Title :
A robust method for fault isolation
Author_Institution :
Dept. of Chem. Eng., Minnesota Univ., Duluth, MN, USA
Abstract :
This work concerns fault classification in the model based directional residual approach. To achieve a balance between optimality and robustness, a minimax method is suggested, which minimizes the maximum expected loss from misclassification. An application example is provided to illustrate the use of this method
Keywords :
decision theory; fault diagnosis; minimax techniques; multidimensional systems; probability; decision theory; fault classification; fault diagnosis; fault isolation; finite dimensional systems; minimax method; model based directional residual; probability; robustness; Chemical engineering; Costs; Diagnostic expert systems; Fault detection; Hardware; Minimax techniques; Robustness; Statistical analysis; Testing; Working environment noise;
Conference_Titel :
American Control Conference, 1997. Proceedings of the 1997
Conference_Location :
Albuquerque, NM
Print_ISBN :
0-7803-3832-4
DOI :
10.1109/ACC.1997.611810
