The use of Bayesian networks in quantifying evidence used to aid the assurance process
Date
2018
Authors
Simmonds, S.
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Conference paper
Citation
SETE 2019: Systems Engineering Test and Evaluation Conference: Unlocking the Future Through Systems Engineering, 2018, pp.138-158
Statement of Responsibility
Conference Name
SETE 2019: Systems Engineering Test and Evaluation Conference (30 Apr 2019 : Canberra, Australia)
Abstract
The Goal Structuring Notation can be used to support argumentation of particular claims, including system safety and system assurance claims. While argument claims need to be supported by evidence, the strength of evidence supporting a claim is often not systematically considered. Restructuring the claim as a Bayesian Network supports systematically quantifying the likelihood of the veracity of the evidence supporting the claim such that the risk can be managed through the system life, in particular in acquisition and modifications during sustainment. By using an Object based Bayesian Network model it is possible to structure a network model to consider the availability and the likelihood of the veracity of relevant evidence for a system. The use of the Object based model with linear and Boolean aggregator functions allows this model to be rapidly expanded to consider multiple systems of systems. The application of this method to a Technical Integrity assurance argument is shown. Similarly, the technique can be used to support a safety argument.
School/Discipline
Dissertation Note
Provenance
Description
Access Status
Rights
Copyright 2019 Engineers Australia
Access Condition Notes: Accepted manuscript available on Open Access