DC-signal injection for stator-resistance estimation in symmetrical six-phase induction motors under open-phase fault

Multiphase machines are often chosen due to their enhanced fault tolerance. Six-phase ones are especially convenient because they may be fed by off-the-shelf three-phase converters. In particular, those with symmetrical windings offer superior postfault capabilities. On the other hand, estimation of...

Full description

Saved in:
Bibliographic Details
Main Authors: Yepes, Alejandro G., Abdel-Majeed, Mahmoud Said, Che, Hang Seng, Abdel-Khalik, Ayman Samy, Ahmed, Shehab, Doval-Gandoy, Jesus
Format: Article
Published: 2023
Subjects:
Online Access:http://eprints.um.edu.my/38250/
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Multiphase machines are often chosen due to their enhanced fault tolerance. Six-phase ones are especially convenient because they may be fed by off-the-shelf three-phase converters. In particular, those with symmetrical windings offer superior postfault capabilities. On the other hand, estimation of the stator resistance is important for purposes such as thermal monitoring and preserving control performance. Resistance estimation by dc-signal injection provides low sensitivity to parameter deviations compared with other techniques. It has previously been shown that the dc signal can be added in the nontorque-producing (x-y) plane of a six-phase machine to avoid the torque disturbances that typically arise in three-phase machines. However, extending this method to the case of an open-phase fault (OPF) is not straightforward because of the associated current restrictions. This article addresses dc-signal injection in a symmetrical six-phase induction motor with an OPF. It is shown that, in contrast to healthy operation, the postfault dc injection should be carefully performed so that minimum copper loss, peak phase current, and zero-sequence braking torque are achieved. A solution that attains optimum performance in all these aspects simultaneously is proposed. Adapted controller and resistance estimation are also presented. Experimental results confirm the theory.