Flux vector control of asynchronous machines.

The present work introduces the dynamic induction machine model using the concept of space vector notation, and shows the fundamentals of torque production to be related to the current and flux space vector interaction. By theoretical analysis the transient response of the current and voltage fed machine is determined in terms of the machine eigenvalues, and this is confirmed using simulation techniques. The concept of torque control using the current and flux space vectors' magnitude and phase is introduced. It is shown that maintaining the flux space vector's magnitude constant is a sensible approach and that transient free torque responses are possible by controlling the current space vector magnitude and phase with respect to the flux space vector, which is termed vector control. Previously only magnitude control was realised, and this is termed scalar control. It is shown that the simplest practical vector controllers work in the rotating rotor flux space vector reference frame, generated from a dynamic machine model using current and speed/position measurements from the real machine. The major disadvantage of the parameter sensitivity of this approach is examined. To complement the theoretical analysis a practical vector controller was to be built. This was designed around a new DSP micro-controller (TMS320C14) and featured a 10 kHz sampling rate. A fully instrumented 7.5 kW test rig was also developed with the induction machine driven from a modified 1 kHz inverter. The measurement of high performance shaft torque is investigated, to allow comparisons between the theoretical and practical results. The practical work centred on the comparison of the dynamic machine model with measurements made on the real machine (torque and flux) and with the theory. These showed excellent steady state performance but the dynamic response was disappointing. The implications of the model not representing the real machine transiently are investigated.