Extending design space optimization heuristics for use with stochastic colored Petri nets

Automatic design optimization of complex systems is a time-consuming task. This paper presents multiphase variants of well-known heuristics for an efficient indirect system optimization with simulation. The tradeoff between accuracy and achievable speedup is analyzed for examples including a benchmark function as well as simulated values for a stochastic colored Petri net application. The analysis are carried out with TOE (TimeNET optimization environment), a software prototype implemented for this task recently.