Developing a multiobjective optimization scheduling system for a screw manufacturer: A refined genetic algorithm approach

Abstract

Over time, the traditional single-objective job shop scheduling method has grown increasingly incapable of meeting the requirements of contemporary business models; thus, a multiobjective scheduling solution is required. Because of changing orders, understanding the schedule and output is a constant challenge when using a traditional manual schedule, particularly among manufacturers that produce various products. The multiobjective optimization genetic algorithm (MOGA) is a relatively superior method of solving multiobjective optimization problems; therefore, we used a MOGA to solve flexible job-shop problems for a middle-scale screw manufacturer in Taiwan. For solving the problems of incorrect jobs assign and diversity problem of traditional genetic algorithm (GA) caused by encoding method when applying traditional GA in the flexible manufacturing environment, a refined GA was proposed. Two-phase test has performed for proposed approach, using a classical benchmark of distributed and flexible jobs-shop scheduling problem, and 80 set of work orders, the empirical results indicated that the proposed model yielded substantial savings, regardless of the total order completion time, machine retooling rate, and average machine load rate.