The transportation sector accounts for various societal and environmental problems, amongst others, enormous amounts of carbon dioxide emissions. As part of a future mobility solution, on-demand ride-pooling services could contribute significantly to a reduction of the total number of vehicles necessary, while simultaneously improving public transport accessibility. However, as they represent a complex system dependent on multiple parameters, it is hard to predict the overall performance for different potential service areas. The aim of this thesis is to study the dynamics and efficiency of such on-demand ride-pooling systems and analyse the emergent behaviour. In order to reduce complexity we propose the concept of the effective topology which maps the underlying street network and a demand distribution defined on it to a single parameter. We derive two such mappings for binary demand distributions, where all starting and end points are drawn from a finite set of request nodes, and analyse them theoretically as well as with respective simulation series. Most of these simulations are performed with a simplified model consisting of a single vehicle of infinite passenger capacity and a dispatching algorithm which does not allow for detours. Furthermore, we examine the route taken throughout the simulation. To this end we define the measure of the link usage characterising the frequency with which the particular streets are driven. A limit analysis of these link usage patterns shows that, in the case of high request submission rates, the vehicle almost exclusively drives on shortest circuits which contain every request node at least once. Finally, we combine our insights and present an approximation of the used efficiency measure for a given service area in the simplified model. This approximation is based on the observation that the average waiting and driving times converge already for relatively low request submission rates.
Effective topology for on-demand ride-pooling networks
Effektive Topolgien für On-Demand Ride-Pooling Netzwerke
2023
Sonstige
Elektronische Ressource
Englisch
Ride-pooling demand prediction: A spatiotemporal assessment in Germany
Elsevier | 2022
|Pricing and equilibrium in on-demand ride-pooling markets
Elsevier | 2020
|