Skip to main navigation Skip to search Skip to main content

Safety Flight Control for a Quadrotor UAV Using Differential Flatness and Dual-Loop Observers

  • Hunan University
  • Beihang University
  • Concordia University

Research output: Contribution to journalArticlepeer-review

Abstract

Focusing on a quadrotor unmanned aerial vehicle (UAV), the presence of unintended actuator faults and external disturbances increases the risk of a crash. Although plenty of efforts have been devoted, how to integrate capability analysis into safety control design is still an open issue. This article presents the design of a safety control system for quadrotor UAVs. Firstly, a system capability analysis method based on a differential flatness algorithm is developed, so that the derivatives (i.e., velocity, jerk, and snap) of flight trajectory can be formulated as flat variables. A tradeoff between system capability and permissible flight maneuverability is made to avoid actuator saturation. Secondly, dual-loop nonlinear disturbance observers are exploited to identify the actuator faults and disturbances, which can thereby be handled by a cascade control scheme. Moreover, the trajectory is regenerated online at the expense of degraded flight maneuverability or even emergency landing, in view of the remaining actuator control authority. When comparing to the existing methods, the gap among safety control, trajectory generation, and system capability analysis is bridged to ameliorate practical flight safety. Finally, flight tests are carried out to demonstrate the unique merits of the proposed system.

Original languageEnglish
Pages (from-to)13326-13336
Number of pages11
JournalIEEE Transactions on Industrial Electronics
Volume69
Issue number12
DOIs
StatePublished - 1 Dec 2022

Keywords

  • Differential flatness (DF)
  • dual-loop observers
  • system capability analysis
  • trajectory regeneration

Fingerprint

Dive into the research topics of 'Safety Flight Control for a Quadrotor UAV Using Differential Flatness and Dual-Loop Observers'. Together they form a unique fingerprint.

Cite this