Control & Autonomy
- Communication and Control Designs for Reconfigurable Systems
- Multi-Platform Phased Mission Prognostics Using Reliability Modelling Techniques
- A Robust Fault-tolerant Tracking Scheme
Communication and Control Designs for Reconfigurable Systems
A. Murgu, I. Postlethwaite, C. Edwards and D.-W. Gu
This paper provides an overview of existing communication and control protocols that can be invoked for architectures used for implementing reconfigurable systems of systems working in NEC environments. Distributed, supervisory, coordinated control concepts and their roles are discussed in relation to the resource allocation problem subject to time constraints associated with reconfigurable systems.
Download the full paper as a pdf file here
Multi-Platform Phased Mission Prognostics Using Reliability Modelling Techniques
D.R. Prescott, J.D. Andrews
This paper examines multi-platform phased missions, which involve a number of platforms working together to achieve a common objective. A prognostics methodology, based on calculating mission failure probabilities, is discussed and a Binary Decision Diagram (BDD) technique that enables these probabilities to be calculated is demonstrated. An example system is considered and is used to show how parts of the application of the methodology may be implemented offline and others online, with the goal of producing mission failure probabilities more quickly when they are required. It is also demonstrated how effort applied to the offline BDD construction parts of the methodology will result in gains in speed of the online BDD quantification processes.
Download the full paper as a pdf file here
A Robust Fault-tolerant Tracking Scheme
I. Postlethwaite, D.-W. Gu, Y. Kim, K. Natesan, M. Kothari, N. Khan and R. Omar
The increasing use of autonomous vehicles (AV) in target tracking and formation keeping tasks has introduced the need to improve sensor accuracy and performance. Detection of sensor faults and corrective action is therefore becoming an integral part of modern AVs. Sensors are vulnerable to internal faults and adverse external conditions such as weather and physical obstacles. In this paper, a multi-agent target tracking problem is considered. We are interested in decentralized target tracking combined with fault detection schemes. We present a fault detection and data fusion algorithm and use a group of robots to show how the algorithm is capable of detecting faulty sensors and keeping the robots in a pre-defined formation to track a target robot.
Download the full paper as a pdf file here
