Implementation and simulation of sensorless control and field weakening for an induction machine

Abstract

AROS electronics produce Permanent Magnet SynchnronousMachines but the market for magnets can be very volatile in that the prices may fluctuate significantly. The Induction machine is then an attractive replacement. The robustness and simple construction makes it one of the most used electrical drives in the industry. However, it is often controlled with a speed sensor or an open loop configuration like the well known Volt/Hertz-control. A field oriented sensorless control would make the Induction machine even more attractive from an economical and maintenance point of view, but the problem is that the flux and the speeds need to be estimated. The largest drawback with sensorless control is that the machine will eventually turn unstable in the low speed region. In order to reach speeds above rated speed, field weakning is required. A field oriented sensorless control model with a flux estimator known as the Statically Compensated Voltage Model has been modelled and implemented together with field weakning in one of AROS’s electronics digital signal processors. The control model was simulated in Matlab/SIMULINK to obtain information about the system robustness and its limitations. The implementation was done in a C-language environment on a 16-bit fixed point processor where tests showed that the system is operating well at nominal speed of 1400 RPM with a nominal torque of 7.5 Nm. The field weakening algorithm made it possible to reach twice the rated speed, 2800 RPM, with a load torque of 5.5 Nm. At about 3100 RPM the machine turns unstable because the maximum voltage that the converter can put out is reached and the current therefore becomes uncontrollable.