Abstract:The reachability analysis algorithm explores the state space of a timed automaton by enumeration of symbolic states. Each symbolic state consists of a location and a time zone which are conjunctions of automatic formulae in the form x-y ≤(<)n. Sometimes the amount of generated symbolic states is very large, the memory required to store the generated symbolic states is not feasible. In this paper, we present an approach to reduce the memory requirement of the reachability analysis algorithm. By analyzing the dependence relation between symbolic states, we can expand some of the symbolic states by removing specific kinds of atomic formulae without changing the reachability analysis result. The expanded states can contain more symbolic states. Removing these contained states can reduce the memory requirement of reachability analysis. The case studies presented in this paper show that our algorithm can save memory in the practical application efficiently.

An Algorithm to Dynamically Reduce the State Space of Timed Automata during the Reachability Analysis