A restricted path-based ridesharing user equilibrium

In this paper, a restricted path-based ridesharing user equilibrium (RUE) is proposed to build up the rationales between ridesharing activities and traffic congestion. In the traffic assignment problem with ridesharing, travelers simultaneously choose routes from origins to destinations and travel modes (including solo driver, ridesharing driver, and ridesharing passenger). The proposed RUE model with nonadditive path costs is more realistic than the existing ones based on some additional model specifications: ridesharing drivers bear an extra cost for taking riders and meanwhile receive an extra subsidy, and each rider is carried by only one driver. The Braess paradox is examined within the context of the proposed RUE model. The numerical results on the Braess network with and without high-occupancy toll (HOT) lane show that (a) the existence of the HOT lane will facilitate more travelers’ sharing rides, reduce the number of total vehicles (drivers), and improve the total vehicle travel time (TVTT); (b) the toll charge of solo drivers on HOT lane will promote the ridesharing activities while this effect is conditional on the levels of the extra cost and subsidy. All numerical results on the Braess network and the grid network show that more travelers will participate in ridesharing when the amount of extra subsidy for ridesharing drivers increases.