abstract
"Most navigation systems calculate the optimal route and time based on real-time vehicle traffic. However, they often seem to fall short in considering the real-time traffic changes during transit, along with anticipated traffic congestion. This project proposes an algorithm that calculates the optimal route and time, integrating real-time traffic information with historical traffic data.”
Key Achievements
- In a single route search, we can provide the optimal route and time by considering both real-time traffic data and historical traffic data, rather than considering them individually.
- For each road, we predict the arrival time and use the variance in the distribution of traffic volume at that arrival time, considering both real-time and historical traffic data, as a weight to provide accurate predictions.
- For instance, our system can simultaneously consider the possibility that a currently congested segment might clear up by the time of arrival, as well as the potential for a currently free-flowing segment to become congested in the future when passing through.
Published state
Overview
real-time speed modeling
prior knowledge