School timetabling problem under disturbances

We present methods to solve the school timetabling problem under disturbances.A multi-objective approach is used to obtain stable schedules.We test the approaches with data of 5 different schools in The Netherlands.The methods presented avoid designing schedules from scratch as a disturbance occurs.The methods generate robust solutions of high quality for different types of schools. School timetables are one or multiple times per year generated to assign class-teacher combinations to class rooms and timeslots. Post-publication disturbances such as absence of teachers typically pose a need for schedulers to rapidly implement some minor changes to avoid empty periods in the timetable. In this paper our aim is to define methods to efficiently solve the school timetabling problem under disturbances. We present three types of solution methods, namely a simple rule-of-thumb, a heuristic and an optimization approach. Exhaustive numerical experiments have been performed with data from five high schools in The Netherlands, each with their unique characteristics in number of classes, number of teachers and number of daily meetings. For each of the three methods we show advantages and disadvantages as well as the effects of resulting changes in the schedules.

[1]  Edmund K. Burke,et al.  A branch-and-cut procedure for the Udine Course Timetabling problem , 2012, Ann. Oper. Res..

[2]  Nelson Maculan,et al.  Strong bounds with cut and column generation for class-teacher timetabling , 2012, Ann. Oper. Res..

[3]  Dominique de Werra,et al.  A generalized class-teacher model for some timetabling problems , 2002, Eur. J. Oper. Res..

[4]  A. Tripathy School Timetabling---A Case in Large Binary Integer Linear Programming , 1984 .

[5]  Nelishia Pillay,et al.  A survey of school timetabling research , 2014, Ann. Oper. Res..

[6]  Alon Itai,et al.  On the Complexity of Timetable and Multicommodity Flow Problems , 1976, SIAM J. Comput..

[7]  Mohammed Azmi Al-Betar,et al.  A harmony search algorithm for university course timetabling , 2010, Annals of Operations Research.

[8]  D. de Werra,et al.  Complexity of some special types of timetabling problems Journal of Scheduling , 2002 .

[9]  R. Alvarez-Valdes,et al.  Constructing Good Solutions for the Spanish School Timetabling Problem , 1996 .

[10]  Stephen C. H. Leung,et al.  A simulated annealing with a new neighborhood structure based algorithm for high school timetabling problems , 2010, Eur. J. Oper. Res..

[11]  Jeffrey H. Kingston,et al.  An XML format for benchmarks in High School Timetabling , 2010, Ann. Oper. Res..

[12]  George H. G. Fonseca,et al.  Variable Neighborhood Search based algorithms for high school timetabling , 2014, Comput. Oper. Res..

[13]  D. de Werra,et al.  An introduction to timetabling , 1985 .

[14]  D. Costa,et al.  A tabu search algorithm for computing an operational timetable , 1994 .

[15]  Efthymios Housos,et al.  A column generation approach for the timetabling problem of Greek high schools , 2003, J. Oper. Res. Soc..

[16]  Elliott N. Weiss,et al.  Class Scheduling to Maximize Participant Satisfaction , 1995 .

[17]  Elliott N. Weiss,et al.  Increasing service levels in conference and educational scheduling: a heuristic approach , 1995 .

[18]  Hanif D. Sherali,et al.  A mixed-integer programming approach to a class timetabling problem: A case study with gender policies and traffic considerations , 2007, Eur. J. Oper. Res..

[19]  Gilbert Laporte,et al.  Recent Developments in Practical Course Timetabling , 1997, PATAT.