National-scale Traffic Model Calibration in Real Time with Multi-source Incomplete Data

Abstract

Real-time traffic modeling at national scale is essential to many applications, but its calibration is extremely challenging due to its large spatial and fine temporal coverage. The existing work is focused on urban-scale calibration with complete field data from single data sources (e.g., loop sensors or taxis), which cannot be generalized to national scale because complete single-source field data at national scale are almost impossible to obtain. To address this challenge, in this article, we design MultiCalib, a model calibration framework to optimize traffic models based on multiple incomplete data sources at national scale in real time. Instead of simply combining multi-source data, we theoretically formulate a multi-source model calibration problem based on real-world contexts and multi-view learning. In particular, we design (i) convex multi-view learning to integrate multi-source data by quantifying biases of data sources, and (ii) context-aware tensor decomposition to infer incomplete multi-source data by extracting real-world contexts. More importantly, we implement and evaluate MultiCalib with two heterogeneous nationwide vehicle networks with 340,000 vehicles to infer traffic conditions on 36 expressways and 119 highways, along with four cities across China. The results show that MultiCalib outperforms baseline calibration by 25% on average with the same input data. Based on the proposed national-scale traffic model calibration, we design a novel dispatching framework integrated with our speed calibration model where we guide a vehicular fleet among national-scale highways with a routing strategy to reduce general traveling time. The results show that a routing strategy based on MultiCalib outperforms a routing strategy based on a state-of-the-art traffic model by 45% on average.