نوع مقاله : علمی - پژوهشی
نویسنده
استادیار، گروه مهندسی عمران، دانشکده مهندسی، دانشگاه بوعلیسینا، همدان، ایران
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسنده [English]
This paper proposes an integer linear mathematical formulation for Vehicle Routing Problem (VRP), where the capital cost for deploying each vehicle is minimized together with other on-link transportation costs. The model has been formulated as a multi-commodity network flow model with capacity constraints. It is well known that the computational complexity to this type of problems is NP-hard. Thus, the ACO algorithm, which has been known to be a powerful meta-heuristic algorithm for solving VRPs in large networks, has been adapted to solve the problem. Although the ACO algorithm has repeatedly been used to solve the capacitated VRP, it has a drawback that cannot consider the capital cost of each vehicle along with other operational costs of the vehicles (associated with the total distance traveled within a day) in its initial form. More specifically, naturally it assumes that each vehicle returns to the depot if it becomes full or the demand finishes, each met first; this paper seeks to propose an adapted ACO algorithm in which this assumption is released. To assess the capability of the proposed model in large-scale networks, the case study of Mashhad city, consisting of 253 traffic analysis zones and over than 3800 links, has been considered. Results show that the proposed algorithm converges to near-to-optimal solutions within two seconds of cpu time, which is encouraging.
کلیدواژهها [English]