Optimisation Of Car Fleet Exploitation Using Statistical And Fuzzy Logic Approaches
Free (open access)
J. M. Boussier, L. Ion, D. Breuil & S. Benhabib
LISELEC is a car sharing system implemented in La Rochelle in 1999. Its goal is to provide the city with a clean transport means offering a mode combining the individual use and the public ownership. This paper presents our contribution to develop a decision aid tool with a double objective: to optimize its exploitation and to anticipate extension (increase of the number of stations or cars, change of vehicle type) for improving the system efficiency. The first step of this project was a detailed analysis of the data exploitation for acquiring knowledge of components behaviour. In this aim, Principal Component Analysis was used to establish correlations between different vehicles flows, departure hour and arrival hour. Via concepts like accessibility and attractiveness of an urban area, a fuzzy logic approach was used in order to model the flows between the Liselec stations. Keywords: urban station, fuzzy logic, principal component analysis. 1 Introduction The continuous increase of the number of vehicles has caused a tremendous growth of the traffic flows on public roads, especially in urban areas; saturated car parks, congested roads, environmental effects explain that the local authorities have look for alternative mobility solutions in order to satisfy users and environmental strategies. Modelling and simulation techniques are frequently used to understand the traffic problems at different levels of abstraction and to help in easy understanding for various managers who have different knowledge of the
urban station, fuzzy logic, principal component analysis.