A Constraint-Based Greedy-Local-Global Search for the Warehouse Location Problem

Constraint optimization problems offer a means to obtain a global solution for a given problem. At the same time the promise of finding a global solution often comes at the cost of significant time and computational resources. In contrast, greedy search and local search represent two alternative approaches, which can lead fast to local optima. In this paper, we explore the advantages of incorporating greedy search and local search into constraint optimization methods without forsaking the pursuit of a global solution. The different instances of our global search process are designed to initially behave akin to a greedy search and a local search while they are integrated in a global search approach. This dual strategy aims to achieve two key objectives: firstly, it accelerates the attainment of an initial solution, and secondly, it ensures that this solution possesses a high level of optimality. Even though constraint programming theoretically finds a global optimum, in practice, this may not always be the case due to time and hardware limitations. Our approach improves upon the general Branch-and-Bound approach in constraint programming, aiming to find a good or optimal solution faster than the conventional method. Finally, we validate our findings using the warehouse location problem as a case study.