Controlling the position and attitude of a helicopter hovering in
the presence of atmospheric turbulence is a difficult task which demands
considerable pilot work-load which becomes even more difficult'when a
load is suspended from the helicopter, because the oscillations of the
load aggravate the situation. Tasks that require a suspended load to be
kept fixed relative to a point in space, while the helicopter remains at
hover, are extremely difficult to achieve.
Several load-positioning systems exist but provide inadequate
solutions to the problem. A brief account of such systems and their
limitations is given before describing the automatic hovering control
system proposed in this thesis. It causes appropriate motion of the
helicopter to achieve the desired stationarity of the load. The
techniques of modern control theory were employed to design this optimal
controller. Digital simulation was used for testing the response of the
resulting optimal system.
The mathematical model of two connected rigid bodies moving in
space (representing the helicopter and suspended load) is described in
detail. Several combinations of cable length-load weight were chosen
and in each case the response of the closed-loop system was investigated.
It was found that considerable reduction of the oscillations of the load
can be achieved when suitable cable arrangements are used. The use of
winch control of lateral displacement of the load also improves the
lateral response of the entire system. An augmented mathematical model was used which included both the dynamics of the control actuators and
the models representing atmospheric turbulence and sensor noise. Since
many of the state variables of the system cannot be physically measured,
it is obvious that only limited information on the state of the system
would be available for processing by such a controller. Therefore two
solutions to the problem were considered:
(i) the use of a state estimator to provide to the controller
the lost feedback information;
and (ii) the use of an output regulator which takes into account
the fact that limited feedback information is available.
The responses of the closed-loop systems using each of these solutions
were investigated and compared. The numerical problems encountered in
this design are analysed and some means .of overcoming them are suggested.
Finally, the best combination of cable arrangement and controller
is described with reference to several important factors such as system
simplicity and performance.
A Doctoral Thesis. Submitted in partial fulfilment of the requirements for the award of Doctor of Philosophy of Loughborough University.