Senior high school course scheduling using genetic algorithm

Abstract

Manual course scheduling can be very complex and take a long time, even sometimes violate hard and soft constraints. Soft constraints usually relate to teacher and student preferences as part of the schedule. The purpose of this study was to apply genetic algorithms (GA) to prevent the violation of hard constraints and minimize the violation of soft constraints. The GA in this study distributing population in some gruoups. The distributed GA generate groups of population and then after each iteration, the migration between groups will be conducted based on given probability of migration. The distributed GA applied to avoid the the premature convergence that could occur in original GA. The probability of migration observed in this study was 0, 0.1, 0.2, 0.3, 0.4, and 0.5. The study shown that the the distributed GA succeed to prevent violation of hard constraints, minimize the soft constraints violation and avoid the premature convergence.