This work addresses the problem of deadlock control in flexible assembly systems (FASs) with multiple resource requirements and parts loaded separately into the system. We use generalized Petri nets to model, analyze, and control these FASs. Two kinds of structural objects in the Petri net model, namely A-circuits and closedΩ-structures, can characterize deadlocks. Then a necessary and sufficient condition is derived for the liveness of FAS based on these two objects. By using integer linear programs, we compute all the problematic closedΩ-structures and A-circuits that are related to deadlocks. A specific design of the monitor is developed for each of these objects so as to prevent it from causing deadlock. By composing all these monitors, a deadlock prevention controller is proposed for a subclass of FASs under consideration. Importantly, this work generalizes the liveness condition of ordinary Petri net models to broader non-ordinary ones. Finally, an example can illustrate the usefulness and correctness of the proposed concepts and methods. Note to Practitioners—In the context of FAS, deadlocks are caused by two patterns: “resource circuit-wait” and “event circuit-wait”. The former indicates a circuit-wait relation among shared resources and has been studied extensively in the past decades, while the latter results from the assembly of different parts and has not received much attention. The analysis regarding event circuit-wait is the key to solving a deadlock problem in FAS. To date, there are a limited number of studies addressing the deadlock control problem for FASs with multiple resource requirements. Given such a complex FAS, based on its Petri net model, this paper 1) characterizes the above two circuit-wait patterns by structural objects, 2) presents a sufficient and necessary liveness condition, and 3) develops an effective deadlock prevention controller for a class of FASs under consideration. The proposed deadlock controller can well solve the deadlocks in concerned FAS applications, like automotive industry, semiconductor manufacturing, and so on. Especially, unlike some existing methods, our controller can be applied to the cases where parts of the same product are permitted to load separately into the system.
No takes yet. Share an insight, caveat, or question.
Feng et al. (2024) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: