Controllability of a portable observational unmanned lightweight airship

P.L. Dalen, van, H. Nijmeijer (Supervisor), C. Bil (Supervisor), I. Wang (Supervisor)

Research output: Book/ReportReportAcademic

1 Downloads (Pure)

Abstract

Four students of InHolland University Delft have started the mechanical design of an autonomous portable observational unmanned lightweight airship (POULA). Their main objectives where to make the airship as light and simple as possible. This has resulted in an airship with only two propellers which can be independently rotated around their common axis. Little or no attention is paid regarding the dynamics and controllability of the airship. Because it was questionable if the strongly nonlinear and underactuated airship would be controllable this had to be investigated first. This investigation is the main subject of the research presented in this report. Because no mathematical model was available the first part of the report concerns the derivation of the kinematic and dynamic model. Initially it was the intention to design a three dimensional nonlinear control system and therefore the model obtained is three dimensional as well. However, this appeared to be a bridge too far for this research. Instead the model is simplified to two dimensions. The purpose of the airship will be to monitor a certain indoor area like a gym with an onboard camera. After the airship has been released in the air the first task of the control system is to let the airship ascent to a desired altitude. After this task is accomplished the airship is ready to track a certain (time-varying) trajectory to surveil the area. A constant altitude is desired during this operation. First the model has been simplified to a two dimensional model in the longitudinal plane. A controller is designed to let the airship ascent to a certain starting position. Because the model is still underactuated it is linearized. Linear state feedback controllers are obtained using the pole placement method. Simulations show that the system can be stabilized and that the controller is able to let the errors converge to zero. When the airship has reached its desired height it is ready to start his surveilling task. Because the desired altitude is constant the model is considered in the horizontal plane. The main assumption is that the airship is neutrally buoyant, i.e. the altitude remains constant during the operation without providing a propeller force upwards. Because of the underactuated design and the non-negligible nonlinear terms it is a very hard control problem. A lot of literature is available how to deal with this problem. Different strategies with different constraints have been proposed. Considering these different constraints and the objectives for the airship there has been chosen to follow the strategy used for controlling an underwater vessel. The method presented in the paper has been adjusted and modified to make it suitable for the airship. First a trajectory planning algorithm is proposed specifying all the necessary reference states. These states are used as an input for the nonlinear controller which is based on Lyapunov techniques. Practical exponential stability is proven. Simulations show that the airship can be stabilized and is able to track a time-varying sinusoidal trajectory with and without model errors and time-varying disturbances. In future research the three dimensional model will have to be used to obtain a three dimensional control system. Literature is available which describe strategies to design this control system. Based on the two dimensional results presented in this report and the available literature it does seem possible to control the airship POULA. It will be possible to surveil indoor areas using the onboard camera.
Original languageEnglish
Place of PublicationEindhoven
PublisherEindhoven University of Technology
Publication statusPublished - 2011

Publication series

NameD&C
Volume2011.019

Bibliographical note

Traineeship report. - DC 2011.019

Fingerprint Dive into the research topics of 'Controllability of a portable observational unmanned lightweight airship'. Together they form a unique fingerprint.

Cite this