%0 Conference Proceedings
%4 sid.inpe.br/mtc-m16c/2015/12.10.17.17
%2 sid.inpe.br/mtc-m16c/2015/12.10.17.17.21
%@issn 2179-4820
%T Optimization of taxi cabs assignment in Geographical Location-based Systems
%D 2015
%A Oliveira, Abilio A. M. de,
%A Souza, Matheus P.,
%A Pereira, Marconi de A.,
%A Reis, Felipe A. L.,
%A Almeida, Paulo E. M.,
%A Silva, Eder J.,
%A Crepalde, Daniel S.,
%@affiliation Universidade Federal de São João Del-Rei (UFSJ)
%@affiliation Universidade Federal de São João Del-Rei (UFSJ)
%@affiliation Universidade Federal de São João Del-Rei (UFSJ)
%@affiliation Centro Federal de Educação Tecnológica de MG (CEFET-MG)
%@affiliation Centro Federal de Educação Tecnológica de MG (CEFET-MG)
%@affiliation Universidade Federal de São João Del-Rei (UFSJ)
%@affiliation Universidade Federal de São João Del-Rei (UFSJ)
%E Fileto, Renato,
%E Korting, Thales Sehn,
%B Simpósio Brasileiro de Geoinformática, 16 (GEOINFO)
%C Campos do Jordão
%8 27 nov. a 02 dez. 2015
%I Instituto Nacional de Pesquisas Espaciais (INPE)
%J São José dos Campos
%P 92-104
%S Anais
%X In this paper, different approaches are evaluated to assign taxi cabs to customers in geographical location-based systems. The main purpose of this work is to identify the solution in which all current customers are met in an acceptable time, however minimizing the distance traveled by existing free taxi cabs. Two aspects are considered: 1) the method to calculate the distance between vehicles and customers; and 2) a vehicle assignment strategy. The methods to calculate the distance between vehicles and customers are: a GPSbased routing (a shortest path algorithm) and the Euclidean distance. On the other hand, as vehicle assignment approaches, the considered strategies are: a greedy algorithm, which assigns each vehicle to the closest customer, and an optimization algorithm, which assigns vehicles considering the whole scenario, minimizing the global distance traveled by taxi cabs to meet the customers. This last strategy considers an optimization model in such a way that the calls are not readily answered. In this case, a short waiting window is implemented, where the calls are stored and then the optimization algorithm is executed, in order to minimize the required distance and to meet all current customers. The combination of the two methods of distance calculation and the two vehicle assignment strategies formed four possible approaches, which are evaluated in a realistic simulator. Results show that the approach which uses the shortest path algorithm and an optimization algorithm reduces the average service time by up to 27.59%, and the average distance traveled by up to 45.79%.
%@language en
%3 proceedings2015_p8.pdf