Bioinspired Optimization Methods and Their Applications

Abstract

In the last decades the Vehicle Routing Problem (VRP) and its ramifications, including the Capacitated Vehicle Routing Problem (CVRP), have attracted the attention of researchers mainly because their presence in many practical situations. Due to the difficulties encountered in their solutions, such problems are usually solved by means of heuristic and metaheuristics algorithms, among which is the Genetic Algorithm (GA). The solution of CVRP using GA requires a solution encoding step, which demands a special care to avoid high computational cost and to ensure population diversity that is essential for the convergence of GA to global optimal or sub-optimal solutions. In this work, we investigated a new binary encoding scheme employed by GA for solving the CVRP. Conducted experiments demonstrated that the proposed binary encoding is able to provide good solutions and is suitable for practical applications that require low computational cost.