Research on a New Combined Algorithm for Computing Minimal Hitting Sets

Abstract

In model-based diagnosis, a key step is to compute the minimal hitting sets (in brief, MHS) from the minimal conflict sets (in brief, MCS). Taking aim at the problem of existing algorithms, a new combined algorithm which can compute MHS was presented. The algorithm is consisted of three parts: the first part is to build a classified hitting sets tree (in brief, CHS-tree); the second part is the Recursive Boolean Algorithm which can compute the MHS from CHS-tree; and the third part given the resolve of the case that some new conflict sets were inserted. The combined algorithm was improved in the aspects as follows. (1) needs not being pruned; (2) can derive all MHS directly; (3) when some new conflict sets were inserted, it was not necessary to rebuild CHS-tree. Finally example of some case was given to demonstrate the algorithm. Experiment results show that the combined algorithm is effective and adapts to resolve practical diagnosis problem.