A GRASP-Tabu Search Algorithm for Solving School Timetabling Problems

Abstract

This work proposes a hybrid approach to solve school timetabling prob- lems. This approach is a GRASP algorithm that uses a partially greedy procedure to construct an initial solution and attempts to improve the constructed solution using a Tabu Search algorithm. When an infeasi- ble solution without overlapping classes is generated, a procedure called Intraclasses-Interclasses is activated, trying to retrieve feasibility. If successful, it will be reactivated. This time, attempting to improve the timetable’s compactness as well as other requirements. Computational results show that the Intraclasses-Interclasses procedure speeds up the process of obtaining better quality solutions.