UACE2017 Proceedings: Quantifying The Impact Of Uncertainty On Sonar Performance Predictions.

• Session:
  Sonar performance modeling and verification - Applications to active and passive sonar
• Paper:
  Quantifying The Impact Of Uncertainty On Sonar Performance Predictions.
• Author(s):
  Mark Prior, Mathieu Colin
• Abstract:
  Sonar performance predictions are vital to the design of new sonar systems, the optimal deployment of existing systems and to the analysis of experimental data. The ocean environment has a strong influence on underwater acoustic propagation, scattering and noise-generating processes. This means that accurate prediction of sonar performance is usually achieved with computer codes that require high-quality input data describing meteorological, oceanographic and geological conditions in the environment of interest.\nData describing environmental conditions are available from databases and forecasts but significant uncertainties are associated with them. These uncertainties may come from a fundamental lack of knowledge of, for example, the properties of seabed sediments. They may also stem from modelling uncertainties in forecasts that predict ocean and weather conditions. Further input-data uncertainties arise from the precision to which the problem under study is defined in terms of temporal and spatial spreads for which ‘representative’ sonar-performance predictions are required. \nA second source of uncertainty arises from the adversarial nature of scenarios for which sonar-performance calculations are typically required. Submerged targets will modify their behavior to minimize detectability and any prediction of sonar performance is subject to potentially large changes arising from choices made outside the control of the sonar operator.\nA methodology is described that captures uncertainties associated with input-data and tactical considerations and produces estimates of their impact on measures of sonar performance. Input-data-uncertainty impacts are quantified via a “scaled sensitivity” parameter that measures the importance of uncertainty, including all inter-parameter dependencies. Tactical uncertainties are handled using game-theoretic approaches.
•   Download the full paper

Contact details

• Contact person:
  Dr Mark Prior
• e-mail:
  
• Affiliation:
  TNO
• Country:
  Netherlands