Modelos integrados de optimización de transporte público : heurísticas para el problema integrado de calendarización y enrutamiento de pasajeros.

Abstract

In this work we are interested in optimization models for public transport planning. Specifically, two important stages are considered: Timetabling and Passenger Routing. The latter refers to finding the number of passengers using each of the different routes available in the transportation network within a certain time horizon, while the definition of the Timetabling problem consists of determining the departure and arrival times of buses at all stops along a transportation network to meet different goals such as: meeting a given frequency, satisfying demand patterns, maximizing the number of good passenger transfer times, and minimizing timeouts. In this work, the timetabling problem will be studied individually, emphasizing the solution approach given by the PESP (Periodic Event Scheduling Problem), whereas a deeper analysis of the integrated model for Timetabling and Passenger Routing will be carried out by the study of Tim-Pass (Integrated Timetabling and Passenger Routing Problem). Both models are well known in the literature. The core of the present study is the analysis of methods that allow obtaining a solution for the integrated model, which is an integer programming model. To do this, two heuristics known as: LB (Lower Bound) and UB (Upper Bound) will be addressed, focusing attention on the LB to evaluate it with a preprocessing algorithm whose approach is oriented to the choice of a subset of OD pairs (origin- destination).