A model-based algorithm for environmentally adaptive bathymetry and sound velocity profile estimation.

摘要

This dissertation presents the derivation and demonstration of novel, Adaptive Bathymetric Estimation algorithms (ABE) for use with forward looking sonar arrays in shallow waters. In addition to providing high-resolution localization of the targets on the ocean bottom in front of the host vehicle, it automatically estimates a local sound speed profile (SSP) model, and corrects the sonar localization for ray bending arising from refraction. The implementation of this algorithm arises from the combination of three advanced technologies: (1) high frequency, high resolution forward looking sonar, (2) advanced satellite based navigation and positioning systems (GPS), and (3) model-based parameter estimation techniques based on the Extended Kalman Filter. During the course of this work, the developed algorithms are analyzed through computer simulations using statistical measures of their performance and consistency. The presentation of the development is staged, with each chapter increasing the internal complexity of the model and adding capability. We then present an in-water experiment which tested the validity of the ABE in real-world operation. We evaluate the algorithm's performance in a shallow water environment, and show results of the success and limitations of the technique.