Bio-inspired optimal site selection of LPG stations for gas-driven cars in an urban region

Abstract

Siting new liquefied petroleum gas (LPG) stations in a metropolitan region to support the initiative of air pollution control is intimately tied to where these LPG stations are located for filling up those vehicles. Unlike traditional siting and routing problems, the LPG stations must be sited to serve the consumers with the highest mobility instead of meeting the regular consumers’ demand. To optimize the spatial allocation of LPG stations in an intelligent transportation network with consideration of the economies of scale, this study presents a bio-inspired computational intelligence algorithm based on the gravity model to express the behavior of drivers in search of a set of neighboring LPG stations. The practical implementation of the proposed method was evaluated using a real-world case study, in which the planning objective is to minimize the expected distance for fueling vehicles. A Monte Carlo simulation is performed to determine the effect of the uncertainty that is caused by the mobility of LPG cars and parameters in the gravity model on the optimal results. The results reveal that the optimal number of LGP stations in the case study is approximately 15, and the minimized expected distance for fueling is 2500 m. The proposed method is useful when the behavior of customers cannot be ignored in siting problems.