A comparison of AUV navigation performance: a system approach

This paper addresses the technical challenges in underwater navigation for autonomous underwater vehicles (AUVs) in terms of selecting an appropriate combination of navigation system and software processing. There are many existing systems that can be used onboard an AUV with different costs and capabilities. It is not immediately clear to ocean scientists which system is appropriate for a given set of navigation requirements. To address this challenge, this paper proposes a novel metric called navigation efficiency which is intended to provide a framework with which the navigation performance can be related directly to a given sensor suite and survey requirements. Three basic survey patterns are considered although more complicated survey patterns can be created using these basic patterns, and the navigation efficiency modified accordingly. To address the limitation in software processing for vehicle motion in the presence of wave disturbances, this paper provides new insights into how the existing error state Kalman filtering techniques can be used to attenuate the effect of wave disturbances. To substantiate the insights, an error state filter is applied to estimating the bias in gyro measurements although the basic idea can be extended to estimating other motion variables. The main contribution of this paper is twofolded. Firstly, it provides ocean scientists an easily understood framework with which to size and specify the needed navigation systems. Secondly, it provides navigation engineers a new appreciation of the use of error state filters to characterize complicated vehicle motion.