Improvement of velocity estimate resolution for a correlation velocity log using surface fitting methods

Boltryk, P., Hill, M., Keary, A., Philips, B., Robinson, H. and White, P. (2002) Improvement of velocity estimate resolution for a correlation velocity log using surface fitting methods. In Proceedings of Oceans '02 MTS/IEEE. vol. 3, IEEE. pp. 1840-1848 .

Abstract

A Correlation Velocity Log (CVL) is an acoustical navigation aid, consisting of a transmitter and a receiver array. Moving patterns in the wave fronts across the receiver army, indicative of motion, are detected by calculating correlation coefficient between bottom reflections from consecutive transmitted pulses. A velocity vector map is constructed by plotting the correlation coefficient on a grid representing the, velocity vector spacing. The peak position on this map surface provides an estimate of the velocity vector of the vessel.
The resolution of the velocity vector map is limited by the minimum diameter of the receiving elements and the maximum size of the array. Current research is focused on exploiting the full potential of the CVL by investigating surface fitting techniques to' interpolate peak position estimates between measurement points.
Detailed simulations and sea trials of a prototype CVL have demonstrated that the correlation coefficient surface is subject to high variance at only a modest distance from the peak. This paper presents numerical comparisons between different surface-fitting methods for performing peak finding on the data. Different least squares schemes are investigated, together with a non-liner iterative implementation of Maximum Likelihood Estimation. The performance of these methods is compared both in terms of rms error and mean, error in peak estimation.
It is shown that CVL performance can he improved significantly by the use of an appropriate peak finding technique.

This record has no associated files available for download.

Contributors

Author: M. Hill

Author: P. White

Download statistics