Evaluation of iron loss influence on speed estimation in sensorless MRAC-based field-oriented induction machines

Rotor speed estimation in sensorless field-oriented control of induction machines is usually performed on the basis of a mathematical model of an induction machine. An estimate of speed is obtained utilising one of the control techniques and its accuracy is affected by parameter variations. When the iron loss is neglected in rite mathematical model, inaccuracy in speed estimation results due to iron loss existence in the machine. This paper evaluates the impact of iron loss on speed estimation for common model reference adaptive control (MRAC) schemes, based on rotor flux, back e.m.f. and reactive power. It is shown that the existence of iron loss lends to a typical speed estimation error of 2 min(-1) to 3 min(-1), which is comparable with errors caused by other detuning effects. A modified sensorless MRAC-based scheme, that is applicable to both rotor flux and back e.m.f: methods, is then proposed. It is verified that the novel scheme enables good compensation of iron loss, yielding a very small speed estimation error of less than 0.5 min(-1).