Detail publikace

Motor Failure Detection for Multicopters

BARÁNEK, R.

Originální název

Motor Failure Detection for Multicopters

Anglický název

Motor Failure Detection for Multicopters

Jazyk

en

Originální abstrakt

The contribution is focused on the problem of motor failure detection for multicopters using sensors usually used for state estimation. Such an algorithm is an essential part of the safety system which would mitigate the consequences of single motor failure of multicopter. The detection algorithm is based on set of Kalman filters for state estimation. Each Kalman filter has different prediction model (each models a different motor failure). The magnitude of corrections applied in the update step of Kalman filter is used as a measure of model correspondence. The algorithm was tested on simulated data for two different scenarios and shows sufficient performance in both cases.

Anglický abstrakt

The contribution is focused on the problem of motor failure detection for multicopters using sensors usually used for state estimation. Such an algorithm is an essential part of the safety system which would mitigate the consequences of single motor failure of multicopter. The detection algorithm is based on set of Kalman filters for state estimation. Each Kalman filter has different prediction model (each models a different motor failure). The magnitude of corrections applied in the update step of Kalman filter is used as a measure of model correspondence. The algorithm was tested on simulated data for two different scenarios and shows sufficient performance in both cases.

Dokumenty

BibTex


@inproceedings{BUT109143,
  author="Radek {Baránek}",
  title="Motor Failure Detection for Multicopters",
  annote="The contribution is focused on the problem of motor failure detection for multicopters using sensors usually used for state estimation. Such an algorithm is an essential part of the safety system which would mitigate the consequences of single motor failure of multicopter. The detection algorithm is based on set of Kalman filters for state estimation. Each Kalman filter has different prediction model (each models a different motor failure). The magnitude of corrections applied in the update step of Kalman filter is used as a measure of model correspondence. The algorithm was tested on simulated data for two different scenarios and shows sufficient performance in both cases.",
  address="LITERA Brno",
  booktitle="Proceedings Of The 20th Conference Student EEICT 2014 Volume 3",
  chapter="109143",
  howpublished="print",
  institution="LITERA Brno",
  year="2014",
  month="april",
  pages="52--56",
  publisher="LITERA Brno",
  type="conference paper"
}