Abstract:

An improved method for hierarchical modelbased diagnosis is proposed based on the work of Chittaro and Ranon. Two abstract processes are achieved simultaneously according to hierarchical representation and abstract condition when a new abstract level is derived. One is a fixed abstract process and the other is an automatic process. The computational complexity for modelbased diagnosis is further decreased by means of the method. Additional levels are minimized and the number of a diagnosis problem is reduced on the premise of  the equal candidate space of the most final abstract level. Furthermore, the final diagnosis can be derived by the refined process at level l(l>0) with out detailing each abstract level until the original level in any case. Finally, experimental results show that the efficiency of our algorithm is better than other algorithms for modelbased hierarchical diagnosis.

Key words: modelbased diagnosis； hierarchical diagnosis； structural abstraction； dynamic hierarchical representation