Please use this identifier to cite or link to this item:
|Title:||Towards model-based system engineering for autonomous cyber-physical systems|
|Abstract:||Cyber-Physical Systems (CPSs) such as smart grids, autonomous automobiles, and medical monitoring systems are becoming increasingly common. Autonomy (i.e., the ability of a system to achieve high-level objective(s) with limited or no external assistance) is integral to future CPS applications, but it comes with non-trivial design challenges. It is especially challenging to validate Autonomous CPSs (ACPSs) that adapt their behaviour to dynamically changing environments. Model-Based Systems Engineering (MBSE) has been recognised as a way to manage the complexity of CPSs as it supports system design at a high level of abstraction, allowing the analysis of designs and identification of key features, including defects, before they reach implementation. However, existing MBSE techniques and frameworks are targeted for generic CPSs that do not involve the challenges found in ACPSs. The aim of our research is to define an MBSE framework that supports the architecting of ACPSs from requirements to validation. As the first step to realising this aim, we propose a MBSE ontology for ACPS. This provides a holistic view on the architecture of ACPS by combining several concepts such as MBSE, CPS and Autonomous Systems into one coherent ontology. Our aim is realised as an Architectural Framework for ACPS (ACPSAF), built on top of the ACPS Ontology. ACPSAF is an encapsulation of a minimum set of practices and requirements for artefacts that describe the ACPS architecture. ACPSAF proposes five Perspectives to capture the architecture of ACPSs across the whole MBSE life cycle. In order to validate the framework, two case studies are carried out where ACPSs are designed using ACPSAF. The results suggest that ACPSAF provides information in the requirements, architecture and analysis of ACPS that can be used with existing ACPS specific techniques such as runtime monitoring approach and automated test case generation to gain confidence in the dependability of ACPSs.|
|Appears in Collections:||School of Computing|
Files in This Item:
|Salim H 2022.pdf||3.88 MB||Adobe PDF||View/Open|
|dspacelicence.pdf||43.82 kB||Adobe PDF||View/Open|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.