The industry's recognized authority for design, engineering and application of equipment and services in the global ocean community
Issue link: http://sea-technology.epubxp.com/i/432125
www.sea-technology.com December 2014 / st 41 To estimate the vehicle position reliably using low-cost sensors and acoustic ranges from multiple beacons, we de- signed a novel EKF navigation algorithm that integrates an initiation, outlier rejection and state noise set method. The R eliable navigation is critical for AUVs, not only for the sake of safe operation and recovery, but also for scientifc surveys. AUVs are now be- ing used for a variety of tasks, including oceanographic surveys, demining and bathymetric data collection in marine and river environments. AUV naviga- tion is a challenging problem, primar- ily due to the rapid attenuation of high- frequency signals and the unstructured nature of the undersea environment. Acoustic EKF navigation systems are still a popular method. The feld is in the midst of a paradigm shift from old technologies, such as long baseline (LBL) that require predeployed and lo- calized infrastructure, toward dynamic system approaches that allow for rapid deployment and fexibility with mini- mal infrastructure. Other systems, like ultrashort baseline (USBL), are em- ployed to track and locate underwater vehicles. However, the EKF navigation algorithm of AUVs is challenging be- cause it is diffcult to design an effec- tive online processing procedure for travel time measurements, which may be affected by noise, drop outs and outliers. Although the inertial naviga- tion unit with high accuracy would perform well in the heading measurement, it costs much more than an electronic compass and consequently leads to a high expenditure in AUV development. We equipped an AUV with a PNI Sensor Corp. (Santa Rosa, California) TCM5 electronic compass and a 300-ki- lohertz acoustic Doppler velocity log (DVL) developed by the Institute of Acoustics, Chinese Academy of Sciences, as heading and velocity sensors. As part of its sensor suite, a customized, multiple-beacon navigation system was equipped on the AUV. Reliable EKF Algorithm For Low-Cost AUV Navigation Method Integrates Initialization, Outlier Rejection and State Noise By Daxiong Ji • Kai Shi • Dongdong Li (Top) The error of initializing methods. The x-axis is the time sequence of sampling, and the period is 24 seconds. (Bottom) Vehicle track comparison of two navigation algorithms (with- out outlier rejection versus with outlier rejection).