Consistent Linear Tracker With Converted Range, Bearing, and Range Rate Measurements

Abstract

Active sonar and radar systems often include a measurement of range rate in addition to the position-based measurements of range and bearing. Due to the nonlinearity of the range rate measurement with respect to a target state in the Cartesian coordinate system, this measurement is not always fully exploited. The state of the art methods to utilize range rate include the extended Kalman filter and the unscented Kalman filter with sequential processing of the position based measurements and the range rate measurement. Common to these approaches is that the measurement prediction function remains nonlinear. The goal of this work is to develop a measurement conversion from range, bearing, and range rate to Cartesian position and velocity that is unbiased and consistent, with appropriate elimination of estimation bias. The converted measurement is then used with a linear Kalman filter. Performance of this new method is compared to state of the art techniques and shown to match or exceed that of existing techniques over a wide range of scenarios.