Pivotal decomposition for reliability analysis of fault tolerant control systems on unmanned aerial vehicles

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Abstract In this paper, we describe a framework to efficiently assess the reliability of fault tolerant control systems on low-cost unmanned aerial vehicles. The analysis is developed for a system consisting of a fixed number of actuators. In addition, the system includes a scheme to detect failures in individual actuators and, as a consequence, switch between different control algorithms for automatic operation of the actuators. Existing dynamic reliability analysis methods are insufficient for this class of systems because the coverage parameters for different actuator failures can be time-varying, correlated, and difficult to obtain in practice. We address these issues by combining new fault detection performance metrics with pivotal decomposition. These new metrics capture the interactions in different fault detection channels, and can be computed from stochastic models of fault detection algorithms. Our approach also decouples the high dimensional analysis problem into low dimensional sub-problems, yielding a computationally efficient analysis. Finally, we demonstrate the proposed method on a numerical example. The analysis results are also verified by Monte Carlo simulations.

Original languageEnglish (US)
Article number5291
Pages (from-to)130-141
Number of pages12
JournalReliability Engineering and System Safety
Volume140
DOIs
StatePublished - Aug 1 2015

Fingerprint

Reliability analysis
Unmanned aerial vehicles (UAV)
Actuators
Fault detection
Decomposition
Control systems
Stochastic models
Automation
Switches
Costs

Keywords

  • Certification
  • Fault detection and isolation
  • Fault tolerant control
  • Unmanned aerial vehicles

Cite this

@article{d1ac6f45368d4d43b6bc12faedc9cdce,
title = "Pivotal decomposition for reliability analysis of fault tolerant control systems on unmanned aerial vehicles",
abstract = "Abstract In this paper, we describe a framework to efficiently assess the reliability of fault tolerant control systems on low-cost unmanned aerial vehicles. The analysis is developed for a system consisting of a fixed number of actuators. In addition, the system includes a scheme to detect failures in individual actuators and, as a consequence, switch between different control algorithms for automatic operation of the actuators. Existing dynamic reliability analysis methods are insufficient for this class of systems because the coverage parameters for different actuator failures can be time-varying, correlated, and difficult to obtain in practice. We address these issues by combining new fault detection performance metrics with pivotal decomposition. These new metrics capture the interactions in different fault detection channels, and can be computed from stochastic models of fault detection algorithms. Our approach also decouples the high dimensional analysis problem into low dimensional sub-problems, yielding a computationally efficient analysis. Finally, we demonstrate the proposed method on a numerical example. The analysis results are also verified by Monte Carlo simulations.",
keywords = "Certification, Fault detection and isolation, Fault tolerant control, Unmanned aerial vehicles",
author = "Bin Hu and {Seiler Jr}, {Peter J}",
year = "2015",
month = "8",
day = "1",
doi = "10.1016/j.ress.2015.04.005",
language = "English (US)",
volume = "140",
pages = "130--141",
journal = "Reliability Engineering and System Safety",
issn = "0951-8320",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Pivotal decomposition for reliability analysis of fault tolerant control systems on unmanned aerial vehicles

AU - Hu, Bin

AU - Seiler Jr, Peter J

PY - 2015/8/1

Y1 - 2015/8/1

N2 - Abstract In this paper, we describe a framework to efficiently assess the reliability of fault tolerant control systems on low-cost unmanned aerial vehicles. The analysis is developed for a system consisting of a fixed number of actuators. In addition, the system includes a scheme to detect failures in individual actuators and, as a consequence, switch between different control algorithms for automatic operation of the actuators. Existing dynamic reliability analysis methods are insufficient for this class of systems because the coverage parameters for different actuator failures can be time-varying, correlated, and difficult to obtain in practice. We address these issues by combining new fault detection performance metrics with pivotal decomposition. These new metrics capture the interactions in different fault detection channels, and can be computed from stochastic models of fault detection algorithms. Our approach also decouples the high dimensional analysis problem into low dimensional sub-problems, yielding a computationally efficient analysis. Finally, we demonstrate the proposed method on a numerical example. The analysis results are also verified by Monte Carlo simulations.

AB - Abstract In this paper, we describe a framework to efficiently assess the reliability of fault tolerant control systems on low-cost unmanned aerial vehicles. The analysis is developed for a system consisting of a fixed number of actuators. In addition, the system includes a scheme to detect failures in individual actuators and, as a consequence, switch between different control algorithms for automatic operation of the actuators. Existing dynamic reliability analysis methods are insufficient for this class of systems because the coverage parameters for different actuator failures can be time-varying, correlated, and difficult to obtain in practice. We address these issues by combining new fault detection performance metrics with pivotal decomposition. These new metrics capture the interactions in different fault detection channels, and can be computed from stochastic models of fault detection algorithms. Our approach also decouples the high dimensional analysis problem into low dimensional sub-problems, yielding a computationally efficient analysis. Finally, we demonstrate the proposed method on a numerical example. The analysis results are also verified by Monte Carlo simulations.

KW - Certification

KW - Fault detection and isolation

KW - Fault tolerant control

KW - Unmanned aerial vehicles

UR - http://www.scopus.com/inward/record.url?scp=84928741612&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84928741612&partnerID=8YFLogxK

U2 - 10.1016/j.ress.2015.04.005

DO - 10.1016/j.ress.2015.04.005

M3 - Article

VL - 140

SP - 130

EP - 141

JO - Reliability Engineering and System Safety

JF - Reliability Engineering and System Safety

SN - 0951-8320

M1 - 5291

ER -