Optimal configuration of diagnostic sensors in manufacturing system based on Petri net

There is an optimal solution for assigning less sensors to a manufacturing system and observing the firing of all faulty transitions. A modified Petri net model and the definition of identifiability are proposed to model the discrete event system (DES). Based onthe PN model and the graph theory, a diagnostic sensor optimal algorithm is developed in this study. The algorithm depends on the topology of the PN model, not the Petri net initial markings. Given the PN model, the minimum number of sensors required and their configuration can be computed by the algorithm.A case study demonstrates the validity and efficiency of the method in a manufacturing system.