Critical stages and their identification in large scale automated manufacturing systems via Petri nets

An automated manufacturing system (AMS) is constituted by numerous, whether sequential or parallel, processes interconnected via a limited number of shared resources. With critical stages as pivots, any monolithic AMS can be partitioned into interacted subsystems. For each of them, there is always an expectation to produce an optimal trajectory in terms of a sequence of events, leading it from the assumptive stating point to the designated destination. The interaction among all subsystems can be resorted to the shared resources, which can be assumed as a quite simple vector, updating the coordinator in a timely manner. Two process structures, i.e., sequential and parallel processes, are formulated in the context of Petri net theory. The former models stages with only precedence relationships whereas the latter models stages with concurrency relationships. Then, the distributed algorithms are developed for such systems with the aid of identified critical places. As a consequence, our research results in a distributed control strategy, polynomial in its complexity and adaptable for realtime control. Sequential and parallel processes as well as their combinations are allowed in the processes of each plant model.