Traffic Simulations: Multi-City Calibration of Metropolitan Highway Networks

2025-01-08

AI Summary

Calibrating traffic flow models for large-scale metropolitan highway networks remains challenging due to sparse, low-resolution sensor data and computational intractability of conventional black-box optimization methods. Method: This paper proposes a path-level travel time–driven demand calibration framework tailored for high-resolution stochastic traffic microsimulators. Unlike traditional link-based approaches relying on sparse detector data, our method employs an interpretable, sample-efficient, path-oriented calibration strategy that avoids heuristic black-box optimizers (e.g., SPSA). Contribution/Results: We demonstrate the first systematic, scalable calibration across six metropolitan networks and 54 diverse scenarios. The framework significantly improves cross-network generalizability and achieves an average 43.5% gain in fitting accuracy over SPSA (up to 80.0%), while drastically reducing simulator evaluations. This establishes a novel paradigm for real-time, large-scale dynamic network calibration.

Technology Category

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

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
• Security and Privacy: Large-scale security measurements

Abstract

This paper proposes an approach to perform travel demand calibration for high-resolution stochastic traffic simulators. It employs abundant travel times at the path-level, departing from the standard practice of resorting to scarce segment-level sensor counts. The proposed approach is shown to tackle high-dimensional instances in a sample-efficient way. For the first time, case studies on 6 metropolitan highway networks are carried out, considering a total of 54 calibration scenarios. This is the first work to show the ability of a calibration algorithm to systematically scale across networks. Compared to the state-of-the-art simultaneous perturbation stochastic approximation (SPSA) algorithm, the proposed approach enhances fit to field data by an average 43.5% with a maximum improvement of 80.0%, and does so within fewer simulation calls.

Problem

• Traffic Flow Optimization
• Highway Network
• Urban Transportation

Innovation

• Travel Demand Prediction
• Path Travel Time Data
• Enhanced Accuracy