An adaptive particle filtering approach to tracking modes in a varying shallow ocean environment

Abstract

Abstract-The shallow ocean environment is ever changing mostly due to temperature variations in its upper layers (< 100m) directly affecting sound propagation throughout. The need to develop processors that are capable of tracking these changes implies a stochastic as well as an "adaptive" design. The stochastic requirement follows directly from the multitude of variations created by uncertain parameters and noise. Some work has been accomplished in this area, but the stochastic nature was constrained to Gaussian uncertainties. It has been clear for a long time that this constraint was not particularly realistic leading a Bayesian approach that enables the representation of any uncertainty distribution. Sequential Bayesian techniques enable a class of processors capable of performing in an uncertain, nonstationary (varying statistics), non-Gaussian, variable shallow ocean. In this paper adaptive processors providing enhanced signals for acoustic hydrophone measurements on a vertical array as well as enhanced modal function estimates are developed. Synthetic data is provided to demonstrate that this approach is viable.

Adaptive wavenumber estimation for mode tracking in a shallow ocean environment

“…The parametric adaption was based on simultaneously estimating the horizontal wavenumbers along with the modes and pressure-field as compared to previous work that concentrated on estimating the modal coefficients as the environmental parameters of interest [25], [26]. These parameters were more challenging from a processor design perspective because of their increased sensitivity to environmental change compared to the modal coefficients.…”

“…Since we are primarily interested in an environmentally adaptive processor, that is, a processor capable of adjusting its parameters to variations in the environment such as temperature, noise, etc. We choose to capture these changes by allowing the horizontal wavenumber to vary as compared to our previous efforts [25], [26] that focused on modal coefficient estimation. Here, we define the parameter vector as…”

“…As mentioned previously [26] the random walk model can provide constraints in the simulation, since the parameter is modeled as Gauss-Markov implying that 95% of the samples must lie within confidence limits controlled by (±1.96 R wθ m wθ m ). This constitutes a soft statistical constraint of the parameter variations [15].…”

“…It has already been shown that this parametrically adaptive approach can operate successfully in this environment [25], [26] when estimating modal coefficients; however, here we construct processors to adapt to the horizontal wavenumber-a more environmentally sensitive parameter. We begin by formulating the problem in a state-space framework and developing the necessary mathematics for processor design.…”

Adaptive particle filtering for mode tracking: A shallow ocean application

“…In this way, we can achieve a detailed analysis of the processor performance prior to fielding and operational version. In this paper we constrain our discussion results to processing the noisy experimental pressure-field measurements [12] as compared to a previous paper [24] that concentrated on the synthetic data set generated from the forward propagator.…”