Instrumentation and Control of Unmanned Air Vehicles

Research output: Book/ReportPh.D. thesis – Annual report year: 2003Research

Documents

View graph of relations

This thesis treats a number of instrumentation and control problems related to autonomous Unmanned Aerial Vehicles (UAV’s). Autonomous Micro Air Vehicles (MAV’s) are of special interest to the Author. These are characterised by their small size, typically below 10kg takeoff weight. Due to their small size, normal avionics are not suited for MAV’s. Instead it is more appropriate to use standard model airplane components and actuators. This has the added benefit of reducing the vehicle cost. However this also means that the vehicle designer has to characterise and design many of the instruments and actuators used for MAV’s.

The first part of this thesis concentrates on obtaining an aerodynamic model for a canard configuration fixed wing UAV. Particular emphasis is placed on treatment of uncertainties in the model and the resulting influence on the UAV dynamics. A model of an electric propulsion system is also proposed, based partly on propeller characteristics obtained by comparing the geometry of the propeller with that of a propeller with known characteristics.

Model airplane actuators are a logical choice for MAV’s because of there availability, price and performance. It is however difficult to obtain published data concerning the dynamics of these actuators. For these reasons a part of this thesis treats the procedure used in experimentally identifying a dynamic model of model airplane actuators. It is shown that a particular make of actuators employ a proportional-derivative bang-bang controller scheme. As a result of this observation, a feedback linearization scheme is proposed and simulated for this type of actuator.

Different lateral guidance strategies are discussed based on the assumption that the desired flight path of the UAV is defined by a number of “waypoints”. It is shown that a “moving point” guidance strategy has certain advantages with respect to autopilot implementation and smooth transition from one heading to another in the vicinity of a waypoint.

The most critical flight phase with respect to guidance and navigation accuracy is the approach and landing. In order to accomplish an autonomous landing it is important to be able to determine the position of the UAV with great accuracy and reliability. The only practical system for accurate navigation at the present, which does not require expensive ground based equipment, is the satellite based Global Positioning System, commonly known as GPS. However this alone does not have sufficient accuracy for the task. By using a differential positioning approach involving a ground based GPS receiver at a known location, it is possible to construct a Differential GPS (DGPS). Such a system has been implemented and studied in detail using a pair of commercially available receivers. It is shown through analysis of experimental data that the most significant error source in DGPS systems is multipath, caused by reflections of the signal from objects in the vicinity of the receiver antennas. Furthermore it is shown that these errors can be correlated with the receiver Signal to Noise Ratio (SNR). Using this information a kinematic Kalman filter is proposed for filtering the raw measurement data. “Simulation” of this filter using actual measurement data shows that a significant improvement in positioning accuracy is obtained.

Finally some issues relating to the design of an Inertial Navigation System (INS) and realtime synchronized instruments are discussed.
Original languageEnglish
Place of PublicationKgs. Lyngby
PublisherTechnical University of Denmark (DTU)
Number of pages258
ISBN (Print)87-91184-15-0
Publication statusPublished - May 2003
Download as:
Download as PDF
Select render style:
APAAuthorCBE/CSEHarvardMLAStandardVancouverShortLong
PDF
Download as HTML
Select render style:
APAAuthorCBE/CSEHarvardMLAStandardVancouverShortLong
HTML
Download as Word
Select render style:
APAAuthorCBE/CSEHarvardMLAStandardVancouverShortLong
Word

Download statistics

No data available

ID: 5641429