Combining fault detection and process optimization in manufacturing systems using first-order hybrid Petri nets

Describes a hierarchical two level modeling and control framework for real-time manufacturing processes, based on hybrid Petri nets, and capable of integrating monitoring and fault detection techniques along with performance optimization procedures. At the higher level, first-order hybrid Petri nets are used to generate the asynchronous concurrent events characterizing the nominal optimal system behavior, also providing the machine production rates. At the lower level, a real time scheduler and a discrete Petri net communicate with the field-bus devices to force the machine to produce at the rates provided by the upper level, and monitor the occurring events by keeping track of part movements. Whenever a process failure is detected, the diagnosis procedure is triggered, and a recovery sequence is imposed.