Efficiency optimizing control of induction motor using natural variables

This paper presents a new approach of optimizing the efficiency of induction-motor drives through minimizing the copper and core losses. The induction-machine model, which accounts for the varying core-loss resistance and saturation dependent magnetizing inductance, uses natural and reference frame independent quantities as state variables. Utilization of the non-linear geometric control methodology of input-output linearization with decoupling permits the implementation of the control in the stationary reference frame. This approach eliminates the need of synchronous reference transformation and flux alignment required in classical vector control schemes. The new efficiency optimizing formulation yields a reference rotor flux, which ensures a minimum loss and yields an improved efficiency of the drive system especially when driving part load. The proposed scheme and its advantages are demonstrated both by computer simulations and some experimental results for motor speed control.