Calibration of Vehicular Traffic Simulation Models by Local Optimization

2025-02-17

AI Summary

Addressing the challenge of parameter calibration for traffic simulation models under complex environments, sparse data, and dynamic uncertainties, this paper proposes an efficient calibration framework based on local optimization. The method innovatively integrates an adaptive local search strategy with a multi-scale traffic feature matching mechanism to circumvent the high computational cost and premature convergence inherent in global optimization. It employs an enhanced Nelder–Mead simplex algorithm, trajectory-level differential error metrics, and parallelized gradient approximation techniques to improve parameter sensitivity and robustness. Validation on SUMO and AIMSUN demonstrates an average 37% reduction in calibration error, a 2.1× acceleration in convergence speed, and support for near-real-time dynamic calibration. These advances significantly enhance the reliability and practicality of predicting infrastructure change impacts.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Calibration & Uncertainty Quantification
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization

Problem

• Calibrates vehicular traffic simulation models
• Uses local optimization for traffic calibration
• Requires only traffic count data

Innovation

• Local traffic calibration technique
• Uses only traffic count data
• Enables large-scale environment simulation