Review of Methods for Real-Time Loss Minimization in Induction Machines

This paper reviews and compares available techniques for minimizing power loss of induction motors. It identifies a separate category of loss minimization techniques (LMTs) - hybrid LMTs. Techniques have different convergence rates, parameter dependence, and convergence errors. Offline techniques are introduced and real-time techniques are detailed. Real-time LMTs are divided into model-based, physics-based, and hybrid categories. Application of a real-time LMT to a hybrid electric vehicle (HEV) is examined. Simulation and experimental results verify the characteristics of each category of real-time LMTs. Sensitivity of real-time LMTs to motor parameters is discussed. It is shown that techniques can be chosen based on application-specific requirements and that hybrid LMTs balance partial parameter independence and fast convergence to achieve the best overall performance.