A polynomial deadlock avoidance policy for a class of assembly processes based on Petri nets

Deadlock avoidance policies (DAPs) are of vitally importance for system control in order that correct deadlock-free resource allocation decisions can be made and continuing system operations can be guaranteed. This paper aims to present a DAP for FASs with assembly processes as their most important operations. The DAP, which is based on a variant of the well-known Banker´s approach in computer science literature, consists of a nominal policy and three concrete algorithms. The time complexity of the proposed DAP is polynomial with the size of the STPNR Petri-net model, where the STPNR (system of tree process net with resources) Petri-net subclass has the advantage to model the flexible assembly systems (FASs) in a parameterized and modular way. Therefore, the DAP can be effectively implemented by an online monitoring and control system for large real-world systems.