An improved symbiotic organisms search algorithm for low-yield stepper scheduling problem

A stepper in a lithography area is the bottleneck machine of a semiconductor manufacturing process. Its effective scheduling in low-yield scenes can improve throughput and profits of a semiconductor wafer fabrication facility. This paper presents an opposition-based Symbiotic Organisms Search with a catastrophe phase algorithm (OBSOS-CA) to minimize the makespan of this scheduling problem. The opposition-based learning technique is used to increase the population diversity in the initial and parasitism phases of Symbiotic Organisms Search (SOS). Moreover, we add a catastrophe phase containing three parts. When the algorithm is trapped in a local optimum, a catastrophe judgement and an extinction operation are used to jump out of the local optimal solution. Meanwhile, variable neighborhood descent is employed in the mutualism phase and commensalism phase of SOS as the explosion operation thereby strengthening the ability of local search. Simulation results demonstrate that OBSOS-CA is effective for a low-yield stepper scheduling problem.