Predicting vehicle trajectories is crucial to ensuring automated vehicle operation efficiency and safety, particularly on congested multi-lane highways. In such dynamic environments, a vehicle's motion is determined by its historical behaviors as well as interactions with surrounding vehicles. These intricate interactions arise from unpredictable motion patterns, leading to a wide range of driving behaviors that warrant in-depth investigation. This study presents the Graph-based Interaction-aware Multi-modal Trajectory Prediction (GIMTP) framework, designed to probabilistically predict future vehicle trajectories by effectively capturing these interactions. Within this framework, vehicles' motions are conceptualized as nodes in a time-varying graph, and the traffic interactions are represented by a dynamic adjacency matrix. To holistically capture both spatial and temporal dependencies embedded in this dynamic adjacency matrix, the methodology incorporates the Diffusion Graph Convolutional Network (DGCN), thereby providing a graph embedding of both historical states and future states. Furthermore, we employ a driving intention-specific feature fusion, enabling the adaptive integration of historical and future embeddings for enhanced intention recognition and trajectory prediction. This model gives two-dimensional predictions for each mode of longitudinal and lateral driving behaviors and offers probabilistic future paths with corresponding probabilities, addressing the challenges of complex vehicle interactions and multi-modality of driving behaviors. Validation using real-world trajectory datasets demonstrates the efficiency and potential.

Transportation Cyber-Physical Systems (T-CPS) enhance safety and mobility by integrating cyber and physical transportation systems. A key component of T-CPS is the Digital Twin (DT), a virtual representation that enables simulation, analysis, and optimization through real-time data exchange and communication. Although existing studies have explored DTs for vehicles, communications, pedestrians, and traffic, real-world validations and implementations of DTs that encompass infrastructure, vehicles, signals, and communications remain limited due to several challenges. These include accessing real-world connected infrastructure, integrating heterogeneous, multi-sourced data, ensuring real-time data processing, and synchronizing the digital and physical systems. To address these challenges, this study develops a traffic DT based on a real-world connected vehicle corridor. Leveraging the Cellular Vehicle-to-Everything (C-V2X) infrastructure in the corridor, along with communication, computing, and simulation technologies, the DT accurately replicates physical vehicle behaviors, signal timing, communications, and traffic patterns within the virtual environment. Building upon the previous data pipeline, the digital system ensures robust synchronization with the physical environment. Moreover, the DT’s scalable and redundant architecture enhances data integrity, making it capable of supporting future large-scale C-V2X deployments. Lastly, the DT’s ability to provide feedback to the physical system is demonstrated through applications such as signal timing adjustments, vehicle advisory messages, and incident notifications. The proposed DT is a vital tool in T-CPS, enabling real-time traffic monitoring, prediction, and optimization to enhance the safety and mobility of transportation systems.

Cellular vehicle-to-everything (C-V2X) is an emerging technology that has the potential to significantly improve traffic safety and efficiency through various applications such as pedestrian alerts, traffic hazard warnings, signal timing information broadcasting, etc. With the help of the Federal Communications Commission's agreement to waive restrictive rules on C-V2X deployment, extensive real-world testing and implementation of C-V2X is on the horizon. However, the performance of C-V2X in urban environments remains unclear as existing studies have been focusing on evaluating it in either simulated or highway environments. To facilitate the understanding of its performance in urban environments, this study has conducted a systematic evaluation of the efficiency and reliability of C-V2X using various metrics based on real-world data collected through roadside units (RSUs) and onboard units in urban environments. Experimental results indicate that C-V2X has relatively stable efficiency (i.e., latency) under different conditions. However, its reliability (i.e., packet error rate and spatial delivery rate) is significantly affected by multiple factors, including the distance from the vehicle to the RSU and to its surrounding buildings. Moreover, this study has proposed a concept of machine learning-assisted RSU planning to facilitate the planning and deployment of C-V2X in urban areas. An XGBoost model is developed to accurately estimate the probability of the delivery of V2X messages using C-V2X. Results from explaining the model suggest that the planning of RSUs should consider the spacing of RSUs and the density of buildings in the area.

This paper proposes a framework for deep Long Short-Term Memory (D-LSTM) network embedded model predictive control (MPC) for car-following control of connected automated vehicles (CAVs) in traffic mixed with human-driven vehicles (HDVs) and CAVs. The framework consists of: 1) lead HDV trajectory prediction through D-LSTM; and 2) CAV car-following control via MPC based on the predicted vehicle trajectory. For the trajectory prediction, two D-LSTM structures are developed based on the availability of preceding vehicle information: 1) ‘sufficient’ historical information of the position and speed of multiple vehicles ahead; and 2) ‘insufficient’ information where preceding vehicle information is unavailable (e.g., due to failed communication). Based on the prediction, a distributed MPC is designed for each scenario by incorporating the predicted trajectory into state space construction. The proposed D-LSTM models are trained and tested with the NGSIM data for validation. Numerical simulation results for various traffic conditions suggest that the proposed strategies perform better than traditional MPC methods in terms of control objective cost reduction, smoother control, and stabilizing effect. The results also indicate that the sufficient information case outperforms the insufficient information case as expected, which highlights the importance of stable communication.