A Depth-Estimation Approach Enabling Closed-Loop Control of Water Surface Gliding

Aerial-aquatic unmanned vehicle(AAUV) is a new class of vehicles capable of operating in multiple domains, specifically in the air and water. These vehicles can glide on the water surface using onboard hydrofoil. Accurate depth measurement is required to ensure closed-loop control of its water surface gliding. Existing methods of measuring water depth usually use a single sensor so that accuracy could be improved, but some methods are not suitable for mounting on autonomous vehicles. To address the above problems, a fusion method based on adaptive weighting and Kalman filtering is designed to make the final depth results more accurate. The accurate depth measurement obtained during AAUV gliding ensures stability during the process. Compared with a single pressure sensor using only the Kalman filter, the average error of the fusion result is reduced by 53.1%, and the maximum error is reduced by 61.6%. Experimental results show that the proposed algorithm is easy to implement and improves accuracy and stability.