Analytical and experimental studies of a hybridised synchronous reluctance machine with a variable X d /X q ratio characteristic is the subject of this study. The machine comprises a round rotor and a salient pole machine element that is mechanically coupled together and integrally wound. There are two sets of poly-phase windings in each stator. The windings of one of the sets are transposed between the two sections of the machine. Either set of the windings may be connected to the supply while the other feeds a balanced capacitance load. It is shown that by tuning of the capacitance load that the X d /X q ratio varies theoretically from zero to infinity at very good power factors. The machine characteristics were verified using the generalised two-axis theory and validated by experimentation. The experimental and analytical results obtained show good agreement. Saturation was investigated by using the variation of the d-axis inductance with machine loading. A set of rotor windings may be introduced for the purpose that the machine be self-starting and self-synchronising as a motor. The rotor field windings when fed with dc, will make the machine also capable of operation as a stand-alone salient-pole generator. For this purpose, it was shown that for a very high saliency ratio, reluctance power can be considerably higher than the excitation power.

The influence of rotor cage on the performance of an autonomous selfexcited reluctance generator supplying loads of different power factors is presented. It is seen that for the same excitation capacitance, terminal load connection, and speed, the self-excited reluctance generator with a cage exhibits a better ability to preserve the voltage waveshape following a sudden addition or removal of load than does a cageless self-excited reluctance generator. This is because during the transient disturbance, the rotor cage supports some DC current in the d -axis, making it act as a field winding. The solution of steady-state model equations formulated directly from the dynamic model are solved using the MATLAB ® (The MathWorks, Natick, Massachusetts, USA) optimization technique 'fsolve.' The cageless self-excited reluctance generator is only capable of yielding less than half the rated power, while the self-excited reluctance generator with a cage is able to produce more than twothirds of the rated power. This is due to the inability of the cageless machine to excite at capacitance values high enough to circulate the rated current in the machine windings.

No abstract

No abstract

No abstract