Scheduling of FMS - Application of an Open Optimization System

Our goal is the development of an open optimization system for scheduling of manufacturing processes. We have already introduced the general structure of this system on the last conference [1] .
At first we regard a manufacturing system as a black box which is controlled by one or more control values (job sequences, job or task priorities, etc.), in general permutations. We prefer a simulation based method for scheduling. This means, we divide an optimization cycle in two independent steps: the search for one or more new control values (optimization module) and the assessment of these control values (simulation module). The assessment is done by a simulation model of the real manufacturing system. For search we use modern stochastic or heuristic algorithms, i.e. Genetic Adaption and Simulated Annealing.
An interface module guarantees the independence and an easy replacement of simulation and optimization module. So we can combine a particular simulation module with different optimization algorithms and vice versa. In addition this interface includes a monitor for watching the optimization process and an editor for creating objective functions. It is possible to create multi objective functions by pre-setting weights for each objective. The data exchange between the modules and the graphical user interface (GUI) is based on the well known Tcl/Tk.