logo SBA

ETD

Archivio digitale delle tesi discusse presso l’Università di Pisa

Tesi etd-09192013-110106


Tipo di tesi
Tesi di laurea magistrale
Autore
INCAINI, RICCARDO
Indirizzo email
r.incaini@gmail.com
URN
etd-09192013-110106
Titolo
Routing Policies for Vehilcle-Sharing Systems
Dipartimento
INGEGNERIA DELL'INFORMAZIONE
Corso di studi
INGEGNERIA ROBOTICA E DELL'AUTOMAZIONE
Relatori
relatore Pallottino, Lucia
relatore Frazzoli, Emilio
Parole chiave
  • Balancing Policies
  • Mobility-on-Demand
  • Optimization
  • Routing Policies
  • Vehicle-Sharing Systems
Data inizio appello
04/10/2013
Consultabilità
Completa
Riassunto
The high cost of private vehicle ownership compared to its low utilization rates, and the increase in parking requirements compared to the decrease in available urban land, make private automobiles an unsustainable solution for the future of dense urban environments.

One of the leading emerging paradigms for future urban mobility systems is vehicle-sharing, which effectively merges private and public mobility, and directly tackles the problems of parking spaces and current low vehicle utilization rates.

In this thesis we consider one-way vehicle-sharing systems that are composed of a finite group of shared vehicles that are located at a set of stations. The user arrives at one station, picks up a vehicle, and drives to his/her destination station where he/she drops the vehicle off. The particularity of one-way systems is that a user may drop the vehicle off at any station in the network, i.e., there is no need to return it to the same station from which it was rented. Hence, when some origin and destination stations are more popular than others, the system will inevitably become out of balance: vehicles will build up at some stations and become depleted at others.

In order to solve this problem we propose different routing policies that allow the vehicle-sharing company to operate efficiently by better managing the distribution of the vehicles among the network stations by taking into account customer demands.

Firstly, we describe a trip price selection mechanism to control the distribution of vehicles among stations. The novelty of our balancing approach is the use of user mobility demand preferences, i.e., the price customers are willing to pay for selected trips, to control the flow of vehicles between stations. Through network-oriented arguments, we provide a pricing-based rental policy that ensures the system operating efficiently in the long term, without employing human drivers or using autonomous empty vehicles.

Secondly, exploiting the idea of using autonomous empty robotic vehicles to maintain the balance of the system, i.e., a uniform distribution of vehicles among the network stations, we propose a model of a one-way electric-vehicle-sharing system taking into account that vehicles need to be charged when their battery level is low. Therefore, the battery constraints are such that even if there is a vehicle at one station it might be unavailable to a customer because its residual energy is not enough to reach the desired destination.

Considering different scenarios we show that, without control policies, the throughput of one-way vehicle-sharing systems plummets. On the other hand, simulations attest the functionality of the reported policies and illustrate their ability to subvert limitations that can plague one-way sharing systems. From the pricing-based routing policy we determine the optimal prices of each route that maximize the long term network throughput while ensuring the stability of the system. From the robotic solution of the balancing problem we prove that (i) there exists a minimum number of vehicles that stabilizes the system, i.e., the waiting number of customers is bounded;
(ii) the minimum number of vehicles that stabilizes the system is a function of the charger and discharger battery functions, and the network topology; (iii) the minimum number of vehicles that stabilizes the system is greater than, or equal to, the one obtained without considering the battery constraints.
File