Rule extraction method in incomplete decision table for fault diagnosis based on discernibility matrix primitive

  • Authors:

  • Wentao Huang;Weijie Wang;Xuezeng Zhao

  • Affiliations:

  • School of Mechatronics Engineering, Harbin Institute of Technology, Harbin, P.R. China;School of Mechatronics Engineering, Harbin Institute of Technology, Harbin, P.R. China;School of Mechatronics Engineering, Harbin Institute of Technology, Harbin, P.R. China

  • Venue:
  • RSKT'08 Proceedings of the 3rd international conference on Rough sets and knowledge technology
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

Compared with extracting rules from complete data, it is more difficult to obtain rules from incomplete data for fault diagnosis. In this paper, based on the rough set theory, a method is proposed to directly extract optimal generalized decision rules from incomplete a decision table for fault diagnosis (IDTFD). The discernibility matrix primitive is defined and characterized to simplify the computing process. A definition of object-oriented discernibility matrix in IDTFD is also proposed. Using these concepts, an object-oriented discernibility function is constructed. With the basic equivalent forms in proposition logic such as distribution laws, absorption laws, a method is proposed to compute the minimal object-oriented reductions and to extract the optimal generalized decision rules in IDTFD. The proposed method is applied in fault diagnosis of operational states of an electric system. The effectiveness of this method is shown in our experiments.