Sensor‐Based Long Baseline Navigation: Observability Analysis and Filter Design

Abstract

AbstractThis paper presents a novel long baseline (LBL) position and velocity navigation filter for underwater vehicles based directly on the sensor measurements. The solution departs from previous approaches as the range measurements are explicitly embedded in the filter design, therefore avoiding inversion algorithms and allowing also the consideration of the cases of reduced numbers of readings, in particular when there are only two or three distance measurements. The nonlinear system dynamics are considered to their full extent and no linearizations are carried out whatsoever. The filter error dynamics are globally exponentially stable (GES) and it is shown, in a realistic simulation environment, that the filter achieves similar performance to the extended Kalman filter (EKF) and outperforms linear position and velocity filters based on algebraic estimates of the position obtained from the range measurements.