Content area
Abstract
The objective of this research is to solve a scheduling problem using memetic algorithms which solve problems with a large number of variables and constraints. This method can be applied to different scheduling problems in which they must share resources, such as scheduling clients for certain services, resource allocation, etc. The problem to be solved in this research is the assignment of the best possible schedule for professors of the Department of Mathematical Sciences, at University of Puerto Rico at Mayaguez, which offers over 180 sections in 13 regular classrooms, two labs and two amphitheatres . It is a difficult process scheduling courses classes at set times due to physical constraints, variation in the number of course credits, fulfill orders of teachers, among others.
The problem can be stated using integer linear programming that is closely related to an assignment problem, but its real-time solution to the problem of assigning professors to classes is computationally infeasible because the variables considered in the problem are binary. It has been shown that the problem of optimizing a scheduling problem is NP-hard, as an alternative is to develop heuristics as an alternative to the solution of this problem whose goal is to determine the best schedule of classes that satisfy as much of preferences of professors, including the satisfaction of requested classes, block scheduling and rooms to obtain the best solution in a reasonable time.
Memetic algorithms that were used are heuristics that solve the scheduling problem in a reasonable time and generate the best possible schedule that satisfies the requests of teachers in terms of classes, time and lounges.
