Design and Implementation of a MIP Model for the Examination Timetabling Problem of Louvain School of Management

Details

Supervisors

Catanzaro, Daniele

Faculty

Louvain School of Management

Degree label

Master [120] en ingénieur de gestion

Abstract

There are two kinds of university timetabling problems, course timetabling and exams timetabling. In this case, we will study the latter. Examination timetabling problem's objective is to allocate exams into timeslots and auditoriums while satisfying hard and soft constraints to avoid any conflict. Hard constraints must not be violated at any costs. Whereas soft constraints should be satisfied as much as possible. This paper studies the current examination timetabling problem of the faculty of Louvain School of Management. The goal here is to implement the mixed integer programming (MIP) technique and automate the timetable creating process and compare the current solution of the faculty with the one to be implemented. An interview of the schedulers in charge of this task is performed to have an understanding of the situation at hand. Thanks to this interview, parameters, variables and preferences of the problem are defined. The next step is to collect the necessary data regarding involved courses, students, teachers and rooms to achieve the implementation of the solution. The problem is solved via the software Fico Xpress Workbench, using the mosel language. After executing the model, it is noticeable that the MIP meets the soft constraints as much as possible. Most of the exams are assigned to rooms close to the faculty, which allows the faculty to meet with the students' preferences. Furthermore, the capacity utilization rate of the rooms is maximized. Afterwards, the faculty's current solution and MIP optimal timetables will be compared with 3 quality metrics, objective function score, rooms preferences and rooms utilization rate. As a conclusion, the MIP technique outperformed the current approach of the faculty in terms of quality and response time.