Modal tracking in a shallow ocean: An adaptive particle filtering approach.

Abstract

It is well-known that the shallow ocean is an ever changing, uncertain environment dominated by temperature fluctuations and ambient noise. The need for a processor that can adapt to these environmental changes while simultaneously tracking the evolution of modal functions is necessary for localization, inversion, and enhancement. An approach to this problem is made possible by developing a sequential Bayesian processor capable of providing a joint solution to both the modal function tracking (estimation) and environmental adaptivity problem. The posterior distribution required will contain multiple peaks and space-varying statistics requiring a sequential (nonstationary) Bayesian approach. The final design that evolves is a so-called particle filter. The particle filter is a sequential Markov chain Monte Carlo processor capable of providing reasonable performance for a multi-mode (probability distribution), space-varying problem. The tracking results are applied to synthesized pressure-field data from the actual experimental parameters. The tracking results are compared to that of the modern unscented Kalman tracker solution.