I. Introduction
Vector control algorithms designed for induction machines (IMs) are based on the respective dynamic mathematical models. For practical reasons, some of these models, and hence the corresponding vector controllers, ignore certain phenomena which are known to occur in the actual IM, such as magnetic saturation of the iron core, temperature variations of the winding resistances, losses in the iron core, stray load losses (SLLs), etc. On the other hand, the accuracy of IM vector control mainly depends on the accuracy with which the utilized mathematical model describes the actual IM behavior. Therefore, in cases when there is a mismatch between the IM model and the actual IM, a detuning takes place in terms of misalignment between the actual and reference flux vector (i.e., flux angle error) or non-unity ratio of actual to reference flux vector magnitudes (i.e., flux magnitude error).