Cooperative acoustic navigation of underwater vehicles without a DVL utilizing a dynamic process model: Theory and field evaluation

Abstract

This paper reports the theoretical development and at‐sea field evaluation of a novel combined underwater acoustic communication and navigation system, known as cooperative acoustic navigation (CAN), for underwater vehicles (UVs) utilizing a second‐order dynamic plant model of the submerged UVs. The present state‐of‐the‐art in CAN is to utilize one‐way travel‐time acoustic modem telemetry together with purely kinematic, constant‐velocity plant process models. We term this approach CAN‐KIN. At present, CAN‐KIN is utilized with an on‐board bottom‐lock Doppler velocity log (DVL) providing frequent, high‐accuracy velocity corrections. However, DVLs are relatively expensive, have significant power requirements, can be physically large, and have limited acoustic bottom‐lock range, which restricts their use to a maximum of 25–420 m above the sea floor. In this study, we investigate the utility of a second‐order dynamic UV plant process model in CAN of UVs equipped with an acoustic modem, attitude, and depth sensors, but lacking a DVL, and a surface ship equipped with an acoustic modem and global positioning system. We term CAN utilizing a dynamic model CAN‐DYN. This paper reports results from at‐sea field trials conducted in the Chesapeake Bay with the Johns Hopkins University Iver3 UV. These experimental results indicate the submerged UV position estimate from CAN‐KIN is poor and even unstable in the absence of DVL velocity observations. These field experimental results also show that CAN‐DYN performs well without a DVL. Our results suggest CAN‐DYN without a DVL does not exhibit instability as does CAN‐KIN without a DVL, performs similarly to CAN‐KIN with a DVL, and outperforms DVL‐based dead reckoning. Additionally, we report an experimental evaluation of the effect of adding (relative) velocity corrections in the form of acoustic range‐rate observations to CAN utilizing a dynamic model without a DVL. We conclude that the addition of infrequent velocity observations, such as those provided by acoustic range rate, does not appear to improve the performance of CAN‐DYN without a DVL.

Underwater AUV Navigation Dataset in Natural Scenarios

“…For instance, Harris et al developed an AUV position estimation algorithm using the ensemble Kalman filter (EnKF) and fuzzy Kalman filter (FKF), which avoids linearization of the AUV's dynamics model [5]. Jin et al proposed a single-source assisted passive localization method that combines acoustic positioning with inertial navigation and concluded that time difference of arrival (TDOA) + AOA yields better results [17].…”

“…Specifically, for stand-alone strapdown inertial navigation systems (SINS), the estimation of relative velocity and position involves the integration of accelerometer and gyroscope sensor data, which introduces errors and leads to significant drift in the estimated position and velocity [4]. The introduction of a DVL assists navigation by measuring bottom-towater relative velocity to improve positioning accuracy, but prolonged error accumulation remains a challenge [5]. Navigation methods based on other perceptual sensors, such as vision and sonar, have been applied in specific scenarios but mostly remain in the laboratory stage [6,7].…”

Stable nullspace adaptive parameter identification of 6 degree-of-freedom plant and actuator models for underactuated vehicles: Theory and experimental evaluation

“…To the best of our knowledge, with the exception of , Harris (2019), Paine (2018), upon which this work is based, there are no previously reported studies for parameter identification of underactuated robotic vehicles that simultaneously identify plant-model parameters and actuation parameters in 6-DOF.…”

A survey on tracking control of unmanned underwater vehicles: Experiments-based approach

“…Expensive cost of sensors and actuators: The high cost of accurate sensors and actuators is one of the financial constraints when designing a new robot Harris and Whitcomb (2021). This problem becomes more critical for the case of building a low-cost version of the robots to be deployed for special intervention missions Ji et al (2021).…”