A Meta-Heuristic Approach for The Constraint Satisfaction Problem

Abstract

This paper studies one of the constraint satisfaction problems (CSP), the university course timetabling problem (CTP). Given the ranges of each variable and a series of constraints, the objective of a CSP is to find a feasible assignment for each variable satisfying all the constraints. By analyzing the course schedule history from one university, we present the mathematical formulation of the CTP and introduce a meta-heuristic approach to solve this challenging problem. Apart from the neighborhood structure for this specific problem, a corresponding fast-incremental evaluation method is also proposed. Experimental results show the high efficiency of the proposed algorithm.