Abstract | Keywords | Export | Availability | Bookmark

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Travel demand modelling is an essential process to facilitate the decisions of transportation planners. This is typically performed through either the traditional four-step model or activity-based models. However, both of these models require the acquisition of household travel surveys, which can be costly to implement, and other socioeconomic data. Furthermore, trip-based models in particular rely on aggregate data, which has several associated shortcomings. This master’s thesis proposes an alternative methodology for determining travel demand within a region. This framework relies on open-source data, in particular from OpenStreetMap.com, and is therefore far less costly to implement. The data also allows for prior assumptions to be updated or checked, when information pertinent to those assumptions is available. To achieve this, a system of extracting useful points of interest from the plethora of data available on OpenStreetMap is proposed. Afterwards, based on data from key-figures, a method for producing a synthetic population of trips through rejection sampling is proposed. Both the benefits and disadvantages of this model over the traditional four-step model and activity-based models are discussed. An attempt to implement the model for the region of Leuven, Belgium is then performed. Firstly, a layer of points of interest is extracted using the proposed methodology, utilising QGIS as the software of choice. Afterwards, production and attraction factors are estimated based on key figures for Flanders, and a crude approximation of production and attraction is performed. Errors in both the POI layer extraction, and production and attraction estimates are discussed. At the time of writing, the goal of creating a sample population which can be compared to existing data has not come to fruition.