Moxnes, Erling, "Stochastic Optimization in Policy Space Using Simulation Models", 1996
ua435
Optimization has not been an important part of system dynamics thus far, and for several good reasons: Simulation models have been sufficient to point out greater potential for improvement than what policy makers have been willing o embrace. Hence presumably minor improvements of policies might have seemed of little value. Even simple simulation results have been difficult to communicate, indicating that the results of complicated optimization efforts would be even harder to convey. Optimization methods have not been able to capture the richness of real decision problems characterized by simultaneous appearance of dynamics, nonlinearities and uncertainties. Hence there has been uncertainty about the transferability of solutions from simple to complex environments.