Solving Traveling Salesman Problem based on Biogeography-based Optimization and Edge Assembly Cross-over

The Biogeography-Based Optimization (BBO) algorithm has recently been of great interest to the researchers for its simplicity of implementation, efficiency, and low number of parameters. The BBO algorithm in optimization problems is one of the new algorithms that have been developed based on the biogeography concept. This algorithm uses the idea of animal migration to find suitable habitats for solving the optimization problems. The BBO algorithm has three principal operators called migration, mutation, and elite selection. The migration operator plays a very important role in sharing information among the candidate habitats. The original BBO algorithm, due to its poor exploration and exploitation, sometimes does not perform desirable results. On the other hand, the Edge Assembly Cross-over (EAX) has been one of the high powers cross-overs for acquiring off-spring, and it increases the diversity of the population. A combination of the BBO algorithm and EAX can provide a high efficiency in solving the optimization problems including the traveling salesman problem (TSP). In this paper, we propose a combination of those approaches to solve the traveling salesman problem. The new hybrid approach is examined with standard datasets for TSP in TSPLIB. In the experiments, the performance of the proposed approach is better than the original BBO and four others widely used metaheuristics algorithms.