Best Paper Award

Hierarchical Control of Concurrent Discrete Event Systems with Linear Temporal Logic Specifications



  This paper deals with the supervisory control problem for concurrent discrete event systems. A concurrent discrete event system consists of several subsystems that are partially synchronized via shared events. The authors proposed a hierarchical control architecture to this problem. The architecture consists of local supervisors and a coordinator. Each supervisor ensures local constraints, while the coordinator guarantees the global constraint through deciding which shared events to prohibit. The authors adopted a game theory-based approach to solving this problem. They obtain local supervisors by Rabin games for the local constraints and then construct a coordinator of them with a Rabin game for the global constraint. The system controlled by this architecture satisfies not only the given constraints but is also deadlock-free.

  The originality of this paper is that Linear Temporal Logic (LTL) is used to describe control specifications. Formal languages have been widely used but may not be appropriate to describe the desired behavior of an actual system. LTL helps us to describe and understand this behavior thanks to its rich expressiveness.

  The usefulness of this paper is that the proposed method is applicable to various systems such as flexible factory automation, cooperative control of a group of robots, and intelligent transportation systems. In the control of such systems, the computation complexity becomes a problem. The proposed hierarchical control architecture can reduce the computation complexity by dividing the whole system into subsystems and designing them individually.

  Due to the novelty and usefulness described above, this paper deserves the IEICE Best Paper Award.