An optimization-aided approach to parametrize scalable models of induction machines in speed variable drives

Abstract

The accurate and time-saving prediction of essential machine variables (output power, torque, and efficiency) is crucial for manufacturers offering a wide power range of induction machines. Many motor variants are typically produced by axially scaling and rewinding the machine. Rescaling procedures of electromagnetic models of induction machines are in everyday use and well known. However, while a high accuracy can be achieved by rescaling in theory, more significant deviations between simulated and measured output parameters of the realized scaled device occur in praxis. These deviations can mainly be attributed to the faulty separation of effects in the distinct machine components, such as the rotor, stator, and bearings. This paper introduces an optimization-aided modeling approach based on the induction machine’s simple equivalent circuit representation. The method is validated by measurement data obtained from many induction machines with various axial lengths and winding configurations.