Modelling Mixed Traffic Equilibrium
May Chew
Cardno, New South Wales
This presentation was delivered at the 2021 Online Conference Series and until October 2022 is only available to registered delegates and Content Access Pass holders via Interchange. For information on accessing this and other presentations please review the Content Access Pass options.
ABSTRACT
A mixed traffic network consists of multiple user classes. This study considers three types of user behaviours: user equilibrium (UE), system optimum (SO) and Cournot-Nash (CN). With the introduction of connected and autonomous vehicles (CAVs) into the network, the conventional traffic network would gradually shift towards a mixed traffic environment with mixed equilibrium. Nevertheless, it is unlikely that the performance of multi-class networks will improve without a significant number of CAVs on the network. This study aims to investigate the impact of the penetration rate of CAVs on the total system travel time (TSTT) of a multi-class network. In particular, a mixed traffic assignment model was implemented using variational inequalities to simultaneously depict the route choices of three classes of users: non-cooperative human-driven vehicles as UE users, cooperative CAVs which are centrally controlled as SO users and CN users where vehicles belonging to a common operator (e.g. ride-hailing company) are fully cooperative but non-cooperative otherwise.
The Frank-Wolfe algorithm was applied with Gauss-Seidel methods to solve the multi-class traffic assignment problem. The model was applied to a toy network, the Sioux Falls Network, and a Sydney arterial sub-network, respectively. From simulations, it was observed that a 10% CAVs penetration rate on the Sydney subnetwork can realize more than 50% of the total TSTT reduction possible. The scheme is also more effective for congested network. The behaviour of the UE, CN and SO player in a mixed equilibrium environment was also shown. This study lays the groundwork for future research and will be relevant in the consideration of future national policies in regulating the number of private human-driven vehicles and CAVs fleet on mixed traffic networks.
Author
May Chew | Cardno
May Chew has completed her undergraduate thesis under the iMOVE CRC undergraduate program and has a keen interest in future mobility. As an undergraduate, she had the opportunity to be part of Smart Innovation Centre, NSW’s hub for collaboration and development of safe and efficient emerging transport technology. She has recently completed her degree in Bachelors of Civil Engineering (Honours)/ Engineering Science in Environmental Engineering and is currently a graduate at Cardno in the Transport Planning team.
