Hybrid algorithm of Tabu Search and Integer Programming for the railway crew scheduling problem

摘要

The crew scheduling problem (CSP) is an important problem for the transportation enterprises, especially when crew's cost is increasing. This justifies the researcher's effort in order to create algorithms which obtain good solutions for this kind of problems. Particularly, hybrid algorithms appear as an efficient alternative in order to solve any combinatorial optimization problem. Tabu Search (TS) is a simple metaheuristic which has been applied in many combinatorial problems, obtaining very competitive solutions. Integer Programming (IP), in this case, allows to the TS algorithm to have a good level of diversification, avoiding that it's converge to local optimal solutions. In this article we presents an hybrid algorithm of Tabu Search and Integer Programming which is applied to the both CSP instances obtained from the literature and a real application case in Chile (MERVAL). The hybrid algorithm obtains good solutions in both cases, and obtains an important minimization in the execution time in the MERVAL case.