High performance sensorless control of induction motor drives for industry applications

This paper deals with some aspects concerning the implementation of a sensorless induction motor control for industry applications. The proposed control method uses an adaptive rotor flux observer and a suitable adaptation law for the speed estimation. It has been found that this method is superior to many other methods when the performance versus complexity criterion is considered. The adaptation law is obtained by a general one, valid for any induction motor parameter estimation. A comparison between the floating and the fixed point DSP implementation of the proposed method is presented. In fact, the appearance of new and powerful fixed point DSP microcontrollers makes the fixed point solution very attractive for many industrial applications, such as retrofits, conveyors control, and a lot of manufacturing process control. Different algorithm implementation issues, such as the online computation of the observer gain matrix and of the discrete motor model are discussed in this paper. Simulation and experimental results for the two cases (fixed and floating point DSP implementation) are shown.