CH36339).

[5]  Amaury Lendasse,et al.  Long-term prediction of time series by combining direct and MIMO strategies , 2009, 2009 International Joint Conference on Neural Networks.

[6]  W. Peizhuang Pattern Recognition with Fuzzy Objective Function Algorithms (James C. Bezdek) , 1983 .

[7]  Chuntian Cheng,et al.  A new indirect multi-step-ahead prediction model for a long-term hydrologic prediction , 2008 .

[8]  Noureddine Zerhouni,et al.  Towards a Neuro-Fuzzy System for Time Series Forecasting in Maintenance Applications , 2008 .

[9]  Noureddine Zerhouni,et al.  Reducing arbitrary choices in model building for prognostics: An approach by applying parsimony principle on an evolving neuro-fuzzy system , 2011, Microelectron. Reliab..

[10]  Rob J Hyndman,et al.  25 years of time series forecasting , 2006 .

[11]  Jie Zhang,et al.  Long-term prediction models based on mixed order locally recurrent neural networks , 1998 .

[12]  Joseph Mathew,et al.  Rotating machinery prognostics. State of the art, challenges and opportunities , 2009 .

[13]  M.J. Roemer,et al.  Prognostic enhancements to diagnostic systems for improved condition-based maintenance [military aircraft] , 2002, Proceedings, IEEE Aerospace Conference.

[14]  Frank L. Lewis,et al.  Intelligent Fault Diagnosis and Prognosis for Engineering Systems , 2006 .

[15]  Nicolas Huck,et al.  Pairs trading and outranking: The multi-step-ahead forecasting case , 2010, Eur. J. Oper. Res..

[16]  Vladan Babovic,et al.  Multi-step-ahead model error prediction using time-delay neural networks combined with chaos theory , 2010 .

[17]  P. Angelov,et al.  Evolving Fuzzy Systems from Data Streams in Real-Time , 2006, 2006 International Symposium on Evolving Fuzzy Systems.

[18]  Hsiao-Tien Pao,et al.  Forecasting energy consumption in Taiwan using hybrid nonlinear models , 2009 .

[19]  M. Farid Golnaraghi,et al.  Prognosis of machine health condition using neuro-fuzzy systems , 2004 .

[20]  Jie Liu,et al.  A multi-step predictor with a variable input pattern for system state forecasting , 2009 .

[21]  Hubert Cardot,et al.  A new boosting algorithm for improved time-series forecasting with recurrent neural networks , 2008, Inf. Fusion.

[22]  Dobrivoje Popovic,et al.  Computational Intelligence in Time Series Forecasting: Theory and Engineering Applications (Advances in Industrial Control) , 2005 .

[23]  Abhinav Saxena,et al.  Damage propagation modeling for aircraft engine run-to-failure simulation , 2008, 2008 International Conference on Prognostics and Health Management.

[24]  Douglas Kline,et al.  Methods for Multi-Step Time Series Forecasting Neural Networks , 2004 .

[25]  Leandro dos Santos Coelho,et al.  Multi-step ahead nonlinear identification of Lorenz’s chaotic system using radial basis neural network with learning by clustering and particle swarm optimization , 2008 .

[26]  Noureddine Zerhouni,et al.  Long term prediction approaches based on connexionist systems - A study for prognostics application , 2011, 2011 IEEE Conference on Prognostics and Health Management.

[27]  Rafael Gouriveau,et al.  Prognostics in switching systems: Evidential markovian classification of real-time neuro-fuzzy predictions , 2010, 2010 Prognostics and System Health Management Conference.

[28]  Peng Wang,et al.  Fault prognostics using dynamic wavelet neural networks , 2001, Artificial Intelligence for Engineering Design, Analysis and Manufacturing.

[29]  Amaury Lendasse,et al.  Time series prediction using DirRec strategy , 2006, ESANN.

[30]  Farid Atry,et al.  Multi-step ahead forecasts for electricity prices using NARX: A new approach, a critical analysis of one-step ahead forecasts , 2009 .

[31]  Kun Yang,et al.  A combining condition prediction model and its application in power plant , 2004, Proceedings of 2004 International Conference on Machine Learning and Cybernetics (IEEE Cat. No.04EX826).

[32]  Bo-Suk Yang,et al.  Machine condition prognosis based on regression trees and one-step-ahead prediction , 2008 .

[33]  D.P. Filev,et al.  An approach to online identification of Takagi-Sugeno fuzzy models , 2004, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[34]  M. Farid Golnaraghi,et al.  A neuro-fuzzy approach to gear system monitoring , 2004, IEEE Transactions on Fuzzy Systems.

[35]  Rob J Hyndman,et al.  Another look at measures of forecast accuracy , 2006 .

[36]  Diyar Akay,et al.  Comparison of direct and iterative artificial neural network forecast approaches in multi-periodic time series forecasting , 2009, Expert Syst. Appl..

[37]  Bo-Suk Yang,et al.  Multi-step ahead direct prediction for machine condition prognosis using regression trees and neuro-fuzzy systems , 2013 .

[38]  Daming Lin,et al.  A review on machinery diagnostics and prognostics implementing condition-based maintenance , 2006 .

[39]  Richard C.M. Yam,et al.  Intelligent Predictive Decision Support System for Condition-Based Maintenance , 2001 .

[40]  James C. Bezdek,et al.  Pattern Recognition with Fuzzy Objective Function Algorithms , 1981, Advanced Applications in Pattern Recognition.