Approximative network partitioning for MFDs from stationary sensor data

Abstract

Macroscopic fundamental diagrams (MFD) are typically estimated for sub-regions in cities, where the zoning conventionally follows the directive of homogeneity in the road network and traffic. Existing network partitioning algorithms fail to work for MFD estimation with stationary sensor data as their link speed and link density measurement are biased by the location of the sensor. However, in this paper, we present an approach for approximative network partitioning for MFDs from stationary sensors.