Abstract
M.Ing.
The synchronous reluctance machine is a reasonably unknown type of machine. Recent developments
have made it possible to deliver more power than an induction machine of similar size.
It must, however, be noted that to accomplish this, the rotor profile design becomes complex
and also expensive to manufacture. The aim of this project is to explore the possibility of
controlling a synchronous reluctance machine `sensorlessly' (with no speed or position feedback
sensors). The specific synchronous reluctance machine was made from an induction machine
by machining a profile into the rotor cage of the induction machine. The required profile was
designed using finite element analysis of the magnetic structure of the machine. A model of the
machine was developed through using the direct and quadrature axis approach. With a model
of the machine an estimator could be designed. This estimator was used to calculate the load
angle of the machine by measuring the phase voltages and line currents and then estimating
from these measurements the actual load angle. The estimator was designed for steady state
conditions and gave sufficiently accurate values for the load angle under these conditions. The
estimator was used as feedback in the torque control of the machine and it was shown that better
results could be achieved with a better rotor profile. The partial cage of the induction machine
present in the rotor made the machine more robust to changes in load. More research can be
done on the modelling of the partial cage as well as the influence it has on the synchronous
reluctance machine. Torque control and speed control were achieved with step changes in load
of up to eighty percent of rated value. The performance of the machine was poor compared to
an induction machine, but should improve with improvement in rotor profile design.