Finding the optimal parameter setting of processes with regard to predefined optimization criteria like minimal costs, minimal resources, maximum profit, etc., is one of the most important tasks of simulation models. Additional demands (such as predefined timespans for processing "just in time" tasks, or predefined quality) have to be taken into account frequently so that the usual approximation methods (e.g., mean value calculations) fail.
With PACE the processes in simulation models can be optimized by means of graphical methods, or fully automatically with the use of mathematical methods. PACE supports the following methods:
Repeated execution of a model having the relevant parameters changed manually before every simulation run until certain outputs (curves, tables, etc.) have the required appearance (e.g., see the demo model "e-commerce").
Automatic repetition of the complete model or of select submodels with parameters changed automatically as specified by the user and listing of the results and/or graphic representation of the results. The optimum can be found by the program or read from the graphical representation (so-called graphic optimization).
Automatic determination of the optimum of a model or submodel using integrated mathematical optimization methods (e.g., Hill Climbing, Simplex, Genetic Methods, Threshold Acceptance).