Urban bicycles renting systems: Modelling and optimization using nature-inspired search methods

详细信息    查看全文

• 作者：Camelia Chira^a ; ^{camelia.chira@itcl.es" class="auth_mail}Author Vitae ; Javier Sedano^aAuthor Vitae ; José ; R. Villar^bAuthor Vitae ; Mó ; nica Cá ; mara^aAuthor Vitae ; Emilio Corchado^cAuthor Vitae
• 关键词：Urban bicycles renting systems ; Vehicle routing problems ; Evolutionary algorithms ; Ant colony systems
• 刊名：Neurocomputing
• 出版年：5 July, 2014
• 年：2014
• 卷：135
• 期：Complete
• 页码：98-106
• 全文大小：984 K

文摘

Urban Bicycles Renting Systems (UBRS) are becoming a common and useful component in growing modern cities. For an efficient management and support, the UBRS infrastructure requires the optimation of vehicle routes connecting several bicycle base stations and storage centers. In this study, we model this real-world optimization problem as a capacitated Vehicle Routing Problem (VRP) with multiple depots and the simultaneous need for pickup and delivery at each base station location. Based on the VRP model specification, two nature-inspired computational techniques, evolutionary algorithms and ant colony systems, are presented and their performance in tackling the UBRS problem is investigated. In the evolutionary approach, individuals are encoded as permutations of base stations and then translated to a set of routes subject to the constraints related to vehicle capacity and node demands. In the ant-based approach, ants build complete solutions formed of several subtours servicing a subset of base stations using a single vehicle based on both apriori (the attractiveness of a move based on the known distance or other factors) and aposteriori (pheromone levels accumulated on visited edges) knowledge. Both algorithms are engaged for the UBRS problem using real data from the cities of Barcelona and Valencia. Computational experiments for several scenarios support a good performance of both population-based search methods. Comparative results indicate that better solutions are obtained on the average by the ant colony system approach for both considered cities.