Machine allocation in a semiconductor wafer fabrication system

General queueing modeling method, Stochastic Timed Automation, is employed to model the Semiconductor Wafer Fabrication System because of its flexibility and fidelity. The machine allocation is formulated as a stochastic integer programming problem with the objective to optimize the throughput of the Semiconductor Wafer Fabrication System. And we propose an Optimal Computing Budget Allocation-based Stochastic Greedy Search algorithm to dynamically allocate machines, each time we allocate only one machine into the best performed station until all the machines are allocated into the Semiconductor Wafer Fabrication System. Numerical examples suggest our method is efficient to find a good allocation solution.