A derivative-free nonlinear algorithm for speed estimation using data from single loop detectors

This paper presents a derivative-free algorithm for speed estimation using occupancy and count outputs from single loop detectors. An unscented Kalman filter (UKF) is used for the nonlinear speed estimation problem and achieved excellent results. Data from a Texas Transportation Institute (TTI) vehicle detector test bed are used for the implementation of the UKF and also for performance evaluation of the implemented algorithm. The results showed that the UKF method has superior performance to other prior methods