Modelling And Control Of An Air Conditioning System With An Emphasis On Cooling Coil Behaviour

A review of the literature on the control of air
conditioning systems has been undertaken. It shows that
the behaviour of sensors and the relationship between
measurements and bulk conditions cause significant
difficulties. A simple air cooling and dehumidifying
experimental plant and a non-linear thermodynamic
mathematical model have been simultaneously developed
to investigate the control of air coolers with an
emphasis on the cooling and dehumidifying process.
The radio frequency interference encountered in using
the frequency inverter for varying the compressor speed
presented major problems. It was suppressed by using
shielding, filtering and common grounding techniques
with high accuracy.
The equations describing the behaviour of the DxX
evaporator in the model are unique. The mathematical
model has been validated and simplified using classical
control theory. Transfer functions derived during the
validation processes were simple mathematical models
and are separately applicable. The techniques employed
are shown to be effective methods for model validation.
Two types of intelligent controller have been derived.
For situations where the bulk air conditions can be
measured, a closed loop intelligent controller with a
combination of simple PID control algorithms’ and
linguistic rules has been developed. For situations in
which the air bulk state is difficult to measure, an
open loop intelligent controller was derived. It
combined some key parameter measurements in the cooling
system and a few mathematical statements with some
linguistic rules. Experiments on the controllers showed
that both control accuracy and stability were achieved.
Two new control concepts for variable speed compressor
systems were investigated with the validated
mathematical model. Promising results from the model
simulation showed that research and development would
be worthwhile.