Receding Horizon Scheduling of On-Demand Urban Air Mobility With Heterogeneous Fleet

Mobility on demand (MoD) is a new paradigm of personal mobility that responds to passengers' demands in real time, and urban air mobility (UAM) is an area of MoD enabled by advances in electric vertical take-off and landing aircraft. This demand-responsive nature of MoD poses a challenge for optimally scheduling vehicles, and it has attracted much attention in recent years. However, there is a lack of research in the MoD scheduling literature: a homogeneous fleet is assumed, but it is not necessarily true all the time. Hence, this article proposes a novel formulation of the scheduling problem for UAM with a heterogeneous fleet and presents particle swarm optimization and a genetic algorithm that utilize a greedy algorithm to keep solutions feasible. The proposed algorithms are implemented with a model-predictive control scheme to effectively manage the demand-responsive nature. As a result, the proposed algorithms can find a near-optimal solution in a short time. Using the algorithms, a numerical experiment with six different fleet mixes is conducted, and impacts of fleet heterogeneity are analyzed. As a result, it is shown that the fleet heterogeneity affects both the quality of service and operational efficiency, and there is a tradeoff: the more vehicles and seats, the better the service, but the less efficient it is.