A nonlinear complementary filter for underwater navigation using inertial measurements

This paper describes a nonlinear complementary filter capable of estimating the course motion variables namely the position, velocity, heading and accelerometers bias of an agile, over-actuated AUV during underwater operations, using the inertial sensors (IMU), the DVL, the depth sensor and the compass. The proposed work is within the framework of the V-Fides project, co-funded by Tuscany Region (Italy) and developed by a team lead by WASS S.p.A. (Whitehead Sistemi Subacquei, Livorno). The aim of the project was to develop and evaluate an high-depth, over-actuated, long endurance Autonomous Underwater Vehicle (AUV). The paper proposes the mathematical development of the observer, together with some experimental results, able to demonstrate the capabilities of the estimation scheme, compared with the estimations obtained via a standard Kalman Filter.