Analysing Ancillary Factors that Influence Speeds to Improve Modelling of Bike Paths

July 8, 2026

This thesis report detailed the creation of a model capable of simulating real world cyclists behaviours through the implementation of a power based model in the software SUMO. Having created the model, the objective of the research was to analyse the most influential parameters involved and to see how much the model can be simplified for the application in other contexts. To achieve the objective, the model was ran through a calibration process to match the behaviour of the simulated cyclists to the video captured data from a bidirectional path in Munkbron, Stockholm. This process was further divided into two stages to match the free riding behaviour as well as the interaction behaviour. The results of the elementary effects sensitivity analysis showed that between environmental factors, wind is the major influencer on travel time, time loss and speed while curvature is the biggest influencer on energy consumption. Concerning the remaining parameters, the curvature response was one of the most sensitive amongst the parameters as well as the aerodynamic drag and the rolling friction for travel time and time loss. To then simplify the model, average values for the parameters were compared with using random uniform distributions and the t-tests showed that there is no reason to reject the hypothesis that there is no statistical difference when using the two models. Lastly, homogenising the bicycle type also rendered no statistical difference in the outputs, meaning it is also possible to not diversify the types.

Authors in the community:

• 

  P

  Philip James Kirchin

  ist1117411

Supervisors of this institution:

• 

  R

  Rosa Melo Félix

  ist155593

civil-engineering - Civil engineering

por - Portuguese

Instituto Superior Técnico