How passenger travel mode choice dynamics shape nighttime public transport service planning? A game-theoretic perspective

Understanding passenger behavioral dynamics is essential for designing efficient and demand-responsive nighttime public transport (NPT) services. This study develops a multi-population evolutionary game framework to analyze heterogeneous passengers’ NPT mode choice and its underlying mechanisms. Replicator dynamics are employed to derive analytical threshold conditions governing behavioral evolution. The insights are incorporated into an operational optimization model that endogenizes passenger responses in service planning. A Beijing case study shows that, under dual-equilibrium conditions, outcomes are influenced by initial preferences, but strong cross-population interactions can weaken path dependence, highlighting the potential for early-stage interventions. Moreover, marginal deviation costs regulate coordination: higher costs eliminate discoordinated choices and enable dual coordinated equilibria, while lower costs sustain heterogeneous mode choices. Behavioral heterogeneity from the value of time and travel distance supports differentiated services, such as long-distance express metro and short-distance feeder buses. Fare and headway exhibit behavioral tolerance thresholds: once passenger acceptance is exceeded, mode choices can shift abruptly, leading to inefficient system performance. Importantly, incorporating endogenous behavioral dynamics into planning reduces total travel costs by 52.96% relative to a behavior-agnostic model. This study underscores the necessity of incorporating behavioral dynamics into NPT planning to support efficient, adaptive, and sustainable services.