Performance evaluation of neural network based approaches for airspeed sensor failure accommodation on a small UAV

Srikanth Gururajan, Mario L. Fravolini, Haiyang Chao, Matthew Rhudy, Marcello R. Napolitano

Research output: Chapter in Book/Report/Conference proceedingConference contribution

8 Citations (Scopus)

Abstract

Traditional approaches to sensor fault tolerance for flight control systems have been based on triple or quadruple physical redundancy. However, recent events have highlighted the criticality of "common mode" failures on the Air Data System (ADS). In fact, since the parameters of flight control laws are typically scheduled as a function of airspeed, incorrect readings from the ADS can lead to potentially catastrophic conditions. In this paper, we describe the evaluation of an analytical redundancy-based approach to the problem of Sensor Failure Accommodation following simulated failures on the ADS of a research UAV, using Artificial Neural Networks (ANNs). Specifically, two different neural networks are evaluated - the Extended Minimal Resource Allocating Network and a Multilayer Feedforward NN. These neural networks are trained and validated using experimental flight data from the WVU YF-22 research aircraft which was designed, manufactured, instrumented, and flight tested by researchers at the Flight Control Systems Laboratory at West Virginia University. The performance of the two approaches is evaluated in terms of the statistics of the tracking error in the estimation of the airspeed, as compared to actual measurements from the ADS, operating under nominal conditions.

Original languageEnglish (US)
Title of host publication2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings
Pages603-608
Number of pages6
DOIs
StatePublished - Oct 15 2013
Event2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Platanias-Chania, Crete, Greece
Duration: Jun 25 2013Jun 28 2013

Publication series

Name2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings

Other

Other2013 21st Mediterranean Conference on Control and Automation, MED 2013
CountryGreece
CityPlatanias-Chania, Crete
Period6/25/136/28/13

Fingerprint

Unmanned aerial vehicles (UAV)
Neural networks
Flight control systems
Sensors
Air
Redundancy
Research aircraft
Fault tolerance
Failure modes
Multilayers
Statistics

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Control and Systems Engineering

Cite this

Gururajan, S., Fravolini, M. L., Chao, H., Rhudy, M., & Napolitano, M. R. (2013). Performance evaluation of neural network based approaches for airspeed sensor failure accommodation on a small UAV. In 2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings (pp. 603-608). [6608784] (2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings). https://doi.org/10.1109/MED.2013.6608784
Gururajan, Srikanth ; Fravolini, Mario L. ; Chao, Haiyang ; Rhudy, Matthew ; Napolitano, Marcello R. / Performance evaluation of neural network based approaches for airspeed sensor failure accommodation on a small UAV. 2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings. 2013. pp. 603-608 (2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings).
@inproceedings{0b1f877375264928af9cd03f4ee1ab7a,
title = "Performance evaluation of neural network based approaches for airspeed sensor failure accommodation on a small UAV",
abstract = "Traditional approaches to sensor fault tolerance for flight control systems have been based on triple or quadruple physical redundancy. However, recent events have highlighted the criticality of {"}common mode{"} failures on the Air Data System (ADS). In fact, since the parameters of flight control laws are typically scheduled as a function of airspeed, incorrect readings from the ADS can lead to potentially catastrophic conditions. In this paper, we describe the evaluation of an analytical redundancy-based approach to the problem of Sensor Failure Accommodation following simulated failures on the ADS of a research UAV, using Artificial Neural Networks (ANNs). Specifically, two different neural networks are evaluated - the Extended Minimal Resource Allocating Network and a Multilayer Feedforward NN. These neural networks are trained and validated using experimental flight data from the WVU YF-22 research aircraft which was designed, manufactured, instrumented, and flight tested by researchers at the Flight Control Systems Laboratory at West Virginia University. The performance of the two approaches is evaluated in terms of the statistics of the tracking error in the estimation of the airspeed, as compared to actual measurements from the ADS, operating under nominal conditions.",
author = "Srikanth Gururajan and Fravolini, {Mario L.} and Haiyang Chao and Matthew Rhudy and Napolitano, {Marcello R.}",
year = "2013",
month = "10",
day = "15",
doi = "10.1109/MED.2013.6608784",
language = "English (US)",
isbn = "9781479909971",
series = "2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings",
pages = "603--608",
booktitle = "2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings",

}

Gururajan, S, Fravolini, ML, Chao, H, Rhudy, M & Napolitano, MR 2013, Performance evaluation of neural network based approaches for airspeed sensor failure accommodation on a small UAV. in 2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings., 6608784, 2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings, pp. 603-608, 2013 21st Mediterranean Conference on Control and Automation, MED 2013, Platanias-Chania, Crete, Greece, 6/25/13. https://doi.org/10.1109/MED.2013.6608784

Performance evaluation of neural network based approaches for airspeed sensor failure accommodation on a small UAV. / Gururajan, Srikanth; Fravolini, Mario L.; Chao, Haiyang; Rhudy, Matthew; Napolitano, Marcello R.

2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings. 2013. p. 603-608 6608784 (2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Performance evaluation of neural network based approaches for airspeed sensor failure accommodation on a small UAV

AU - Gururajan, Srikanth

AU - Fravolini, Mario L.

AU - Chao, Haiyang

AU - Rhudy, Matthew

AU - Napolitano, Marcello R.

PY - 2013/10/15

Y1 - 2013/10/15

N2 - Traditional approaches to sensor fault tolerance for flight control systems have been based on triple or quadruple physical redundancy. However, recent events have highlighted the criticality of "common mode" failures on the Air Data System (ADS). In fact, since the parameters of flight control laws are typically scheduled as a function of airspeed, incorrect readings from the ADS can lead to potentially catastrophic conditions. In this paper, we describe the evaluation of an analytical redundancy-based approach to the problem of Sensor Failure Accommodation following simulated failures on the ADS of a research UAV, using Artificial Neural Networks (ANNs). Specifically, two different neural networks are evaluated - the Extended Minimal Resource Allocating Network and a Multilayer Feedforward NN. These neural networks are trained and validated using experimental flight data from the WVU YF-22 research aircraft which was designed, manufactured, instrumented, and flight tested by researchers at the Flight Control Systems Laboratory at West Virginia University. The performance of the two approaches is evaluated in terms of the statistics of the tracking error in the estimation of the airspeed, as compared to actual measurements from the ADS, operating under nominal conditions.

AB - Traditional approaches to sensor fault tolerance for flight control systems have been based on triple or quadruple physical redundancy. However, recent events have highlighted the criticality of "common mode" failures on the Air Data System (ADS). In fact, since the parameters of flight control laws are typically scheduled as a function of airspeed, incorrect readings from the ADS can lead to potentially catastrophic conditions. In this paper, we describe the evaluation of an analytical redundancy-based approach to the problem of Sensor Failure Accommodation following simulated failures on the ADS of a research UAV, using Artificial Neural Networks (ANNs). Specifically, two different neural networks are evaluated - the Extended Minimal Resource Allocating Network and a Multilayer Feedforward NN. These neural networks are trained and validated using experimental flight data from the WVU YF-22 research aircraft which was designed, manufactured, instrumented, and flight tested by researchers at the Flight Control Systems Laboratory at West Virginia University. The performance of the two approaches is evaluated in terms of the statistics of the tracking error in the estimation of the airspeed, as compared to actual measurements from the ADS, operating under nominal conditions.

UR - http://www.scopus.com/inward/record.url?scp=84885224908&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84885224908&partnerID=8YFLogxK

U2 - 10.1109/MED.2013.6608784

DO - 10.1109/MED.2013.6608784

M3 - Conference contribution

AN - SCOPUS:84885224908

SN - 9781479909971

T3 - 2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings

SP - 603

EP - 608

BT - 2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings

ER -

Gururajan S, Fravolini ML, Chao H, Rhudy M, Napolitano MR. Performance evaluation of neural network based approaches for airspeed sensor failure accommodation on a small UAV. In 2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings. 2013. p. 603-608. 6608784. (2013 21st Mediterranean Conference on Control and Automation, MED 2013 - Conference Proceedings). https://doi.org/10.1109/MED.2013.6608784