An Induction Motor Model for Virtual Drive Simulation Systems

Abstract

In order to improve the control strategies and in turn increase the efficiency of the system, it is necessary to use accurate machine models. Finite element method (FEM) based machine models are considered sufficiently accurate. However, they cannot run in real time due to their computational costs. Traditional analytical models containing only constant-valued parameters can run in real time; however, they cannot model the machine power losses accurately enough.

This thesis proposes an analytical induction motor model. The electromagnetic (EM) loss calculation has been improved compared to the previous models by separating the rotor iron loss from the stator no load iron loss, by modelling the stray load loss (SLL) in the stator iron and by considering the loss due to the skin effect in the rotor bars. In the case of the 45 kW test motor and compared to the FEM computations, the error in the total EM loss is less than 0.1 % in the no load condition, and less than 0.6 % for the rated loading. The proposed model is suitable for time-domain simulations and it can run in real time. Therefore, it can be integrated with an electric drive simulation, which can be used in a variety of applications. The model has been implemented in the MATLAB Simulink environment.

FEM-based program FC-SMEK was used for computing the machine losses and for validating the results. The nonlinear least squares data fitting method lsqnonlin was used for identifying the parameters of the proposed model.