A Markovian Traffic Equilibrium Model for Ride-Hailing

A team of researchers including Song Gao, Hanyu Cheng, Chiwei Yan, and Guocheng Jiang has introduced a new Markovian traffic equilibrium model specifically designed for ride-hailing systems. Published on arXiv, the model treats each vehicle—whether empty or hired—as an agent making sequential decisions on order acceptance and route choice. The goal is to maximize total discounted return over an infinite horizon, framed as a semi-Markov decision process. This approach endogenizes both the competition among empty vehicles for passenger demand and the traffic congestion that arises from road usage at the link level.

The researchers characterize the equilibrium as the solution to a fixed-point system and prove its existence. They develop relaxed fixed-point iteration algorithms for computing the equilibrium, with convergence guarantees for specialized network structures. Computational experiments on realistic networks demonstrate the model's practical value for transportation planning. Ablation analyses reveal that ignoring either traffic congestion or drivers' forward-looking behavior can lead to potentially substantial biases in policy evaluation, highlighting the model's importance for accurate simulation and decision-making in ride-hailing markets.