Aplicação da metaheurística busca em vizinhança variável no fretamento de ônibus intermunicipais.

Abstract

In the vast majority of cities in Brazil, the bus transportation is the only public transportation available to the population. In cities with more inhabitants, the number of vehicles in traffic causes problems such as congestions and environmental pollution. Thus, there is a need to encourage bus transportation. A measure that can be taken in order to increase the use of the bus as preferred transport is the cost reduction of tariffs. As one of the biggest shares in the cost of the salary rates of drivers and conductors, a range of appropriate and effective work could significantly reduce these costs. The problem of scale daily crew consists in assigning the task of driving of vehicles by the crews, so that all trips made by a company to run with the lowest possible cost. The process of scale is a crucial step in the planning of operations for a company in the transportation and belongs to the class of problems difficult to solve, whose function of computational time for the resolution is not polynomial. Numerous methods and techniques have been applied to solve this problem, showing widely varying results. This work makes use of a specific software developed for the generation of inter-scale and, afterwards, apply a metaheuristic called Variable neighborhood search (VNS - Variable Neighborhood Search), to solve the problem of adjustments for charter trips, in order to obtain better solutions than those related to the real problems today. During this work, we developed a computer system by addressing the process of generation of the scale for the regular lines, lines for tourism and for charter. Are presented results of computational experiments with real data from a local transportation company. An exact method was used to generate solutions and the results show a gain in comparison with the solutions obtained by the methodology currently employed in the generation of the scale by the company.