Regression kernel for prognostics with support vector machines

Estimating the remaining useful life (RUL) of systems and/or equpipments has been an important goal for reliable, safe, and profitable operation of industrial plants. However, traditional mathematical and statistical modeling based approaches are difficult to design and they adapt poorly to the ever changing operating and environmental conditions in real-world industries. With recent developments in computational technologies, data storage, and industrial automation recording and storage of large amounts of historical plant data from embedded sensors and maintenance records have become easy. Availability of large data sets together with advancements in data driven machine learning algorithms has been the key driver for prognostic and diagnostic research in the industry as well as by academia. Nevertheless, developing generalized machine learning algorithms for the prognostic domain has been challenging due to the very nature of the problem. This paper describes some of these challenges and proposes a modified regression kernel that can be used by support vector regression (SVR) for prognostic problems. The method is tested on a simplified simulated time-series data set that is modeled to represent the challenges presented.

[1]  Frank L. Lewis,et al.  Intelligent Diagnosis and Prognosis of Industrial Networked Systems , 2011 .

[2]  Ravi Sankar,et al.  Time Series Prediction Using Support Vector Machines: A Survey , 2009, IEEE Computational Intelligence Magazine.

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

[4]  Robert Babuska,et al.  Railway Track Circuit Fault Diagnosis Using Recurrent Neural Networks , 2017, IEEE Transactions on Neural Networks and Learning Systems.

[5]  Tomoko Matsui,et al.  A Kernel for Time Series Based on Global Alignments , 2006, 2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07.

[6]  Shie-Jue Lee,et al.  A multiple-kernel support vector regression approach for stock market price forecasting , 2011, Expert Syst. Appl..

[7]  Donghua Zhou,et al.  Remaining useful life estimation - A review on the statistical data driven approaches , 2011, Eur. J. Oper. Res..

[8]  Bernhard Schölkopf,et al.  A tutorial on support vector regression , 2004, Stat. Comput..

[9]  Francis Eng Hock Tay,et al.  Modified support vector machines in financial time series forecasting , 2002, Neurocomputing.

[10]  Kay Chen Tan,et al.  Multimodal Degradation Prognostics Based on Switching Kalman Filter Ensemble , 2017, IEEE Transactions on Neural Networks and Learning Systems.

[11]  Enrico Zio,et al.  Failure and reliability prediction by support vector machines regression of time series data , 2011, Reliab. Eng. Syst. Saf..

[12]  Benoît Iung,et al.  Remaining useful life estimation based on stochastic deterioration models: A comparative study , 2013, Reliab. Eng. Syst. Saf..

[13]  Ying Peng,et al.  Current status of machine prognostics in condition-based maintenance: a review , 2010 .

[14]  Jiuping Xu,et al.  PHM-Oriented Integrated Fusion Prognostics for Aircraft Engines Based on Sensor Data , 2014, IEEE Sensors Journal.

[15]  A. A. Rodríguez,et al.  Support Vector Method for Robust ARMA System Identification , 2009 .

[16]  Bo-Suk Yang,et al.  Support Vector Machine for Machine Fault Diagnosis and Prognosis , 2008 .

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