On the flexibility of constraint programming models: From single to multiple time windows for the traveling salesman problem

One of the major strengths of Constraint Programming is the flexibility and expressiveness of models in that computational paradigm, which make it easy to add problem-dependent constraints without having to modify the solution strategy. We show here what needs to be done in order to adapt a constraint programming algorithm for the traveling salesman problem with time windows so that it can handle multiple time windows. Computational results are also presented on a set of instances created for that little-studied problem.