A multi-objective dynamic scheduling approach using multiple attribute decision making in semiconductor manufacturing

This paper proposes a multi-objective dynamic scheduling approach that combines three attributes based on a hybrid multiple attribute decision making (MADM) technique. With consideration of two kinds of uncertainties, three advanced attributes are specifically designed to optimize four key performance measures. Each attribute is converted into a relative closeness to the ideal alternative, and collaborate with the others at each scheduling instant. Furthermore, a cutoff method is used to make the proposed approach adapt to full-scale semiconductor manufacturing systems (SMSs) better. Results in two SMS models show that the proposed approach is validated for the real-time multi-objective scheduling problem.