Investigation of speed estimation using single loop detectors.

The ability to collect or estimate accurate speed information is of great importance to a large number ofIntelligent Transportation Systems (ITS) applications. Estimating speeds from the widely used singleinductive loop sensor has been a difficult, yet important challenge for transportation engineers. Based onempirical evidence observed from the sensor data from two metropolitan regions in Northern Virginia andCalifornia, this research effort developed a Kalman filter model to perform speed estimation for congestedtraffic. Taking advantage of the coexistence of dual loop and single loop stations in typical freewaymanagement systems, a calibration procedure was proposed for seeding and initiating the algorithm.Empirical evaluation showed that the proposed algorithm can produce accurate speed estimates (on the orderof 1-3 miles/hour error) under congested traffic conditions