Neural Network Parameter Estimation for the Modified Bistatic Scattering Strength Model (BISSM)

Abstract

This technical note investigates the estimation of environmental parameters in the Bistatic Scattering Strength Model (BISSM) given backscatter strength and bathymetric data. A monostatic version of the model is derived, since this will be the form of data provided by acoustic imaging sensors. Feedforward neural networks, using the backpropagation learning algorithm, are used to perform the estimation of parameters for the nonlinear BISSM equation. The parameters that can be estimated are identified, and neural networks have been developed to estimate these parameters. Using noise-free artificial data generated with the BISSM equation, the networks provided excellent estimates of the desired parameters. The primary impetus for this work is a need for the Naval Oceanographic Office (NAVOCENO) to provide relevant survey support for Low-Frequency Active Acoustics (LFAA) programs and future Low-Frequency Active (LFA) operational systems. It has been recognized by the Commander, Naval Oceanographic Command (CNOC) that such support will require knowledge of certain bottom and subbottom properties and high-resolution geomorphology. The BISSM model has been proposed as a model for aspects of active bottom reverberation.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1991
Accession Number
ADA247913

Entities

People

  • Brian S. Bourgeois

Organizations

  • United States Naval Research Laboratory

Tags

Communities of Interest

  • Air Platforms
  • Materials and Manufacturing Processes
  • Sensors

DTIC Thesaurus Topics

  • Acoustic Scattering
  • Acoustics
  • Algorithms
  • Angle Of Incidence
  • Backscattering
  • Computations
  • Computing System Architectures
  • Frequency
  • Grazing Angles
  • Low Angles
  • Measurement
  • Network Architecture
  • Neural Networks
  • Probability
  • Probability Density Functions
  • Scattering
  • Signal Processing

Readers

  • Acoustical Oceanography.
  • Neural Network Machine Learning.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - DoD AI Strategy
  • AI & ML - Neural Networks