Genetic Algorithm for the Timetabling Problem in Universities

The timetabling problem in universities is to assign the classrooms into classes. This problem is somehow similar to that of assigning machines into jobs in manufacturing systems to minimize the makespan and tardiness. With this observation, the timetabling problem is described by a triplet as a parallel machine scheduling problem with machine eligibility restrictions. Then, a mathematical model is developed for the problem. By analyzing the features of the model, a heuristic that combines dispatching rules and genetic algorithm is proposed to solve the problem. Techniques for coding, individual fitness, crossover, and mutation are presented to implement the genetic algorithm. The proposed method is tested by using a practical case problem and it is shown to be effective and feasible.