Optimal design of fault-tolerant Petri net controllers

This paper proposes an approach for the optimal design of fault-tolerant Petri net controllers. Given a system controller that is modeled as a Petri net, a fault-tolerant Petri net controller can be obtained by embedding the given Petri net controller into a larger Petri net controller that retains the functionality of the original controller, and uses additional places, tokens, and connections to allow the detection and identification of faults that may occur in the controller places. An algorithm is developed to systematically design this fault-tolerant Petri net controller in an optimal sense. The optimality is in terms of minimizing the sum of arc weights of the (input and output) incident matrices of the redundant controller. Such minimum sum of arc weights is useful for characterizing the minimum hardware cost required to implement the redundant controller. An example of the optimal design of a fault-tolerant Petri net controller for an automated guided vehicle (AGV) system is also provided to illustrate our approach.