Enhancement of Deadlock Detection and Recovery Policy for Flexible Manufacturing Systems by Using Resource Flow Graph Technique

It is necessary to solve the unexpected deadlock problem during competition between different manufacturing processes of Flexible Manufacturing Systems (FMSs). Once system deadlock occurs, the lack of shared resources and machines causes process outages. To prevent such errors, we introduced a Resource Flow Graph (RFG) based deadlock detection and recovery policy. This method was used to analyze the flow relation between operation and resource to generate corresponding control transitions. However, in related research on designing optimal deadlock solutions, the marking explosion problem often arises as the size and complexity of the model increase. Thus, we improved RFG-based methods by considering the subnet Pre-Idle Places of Petri Net (PN) model and reduced the scale of RFG. This method reduced the computational complexity and costs. We also introduced three classic PN models of FMS to demonstrate the proposed control policy in this study.