Recently, an open-source light-weight dynamic traffic assignment (DTA) package, namely DTALite, has been developed to allow a rapid utilization of advanced dynamic traffic analysis capabilities. Aiming to bridge the modeling gaps between multi-agent simulation and optimization in a multimodal environment, we further design and develop DTALite-S to simplify the traffic flow dynamic representation details in DTALite for future extensions. We hope to offer a unified modeling framework with inherently consistent space–time network representations for both optimization formulation and simulation process. This paper includes three major modeling components: (1) mathematic formulations to describe traffic and public transportation simulation problem on a space–time network; (2) transportation transition dynamics involving multiple agents in the optimization process; (3) an alternating direction method of multipliers (ADMM)-based modeling structure to link different features between multi-agent simulation and optimization used in transportation. This unified framework can be embedded in a Lagrangian relaxation method and a time-oriented sequential simulation procedure to handle many general applications. We carried out a case study by using this unified framework to simulate the passenger traveling process in Beijing subway network which contains 18 urban rail transit lines, 343 stations, and 52 transfer stations. Via the ADMM-based solution approach, queue lengths at platforms, in-vehicle congestion levels and absolute deviation of travel times are obtained within 1560 seconds.The case study indicate that the open-source DTALite-S integrates simulation and optimization procedure for complex dynamic transportation systems and can efficiently generate comprehensive space-time traveling status.

Computational graphs (CGs) have been widely utilized in numerical analysis and deep learning to represent directed forward networks of data flows between operations. This paper aims to develop an explainable learning framework that can fully integrate three major steps of decision support: Synthesis of diverse traffic data, multilayered traffic demand estimation, and marginal effect analyses for transport policies. Following the big data-driven transportation computational graph (BTCG) framework, which is an emerging framework for explainable neural networks, we map different external traffic measurements collected from household survey data, mobile phone data, floating car data, and sensor networks to multilayered demand variables in a CG. Furthermore, we extend the CG-based framework by mapping different congestion mitigation strategies to CG layers individually or in combination, allowing the marginal effects and potential migration magnitudes of the strategies to be reliably quantified. Using the TensorFlow architecture, we evaluate our framework on the Sioux Falls network and present a large-scale case study based on a subnetwork of Beijing using a data set from the metropolitan planning organization.

Dynamic path generation in complex transportation networks is essential for intelligent transportation systems. Traditional methods, such as shortest path algorithms or heuristic-based models, often fail to capture real-world travel behaviors due to their reliance on simplified assumptions and limited ability to handle long-range dependencies or non-linear patterns. To address these limitations, we propose PathGen-LLM, a large language model (LLM) designed to learn spatial–temporal patterns from historical paths without requiring handcrafted features or graph-specific architectures. Exploiting the structural similarity between path sequences and natural language, PathGen-LLM converts spatiotemporal trajectories into text-formatted token sequences by encoding node IDs and timestamps. This enables the model to learn global dependencies and semantic relationships through self-supervised pretraining. The model integrates a hierarchical Transformer architecture with dynamic constraint decoding, which synchronizes spatial node transitions with temporal timestamps to ensure physically valid paths in large-scale road networks. Experimental results on real-world urban datasets demonstrate that PathGen-LLM outperforms baseline methods, particularly in long-distance path generation. By bridging sequence modeling and complex network analysis, PathGen-LLM offers a novel framework for intelligent transportation systems, highlighting the potential of LLMs to address challenges in large-scale, real-time network tasks.

This paper presents a rapid tree-based reoptimization framework for solving traffic assignment problems in many critical rapid decision making applications. In recent years, regional transportation decision makers and planning organizations are faced with many important decision-making situations with significantly changed origin-destination (O-D) demand patterns, partially due to pattern shifts in land use and emerging transportation modes such as shared bikes and shared vehicles. Thus, there is a critical need for a faster-than-real-time decision-making support system for enabling more informed planning processes. In our approach, we propose a new reoptimization method to recalculate new paths according to baseline traffic assignment results when responding to a new set of traffic demands or supply scenarios. Through smart indexing of previously-calculated traffic assignment outputs, our proposed algorithm can generate new network flow distributions quickly with satisfactory convergence performance. Finally, numerical experiments are performed to demonstrate the adaptive converging behavior and computational efficiency of our tree-based reoptimization algorithm.