Network-wide Traffic State Estimation using the Macroscopic Fundamental Diagram: A data fusion approach
Marianthi Mermygka · Research Repository (Delft University of Technology) · 2016
an incident occurred in roads with low demand, the proposed method was still able to predict the traffic state.However, in some cases that the incident was placed in spots with higher demand, the derived traffic state was out of bound.In these cases, it was observed that the average network speed drop after five minutes was remarkably high and over 20%.This indicates that the speed drop can potentially be used as a sign that an incident occurs in the network and the MFD cannot describe the traffic state accurately any more.Concluding, this project managed to estimate the traffic state of a simulated network using the MFD and speed data.The results showed that the derived traffic state was within acceptable error bounds to describe accurately the traffic situation in the network.This means that the traffic state derived from the proposed process can be used as a solid and reliable base for traffic state prediction and traffic control strategies aiming at the optimization of the traffic system.Further research could focus on exploring how efficiently the proposed process can be used for these purposes.Furthermore, the application of the same traffic state estimation process in other networks could offer insights in the potential generalization of the method and the establishment of the MFD as a simple but powerful traffic state estimation tool.