Matrix Approach to Verification of Nonblocking for Partially Observed Discrete Event Systems

To ensure the safety of discrete event systems, detecting the nonblocking property of the system is an effective approach. This paper proposes a matrix-based method to address the nonblocking property detection problem for partially observed discrete event systems. Firstly, we introduce nondeterministic output functions to represent the state dependency and nondeterministic of partial event observations, and model the partially observed discrete event systems as Mealy finite state machines with nondeterministic output functions. Furthermore, we utilize a novel matrix method called semi-tensor product to establish algebraic expressions for the finite state machine. Based on the obtained algebraic expressions, we analyze the nonblocking property verification problems of the finite state machine from a state perspective. We propose an algebraic algorithm to identify potentially blocking states in the finite state machine.