Leizhen Wang

Leizhen Wang, Peibo Duan, Cheng Lyu et al.

Modern navigation systems and shared mobility platforms increasingly rely on personalized route recommendations to improve individual travel experience and operational efficiency. However, a key question remains: can such sequential, personalized routing decisions collectively lead to system-optimal (SO) traffic assignment? This paper addresses this question by proposing a learning-based framework that reformulates the static SO traffic assignment problem as a single-agent deep reinforcement learning (RL) task. A central agent sequentially recommends routes to travelers as origin-destination (OD) demands arrive, to minimize total system travel time. To enhance learning efficiency and solution quality, we develop an MSA-guided deep Q-learning algorithm that integrates the iterative structure of traditional traffic assignment methods into the RL training process. The proposed approach is evaluated on both the Braess and Ortuzar-Willumsen (OW) networks. Results show that the RL agent converges to the theoretical SO solution in the Braess network and achieves only a 0.35% deviation in the OW network. Further ablation studies demonstrate that the route action set's design significantly impacts convergence speed and final performance, with SO-informed route sets leading to faster learning and better outcomes. This work provides a theoretically grounded and practically relevant approach to bridging individual routing behavior with system-level efficiency through learning-based sequential assignment.

Yuanyuan Wu, Zhenlin Qin, Leizhen Wang et al.

Understanding and modeling individual travel behavior responses is crucial for urban mobility regulation and policy evaluation. The Markov decision process (MDP) provides a structured framework for dynamic travel behavior modeling at the individual level. However, solving an MDP in this context is highly data-intensive and faces challenges of data quantity, spatial-temporal coverage, and situational diversity. To address these, we propose a group-effect-enhanced generative adversarial imitation learning (gcGAIL) model that improves the individual behavior modeling efficiency by leveraging shared behavioral patterns among passenger groups. We validate the gcGAIL model using a public transport fare-discount case study and compare against state-of-the-art benchmarks, including adversarial inverse reinforcement learning (AIRL), baseline GAIL, and conditional GAIL. Experimental results demonstrate that gcGAIL outperforms these methods in learning individual travel behavior responses to incentives over time in terms of accuracy, generalization, and pattern demonstration efficiency. Notably, gcGAIL is robust to spatial variation, data sparsity, and behavioral diversity, maintaining strong performance even with partial expert demonstrations and underrepresented passenger groups. The gcGAIL model predicts the individual behavior response at any time, providing the basis for personalized incentives to induce sustainable behavior changes (better timing of incentive injections).

Leizhen Wang, Peibo Duan, Hao Wang et al.

In traffic engineering, the fixed-time traffic signal control remains widely used for its low cost, stability, and interpretability. However, its design depends on hand-crafted formulas (e.g., Webster) and manual re-timing by engineers to adapt to demand changes, which is labor-intensive and often yields suboptimal results under heterogeneous or congested conditions. This paper introduces the EvolveSignal, a large language models (LLMs) powered coding agent to automatically discover new traffic signal control algorithms. We formulate the problem as program synthesis, where candidate algorithms are represented as Python functions with fixed input-output structures, and iteratively optimized through external evaluations (e.g., a traffic simulator) and evolutionary search. Experiments on a signalized intersection demonstrate that the discovered algorithms outperform Webster's baseline, reducing average delay by 20.1% and average stops by 47.1%. Beyond performance, ablation and incremental analyses reveal that EvolveSignal modifications-such as adjusting cycle length bounds, incorporating right-turn demand, and rescaling green allocations-can offer practically meaningful insights for traffic engineers. This work opens a new research direction by leveraging AI for algorithm design in traffic signal control, bridging program synthesis with transportation engineering.