LAUSR

In recent years, the research on robust deadlock control has become increasingly popular in automated manufacturing systems (AMSs) because resource failures may lead any system to stagnation, e.g., deadlock. In this article, we study robust supervisory control issues in AMSs with multiple unreliable resources. Petri nets are used to model the considered AMSs that allow multi-quantity and multi-type of resource acquisitions. A set of integer linear programming formulations are introduced to detect a class of deadlocks that have the maximal number of dead transitions. By analysis, a deadlock is characterized by a saturated circuit, which only consists of a set of unmarked resources and a set of critical transitions. Based on the circuit, a linear marking constraint is developed to prevent such circuits from being saturated. A control place (monitor) with its control variable is thus designed for the constraint to prevent the deadlock from appearing even if some resource failures occur. Therefore, we can synthesize a robust deadlock supervisor, which can guarantee that the controlled system can implement the continual operations even if some unreliable resources fail. Finally, the theoretical analysis and comparative study are provided to elucidate the effectiveness and efficiency of our proposed method. Note to Practitioners—In practice, resource failures in automated manufacturing systems (AMSs) are common. Deadlock prevention control in AMSs allowing resource failures has gained more and more attention from researchers and practitioners. Most prior research is based on the enumeration of either siphons or perfect resource transition circuits whose number exponentially increases with the system scale. This means that the synthesized supervisor has a much complex structure. In this article, based on a special kind of circuits at a deadlock marking detected by using a set of mathematical formulations, we develop an effective and efficient method for AMSs with multiple unreliable resources to iteratively control deadlocks such that the controlled system can continue to operate smoothly even if some unreliable resources fail. The computational and comparative results show that our proposed approach can acquire more permissive states with a simpler supervisor.