Impact of automated vehicles on traffic assignment, mode split, and parking behavior

Implementing autonomous vehicles (AVs) can provide plenty of advantages, such as improving network capacity and reducing crashes. If morning commuters take AVs to the workplace, they can assign AVs an induced trip to a low-cost parking lot or even park AVs at their origins with no cost. This travel behavior might result in a more congested city network. This study firstly considers the equilibriums in route choice, mode choice, and parking lot choice in one optimization model. The solution algorithm is proposed to solve the optimization program iteratively. The Sioux Falls network is used to test the potential impacts of induced AV trips and the preference to choose an AV. The results using the test network find that scenarios with AVs increase the average travel time by approximately 50%, while 14.60% to 32.27% (scenario-based) of the area that is currently used as parking spaces can be re-purposed.