University of Twente Student Theses
Impacts of AVs on the Capacity of a Dutch Roundabout
Geerlings, K.H. (2024) Impacts of AVs on the Capacity of a Dutch Roundabout.
|
PDF
7MB |
| Abstract: | The main goal of this report is to research the impacts of autonomous vehicles (AVs) on the capacity of a Dutch roundabout. This was done by answering the sub-questions; How can driving behaviour of AVs translate over to the micro-simulation model in Vissim; How to gather data from the roundabout model in Vissim to determine the capacity; What is the roundabout capacity with different AV types and AV penetration rate scenarios; What are the effects of pedestrian and cyclist crossings on the roundabout capacity. Autonomous vehicles (AVs) labelled as level 4 and 5 by levels of driving automation have the possibility to change the current traffic infrastructure landscape. They can make transportation more efficient, safer and accessible for everyone. The Dutch Ministry of Infrastructure has become aware and is now adjusting regulations to make testing of AVs possible in the current traffic network. The effects of AVs need to be modelled and researched with modelling tools in advance for proper testing. Literature review shows that the impact of AVs on the entry capacity of the roundabout depends on many variables. The geometrical aspects are important for currently existing linear regression models that use relationships between one dependent variable and multiple independent variables to estimate the capacity of roundabouts. Linear regression models are based on empirical data, and since AVs are not existing in the current traffic environment it is not possible to use this method. Another method is gap acceptance modelling. It uses the critical gap, follow-on headway and circulating traffic parameters, such models are based on assumptions, variables and distributions based on the current traffic network. How AVs will perform on roundabouts is still unclear and therefore a different method was used. Vissim is a microsimulation model software package that can model any traffic situation and change almost all variables and their distributions, as well as geometrical elements of road design elements. For this research the geometrical elements have been based on Goudappel and CROW standards. The capacity of a roundabout is hard to determine due to the many variables at play. Therefore, a smaller approach was taken where only one entry leg was observed. Meaning that the term capacity for this report means entry capacity of the roundabout. Vehicle behaviour is dependent on many variables. Therefore, three different AV types have been considered, cautious, normal and aggressive AVs. These behaviour types are based on how aggressive they are in their behaviour as their name implies. Cautious AVs drive more cautiously than conventional vehicles (CVs), while normal AVs are equivalent to CVs, and aggressive AVs drive more aggressive than CVs. AVs are expected to have the same behaviour variables as CVs, but without some of the human error-based parameters. The AV behaviour types can be translated to Vissim using the car following model Wiedemann 99 parameters. Next to car following, the roundabout entry behaviour is also important. This is mostly based on time gaps and minimum clearances, which differ for every vehicle type. To find the influence of these different AV types, multiple scenarios were created and simulated. The scenarios were based on three different ratios regarding entry and opposing traffic flows, five AV penetration rates from 20% to 100% in increments of 20%, and four additional scenarios based on the presence of pedestrians and cyclists. Cautious AVs caused drops in capacity, while normal AVs were comparable to CVs. Aggressive AVs caused the highest increases in capacity as expected. The ratio of 25/75 had the most varying results, as cautious vehicles caused a decrease in capacity of 20%. When pedestrians and cyclists were involved, a further drop was observed to a capacity decrease of 26%. While Aggressive AVs had an increase of 21% at 100% penetration rate. Overall, the increase in the level of aggressiveness, as well as penetration rate and lack of pedestrians and cyclists led to an increase in capacity. |
| Item Type: | Essay (Bachelor) |
| Faculty: | ET: Engineering Technology |
| Programme: | Civil Engineering BSc (56952) |
| Link to this item: | https://purl.utwente.nl/essays/103855 |
| 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