Newcastle University ePrints

The timetabling of lecturers, seminars, practical sessions and examinations is a core business process for academic institutions. A feasible timetable must satisfy hard constraints, an optimum timetable will additionally satisfy soft constraints, which are not absolutely essential. An Ant Colony based Timetabling Tool (ANCOTT) has been developed for solving timetabling problems. New variants of Ant Colony Optimisation (ACO) called Best-Worst Ant System (BWAS) and Best-Worst Ant Colony System (BWACS) were embedded in the ANCOTT program. Local Search (LS) strategies were developed and embedded into BWAS and BWACS to enhance their efficiency and to help find the best timetable with the lowest number of soft constraint violations. Statistical tools for experimental design and analysis were adopted to investigate the factors affecting the BWAS performance. Eight benchmarking instant problems were used for benchmarking the performance. The proposed LS enhanced both BWAS and BWACS performances by up to 70% but required longer execution time.