Simulation Based Scheduling Applying Petri Nets with Sequences and Priorities

The paper deals with modelling and simulation of scheduling and sequencing problems based on Petri nets. In particular, the Petri-net models are simulated over the time domain and a simulation-based optimisation is implemented to optimise the input sequences. A new conflict resolution is implemented and a sophisticated way of defining firing sequences is developed. This enables the optimisation of scheduling problems by automated changing and evaluating the used sequences. The optimisation problem is solved by heuristic algorithms, including genetic algorithms, simulated annealing and threshold accepting. All these methods are implemented in the so called MATLAB PetriSimM toolbox which offers the capability of modelling, simulation, and optimisation of Timed, Coloured, and Stochastic Petri nets. For comparison, a Petri net based heuristic search is implemented, which is based on the reachability tree exploration. The algorithms are compared and tested in a case study dealing with modelling, simulation and optimisation of a production cell.