Magnetometers independent heading estimation strategy for UUV based on position and speed observations

For Unmanned Underwater Vehicles (UUV), the problem of position estimation is crucial for many different aspects ranging from good navigation performance to suitable geo-referencing capabilities of the payload data. The possibility of accurately estimating the UUV orientation has a positive impact on it, as the most of position estimation algorithms rely on the orientation signal. Among the orientation degrees of freedom, the heading angle has the highest importance in this process. Because its determination is strongly based on the use of magnetometers, it is also the one mostly affect by environmental unpredictable disturbances. In the recent past, the authors worked on a complementary filter capable of detecting and ignoring the magnetometers samples affected by disturbances. This way the navigation system of the vehicle can always exploit a reliable orientation information. In order to allow the UUV working under strongly disturbed conditions (e.g. inspections of modern wrecks), in this paper the bases for a heading estimation strategy that could completely avoid the use of magnetometers is proposed. The only involved devices are positioning sensors (Global Positioning System - GPS on the surface or Ultra-Short BaseLine USBL system underwater) velocity sensor (Doppler Velocity Log - DVL) and gyroscopes. The results of an observability analysis are provided highlighting that the heading degree of freedom results observable except for few trajectories along which a real vehicle reasonably never travels. Experimental results are provided as a first validation of the proposed approach. MARTA AUV was controlled to complete a mission on an area free of magnetic disturbances; the outcome of the proposed algorithm is compared with the one of the onboard orientation estimation algorithm that could exploit also the magnetometers signal. The slightly difference between the two estimations is a first encouraging result towards its on-line application.