Range Safety Application of Kernel Density Estimation
Abstract
This report describes the kernel density estimation technique and its application to range safety applications. The kernel density estimation technique is shown to be suitable for developing probabilistic risk assessments from ground impact data generated for guided weapon systems via Monte Carlo simulations. An advantage of this technique is that it can be used to predict the probability density function for minimal simulated ground impacts with apparently random distribution. Several techniques have been proposed to ameliorate the identified limitations of the kernel density estimation technique, including a covariant form for two-dimensional data. Analysis of the available simulated guided weapon ground impact data has identified that around six hundred impact points are sufficient for generating a probability distribution.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2010
- Accession Number
- ADA522670
Entities
People
- Duncan Fletcher
- Gary Glonek
- Jeremy Mcmahon
- Michael Jokic
- Michael Rumsewicz
- Oleg Mazonka
- Timothy Staniford
Organizations
- Defence Science and Technology Group