University of Twente Student Theses
As of Friday, 8 August 2025, the current Student Theses repository is no longer available for thesis uploads. A new Student Theses repository will be available starting Friday, 15 August 2025.
Predicting the volume of cyclists at road segments based on environmental characteristics
Witjes, F.L.D. (2025) Predicting the volume of cyclists at road segments based on environmental characteristics.
|
PDF
5MB |
| Abstract: | In the Netherlands the number is cyclists is significant and expected to grow even further in the coming years. This makes it important for municipalities and other policy makers to have an insight cyclist intensities, as they can use this information when new developments aƯecting cyclists are made. Currently, models that estimate cyclist intensities often use the four step model as their core approach, which estimates the shortest route between the origin and destination of a trip. In research it was found that many trips do not follow the shortest route, but are instead influenced by environmental factors to take a diƯerent route (De Jong et al., 2023). Models that follow a spatial planning approach, which are often used to estimate pedestrian intensities use only environmental characteristics to predict and could be a solution to this problem. An example of a model that functions as such is the Loopmonitor of Witteveen+Bos, which predicts pedestrian intensities. In this research it was evaluated if a model that only took into account environmental characteristics, like the Loopmonitor, can predict cyclist intensities on a case study of Apeldoorn, the Netherlands. Based on a literature study relevant characteristics that aƯect cyclists were determined, whilst also characteristics that are used in spatial planning analyses were found. These characteristics were used in a regression model, where Random Forest regression was found to be the most suitable regression type. The findings from the regression model reveal that network characteristics were the most influential, whilst infrastructural characteristics had very little eƯect on the model results. In the validation of the results a Mean Absolute Error (MAE) of 436 cyclists on a daily basis was found, with the Root Mean Square Error (RMSE) being 681. These errors were more significant for locations with many cyclists. In a comparison with a four step model, the Fietsmonitor of Witteveen+Bos it was found that the constructed model has a greater predictive capability on Apeldoorn. From this research it is furthermore recommended to incorporate network analysis more into research related to predicting cyclist intensities, as these were determined to be very impactful in this study. Lastly, it is recommended to compare the model and the Fietsmonitor again on a diƯerent city, so that an independent comparison can also be made about the accuracies of both models. |
| Item Type: | Essay (Bachelor) |
| Faculty: | ET: Engineering Technology |
| Programme: | Civil Engineering BSc (56952) |
| Link to this item: | https://purl.utwente.nl/essays/107163 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page