Abstract
The ability to make oceanographic scientific measurements without the need for fixed hardware is of interest to the US Navy. The Office of Naval Research (ONR) has been exploring the feasibility of using autonomous underwater vehicles (AUVs) to conduct such measurements. To this end, the University of Idaho (UI) has developed a fleet of AUVs.
To better operate in the presence of ocean currents, the UI AUVs have been equipped with a higher powered motor. As AUV operations increased in speed, the control gains have been investigated through the use of simulation and field testing to achieve more stable performance.
When making scientific measurements it is crucial to both temporally and spatially local-ize the AUVs. To accomplish this an extended Kalman filter (EKF) is employed to estimate the position and orientation of the UI AUVs. To improve this state estimation in oceanic environments, a new EKF was designed to directly estimate the ocean currents. These esti-mated currents have been utilized to adapt the heading controller to account for the effects of these currents.
The new EKF and adapted heading controller were simulated and subsequently field tested. Although simulation results were promising, an ineffective speed correlation caused large estimation divergence during field testing. Despite this divergence, the concept of esti-mating currents and leveraging them to improve navigation shows feasibility.