Estimation of roadway traffic density on freeways using presence detector data

A simple data processing algorithm is presented for estimating the density of vehicular traffic flow on freeways between two conventional (presence) type loop detectors. Using flow and occupancy data from adjacent detector stations, a simple scalar Kalman filter is derived under assumption of homogeneous (in space) flow conditions. Correction for inhomogeneous conditions is provided by a bias detection-compensation technique based on the generalized likelihood ratio method. Accurate performance over a wide range of speed-density flow conditions is demonstrated.