IntTrajSim: Trajectory Prediction for Simulating Multi-Vehicle driving at Signalized Intersections

2025-06-10

AI Summary

To address the limited fidelity of rule-based traffic simulators in reproducing realistic driving behaviors, this paper proposes a data-driven microscopic simulation system for signalized intersections. Methodologically, we first establish a generative trajectory model evaluation framework tailored to traffic engineering requirements; design a simulation-in-the-loop validation architecture; and introduce a multi-head self-attention trajectory prediction model augmented with signal phase information. Our approach integrates deep generative modeling, trajectory forecasting, and feedback control, and is embedded within mainstream traffic simulation engines. Experimental results demonstrate that the proposed model significantly outperforms baseline methods across both macroscopic metrics (e.g., capacity, delay) and microscopic metrics (e.g., car-following fidelity, conflict rate), achieving high behavioral realism and interpretability. This work establishes a novel paradigm for evaluating intelligent transportation systems through data-informed, closed-loop simulation.

Technology Category

• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Planning, Routing, and Scheduling: Model-Based Reasoning
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Web Mining and Content Analysis: Web data generation and simulation
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Traffic simulators are widely used to study the operational efficiency of road infrastructure, but their rule-based approach limits their ability to mimic real-world driving behavior. Traffic intersections are critical components of the road infrastructure, both in terms of safety risk (nearly 28% of fatal crashes and 58% of nonfatal crashes happen at intersections) as well as the operational efficiency of a road corridor. This raises an important question: can we create a data-driven simulator that can mimic the macro- and micro-statistics of the driving behavior at a traffic intersection? Deep Generative Modeling-based trajectory prediction models provide a good starting point to model the complex dynamics of vehicles at an intersection. But they are not tested in a"live"micro-simulation scenario and are not evaluated on traffic engineering-related metrics. In this study, we propose traffic engineering-related metrics to evaluate generative trajectory prediction models and provide a simulation-in-the-loop pipeline to do so. We also provide a multi-headed self-attention-based trajectory prediction model that incorporates the signal information, which outperforms our previous models on the evaluation metrics.

Problem

• Simulate realistic multi-vehicle driving behavior at signalized intersections
• Evaluate generative trajectory models using traffic engineering metrics
• Develop data-driven simulator for intersection safety and efficiency analysis

Innovation

• Data-driven simulator for intersection traffic behavior
• Simulation-in-the-loop pipeline for evaluation
• Multi-headed self-attention-based trajectory prediction model