论文信息 - DEVELOPING DATA MINING-BASED PROGNOSTIC MODELS FOR CF-18

DEVELOPING DATA MINING-BASED PROGNOSTIC MODELS FOR CF-18

The CF-18 aircraft is a complex system for which a variety of data are systematically being recorded: operational flight data from sensors and Built-In Test Equipment (BITE) and maintenance activities recorded by personnel. These data resources are stored and used within the operating organization but new analytical and statistical techniques and tools are being developed that could be applied to these data to benefit the organization. This paper investigates the utility of readily available CF-18 data to develop data mining-based models for prognostics and health management (PHM) systems. We introduce a generic data mining methodology developed to build prognostic models from operational and maintenance data and elaborate on challenges specific to the use of CF-18 data from the Canadian Forces. We focus on a number of key data mining tasks including: data gathering, information fusion, data pre-processing, model building, and evaluation. The solutions developed to address these tasks are described. A software tool developed to automate the model development process is also presented. Finally, the paper discusses preliminary results on the creation of models to predict F404 No. 4 Bearing and MFC (Main Fuel Control) failures on the CF-18.Copyright © 2010 by National Research Council of Canada

Sylvain Létourneau | Jeff W. Bird | Marvin Zaluski | Chunsheng Yang

[1] Grigorios Tsoumakas,et al. Effective Voting of Heterogeneous Classifiers , 2004, ECML.

[2] Larry A. Rendell,et al. A Practical Approach to Feature Selection , 1992, ML.

[3] Mark A. Hall,et al. Correlation-based Feature Selection for Discrete and Numeric Class Machine Learning , 1999, ICML.

[4] Ian H. Witten,et al. Data mining: practical machine learning tools and techniques, 3rd Edition , 1999 .

[5] Bernard Zenko,et al. Stacking with Multi-response Model Trees , 2002, Multiple Classifier Systems.

[6] Thomas G. Dietterich. An Experimental Comparison of Three Methods for Constructing Ensembles of Decision Trees: Bagging, Boosting, and Randomization , 2000, Machine Learning.

[7] A. Atamer. Comparison of FMEA and field-experience for a turbofan engine with application to case based reasoning , 2004, 2004 IEEE Aerospace Conference Proceedings (IEEE Cat. No.04TH8720).

[8] Ralph Weischedel,et al. PERFORMANCE MEASURES FOR INFORMATION EXTRACTION , 2007 .

[9] Chunsheng Yang,et al. Learning to predict train wheel failures , 2005, KDD '05.

[10] R. W. Cue,et al. Engine Performance Monitoring and Troubleshooting Techniques for the CF-18 Aircraft , 1991 .

[11] Stan Matwin,et al. Data mining to predict aircraft component replacement , 1999, IEEE Intell. Syst..