Sensorless vector control with online in-phase inductance identification for synchronous reluctance motors

Abstract

This paper proposes online inductance identification methods, including two identification principles, three identification algorithms, and a compensation signal generation method for the sensorless drive of synchronous reluctance motors. The proposed methods have the following features. (a) The identification object is the “in-phase inductance”, which is only required for the “mirror-phase flux estimation”. (b) The inductance can be identified even in the steady state, where the frequency of the signals in the dq synchronous reference frame is substantially zero, and identifications are the most difficult. (c) The superimposition of a special signal, such as an M-sequence signal, is not required for the identification. (d) A compensation signal is essential for making the identified value converge to the true one. A compensation signal generation method that is applicable in the minimum copper loss control is proposed. (e) The identification can be performed, in principle, in the γδ general reference frame but it is practically performed in the αβ stationary reference frame. (f) Recursive calculations are possible, and the amount of calculations required for the identification is very small. This study also verifies the validity of the proposed identification methods with the aforementioned features through extensive numerical experiments.