Loop reduction techniques for reachability analysis of linear hybrid automata

Abstract

The problem of reachability analysis of linear hybrid automata (LHA) is very difficult. This paper considers to improve the efficiency of the reachability analysis by optimizing the structures of LHA. We identify two types of loops called the flexible loops and the zero loops, and present the techniques to replace the repetitions of those loops in the behavior of LHA with finite sequences of locations and in the meantime simplify the associated constraints. The techniques work not only for the polyhedral computing based algorithms but also for the bounded model checkers.