Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.820874
Title: Statistically assessing forward-wise and inverse uncertainties in outdoor acoustics
Author: Parry, Jordan Aaron
ISNI:       0000 0004 9356 9828
Awarding Body: University of Sheffield
Current Institution: University of Sheffield
Date of Award: 2020
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
Predicting outdoor sound in uncertain conditions is a difficult task and there are limited data and statistical research which enable us to relate accurately the variations in the conditions in the propagation path to the fluctuations in the received acoustical signal. This research aimed to create better understanding of the propagation of uncertainty, using both forward and inverse case studies, in varied conditions with widely accepted engineering models, so further improvements could be made in our academic understanding and to industrial practices. The separation of the direction of uncertainty allows for more focus to be focused on each given condition. Firstly, the forward problem is approached by simplifying the model used and conditions in present, to better understand the statistical behaviour across evolving parameter uncertainties. A further study, inspired by current acoustical standards, evaluated whether improvements to data capture could be made by manipulating the physical way the data was obtained, in the presence of varying parameter uncertainties. The inverse problem was investigated for a very specific application of small arms fire, yet the methodology was expanded to show how powerful computationally cheap statistical methods can be used in investigating parameter interactions under given uncertainties, while also accurately inverting the desired parameters. Investigations have proved successful in characterising, in general and for specific scenarios, the foundational uncertainties in outdoor sound propagation. Methods have been presented that allow for simple yet powerful study into the statistical behaviours of a wide range of outdoor sound propagation problems. Characterising uncertain acoustic data using statistical representations serves to be extremely beneficial, while a physical two-microphone method is shown to be theoretically efficient in negating a large proportion of the uncertainty present, while capturing acoustical data known to be useful for source localisation and characterisation. It is also shown in which direction research should be established in relation to military applications, after showing efficient ways in which computational models be applied to invert important parameters from readily obtainable data.
Supervisor: Horoshenkov, Kirill Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.820874  DOI: Not available
Share: