摘要:AbstractEfficient delivery planning for freight transport is required for reducing environmental load and logistical costs. By using an actual traffic condition, a more realistic delivery plan, increased delivery work efficiency, and CO2 reduction will be possible. An algorithm for quickly obtaining a best solution is required to solve actual vehicle routing problems of large-scale networks and fluctuating delivery costs. The purpose of our research was to solve vehicle routing problems on a large-scale road network that can be applied to actual vehicle routing problems. Our proposed algorithm uses a hierarchical traffic network model composed of two layers according to the frequency of road use. The algorithm is based on tabu search, and the hierarchical traffic network is applied to a generation strategy of the neighborhood solution. Specifically, priority of movement of neighborhood solutions for generating a neighborhood solution is given to delivery points connected in frequently used roads. By using travelling salesman and simple vehicle-routing benchmark problems as a first step in our computing, we confirmed that the proposed algorithm can quickly provide a better solution than the non-application of the proposed algorithm. Furthermore, the effectiveness of the proposed algorithm is discussed by using a virtual vehicle routing problem based on an actual road network.