A Petri nets based deadlock avoidance policy for flexible manufacturing systems
A.R. Wang, Zonghui Li, J. Yjia
Abstract
A.R. Wang, Zonghui Li, J. Yjia
Abstract
Developed in this paper is a new deadlock avoidance strategy for flexible manufacturing systems, which is based on Petri nets and their structural analysis. The deadlock avoidance policy only uses partial reachability graph of the Petri net without the computations for the whole reachability graph, this deadlock avoidance approach is first to compute some special states or markings, such as deadlock markings, bad markings, and dangerous markings we call. In a bad state (marking), the system (Petri net) will inevitably reach a deadlock, whereas in a dangerous state, the system may not reach a deadlock if the firings of enabled transitions are properly controlled by supervisory control The deadlock avoidance policy presented here is in fact to make the system never reach a bad state. The major advantage of the technique lies in the fact that It is suitable for much larger Petri nets than that of the most existing deadlock avoidance policies and the supervisory control is maximally permissive. Meanwhile, the computation is successfully avoided for the whole reachability graph of the Petri net of the system.
OpenAlex reports 5 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Developed in this paper is a new deadlock avoidance strategy for flexible manufacturing systems, which is based on Petri nets and their structural analysis. The deadlock avoidance policy only uses partial reachability graph of the Petri net without the computations for the whole reachability graph, this deadlock avoidance approach is first to compute some special states or markings, such as deadlock markings, bad markings, and dangerous markings we call. In a bad state (marking), the system (Petri net) will inevitably reach a deadlock, whereas in a dangerous state, the system may not reach a deadlock if the firings of enabled transitions are properly controlled by supervisory control The deadlock avoidance policy presented here is in fact to make the system never reach a bad state. The major advantage of the technique lies in the fact that It is suitable for much larger Petri nets than that of the most existing deadlock avoidance policies and the supervisory control is maximally permissive. Meanwhile, the computation is successfully avoided for the whole reachability graph of the Petri net of the system.
Key concepts: Petri net, Reachability, Deadlock, Deadlock prevention algorithms, Computer science, Distributed computing, Graph, Computation