Controller Design for a Remotely Controlled Airship
Akshat Goyal,
D.Deg,
06,
58 pp.
Department of
Aerospace Engineering
Indian Institute of Technology Bombay,
Powai, Mumbai 400 076.
Supervisor(s):
R. K. Pant, R. Gudi
An airship is a lighter than air vehicle that derives
its lift from the principal of buoyancy. It is considered to be one of
the most capable vehicles for long endurance flights at low speeds.
Airships find applications in various areas like advertisements,
surveillance, scientific explorations, telecommunications etc. An
airship is a highly non-linear system and modeling its dynamics poses
several challenges. Terms like added mass and inertia and significant
effect of Coriolis forces adds to the complexity of modeling and
control system design.
This report covers literature survey on modeling of airship dynamics,
stability, control system design and use of neural networks for
control. Non-linear equations of motion are presented and an earlier
attempt at modeling the dynamics is discussed. After discussing the
problems with the available model a new model is presented. A control
system design scheme involving the linear model of the airship is
discussed and a methodology for creating a linear model from the
non-linear model is presented. PI and PID controller design using
Direct Synthesis method is discussed and the designed controllers are
applied to the non-linear model. The developed controllers are then
tested in turbulent conditions. Use of neural networks for controls is
discussed. A general neural architecture is presented and a training
methodology is described. A prediction model for the airship model is
created in neural network and validated with a suitable test data.