Validation of a Subsea Navigation Analysis Program

Abstract

Subsea navigation is a critical technology for successful employment of Unmanned Undersea Vehicles (UUVs). Lockheed Martin (LM) has developed a Subsea Navigation Analysis Program (SNAP) to provide rapid assessment of the performance of instrument suites and maneuvers for UUV navigation. SNAP utilizes a library of detailed error models of commercial navigation instruments, coupled with a path-dependent covariance analysis to perform navigation analysis many times faster than real-time. In order to validate the SNAP tool, a Mobile Inertial Test System (MITS) was designed and integrated into a small craft owned and operated by LM in Riviera Beach, FL. The MITS was outfitted with Commercial - Off the Shelf (COTS) Global Positioning System (GPS) receivers and a Kearfott SeaDevil underwater navigation system. The SeaDevil is an integrated Kearfott Inertial Navigation Unit (INU) and a RDInstruments Workhorse Navigator Doppler Velocity Log (DVL). The SNAP tool was validated and precision alignment techniques were developed during two at-sea test periods. In addition to validating the SNAP tool, an analysis of alignment accuracy was performed. Alignment is the process of providing the INU with data which allows determination of true heading. The SeaDevil allows for gyrocompass alignment aboard the small craft while stationary at the dock. Alignment can also be achieved by providing the INU position updates from a GPS reference while the system is moving, a technique known as Moving Base Alignment (MBA). The position reference was passed to the INU through a data system, rather than providing direct NMEA GPS data to emulate the conditions which would be experienced by the UUV during pre-dive preparations in a ship-board hangar. The alignment quality was estimated by the INU's Kalman Filter as part of normal Kalman Filter processing, and was calculated based on the cross-track errors observed in the data. Both MBA and gyrocompass alignment techniques proved effective, although the quality of the MBA alignment suffered when a noisy / high-latency position reference was used. The test program consisted of two missions which were run in the in-shore waters near Riviera Beach. The first mission involved shuttling between channel markers located approximately 1 Nautical Mile (NM) apart. This mission emulated a reconnaissance / search mission. The second mission was a transit of approximately 10 NM. Both test cases provided valuable insight into UUV performance, although the transit was found to yield the most beneficial information. The SNAP tool was found to provide good predictions of both alignment performance and navigation performance for all of the tests conducted. The tests also confirmed that the SeaDevil Kalman Filter provides a reasonable estimate of navigation uncertainty and can be used to monitor operational UUV navigation performance. The MITS was found to be a versatile and valuable test tool for assessing navigational performance. The initial design of the MITS was found to be too susceptible to weather factors, causing most of the testing to be performed in-shore, where salinity gradients complicated the testing. An upgraded MITS has been developed, which is suitable for mounting on a number of platforms and has improved weather tolerance. Future inertial testing at the LM Riviera beach site will be conducted with this improved test system.