A new hybrid ant colony optimization algorithm for the vehicle routing problem

This paper presents a novel hybrid ant colony optimization approach called SS_ACO algorithm to solve the vehicle routing problem. The main feature of the hybrid algorithm is to hybridize the solution construction mechanism of the ant colony optimization (ACO) with scatter search (SS). In our hybrid algorithm, we use ACO algorithm and greedy heuristic to generate the initial solutions which are then formed the reference set. Within the scatter search framework, after two-solution combination method for the reference set has been applied, we employ ACO method to generate new solutions through updating the common arc pheromone mechanism. Moreover, during implementing the hybrid algorithm, cyclic transfers, a new class of neighborhood search algorithm can also be embedded into the scatter search framework as neighborhood search to improve solutions. Despite the size of the cyclic transfer neighborhood is very large, a restricted subset of the cyclic transfer neighborhood is adopted to reduce the computational requirements to reasonable levels. Finally, the experimental results have shown that the proposed hybrid method is competitive to solve the vehicle routing problem compared with the best existing methods in terms of solution quality.