A self-tuning thermal protection scheme for induction machines

Abstract

This paper proposes a new method for improving the performance of modern thermal protection devices for induction machines. Microcomputer-based thermal models have begun replacing the traditional electromechanical elements for thermal overload protection in induction motor drives, because they more reliably estimate temperatures in the machine and provide the user with greater flexibility in responding to overload conditions. In a totally enclosed fan-cooled induction motor, however, cooling conditions can change because of a broken fan or clogged air vents, leading to inaccurate temperature estimation by the software model. This paper proposes a scheme which updates the parameters of the thermal model in the event that they change. The tuning scheme is driven by the difference between two speed estimates; one is based on the slip relation and the other on parameter-independent saliency harmonics in the current spectrum. As the rotor resistance changes, the slip relation speed estimate produces an error which drives the tuning mechanism. Since the rotor resistance is proportional to the rotor temperature, this provides a temperature curve from which the thermal time constant of the rotor can be derived. An assumed thermal model of the machine is then updated to account for any change in the thermal time constant. Unusual temperature increases under normal operation can also be detected.