Fault-Tolerant Synthesis for Multi-process Systems via Resource Sharing: A Discrete-Event Approach

This study introduces a novel framework for fault-tolerant supervisory control in Automated Manufacturing Systems (AMS), addressing the critical challenge of maintaining operational integrity in the presence of faults. Leveraging the Ramadge-Wonham supervisory control paradigm, we propose an innovative modeling approach that enables AMS to recover from faults and adhere to predefined specifications through efficient resource sharing. Our methodology commences by representing the system as a discrete-event system, incorporating resource sharing relations. We then introduce an incremental abstraction strategy to derive the system's fault model from its pre-fault abstraction. This paper contributes to the field by establishing conditions for the existence of a fault-tolerant supervisor in large AMS and detailing a method for synthesizing such a supervisor when feasible. The effectiveness of our proposed fault-tolerant synthesis method is validated through simulation on a model manufacturing system. This research not only addresses a significant industrial challenge but also offers a scalable and efficient solution for fault-tolerant supervisory control in complex manufacturing environments.