Freeway Traffic State Estimation using Fixed and Mobile Sensing Data with Microscopic Simulation Evaluation.

Freeway traffic state estimation based on macroscopic traffic flow model METANET and extended Kalman filtering used to be conducted with fixed sensing data only. Recently the work was extended to the mixed sensing case of fixed and mobile sensors, and evaluated using the NGSIM data, with some significant conclusions. However, the highway stretch covered by NGSIM is unfortunately very short, the corresponding data duration is quite limited, and the involved congestion is not strong. To further demonstrate the performance of the designed traffic state estimator and verify the conclusions obtained with NGSIM, the estimator is studied and evaluated in this paper in microscopic simulation based on AIMSUN for a long freeway stretch of on/off-ramps and heavy congestion.