Modified active power-MRAS for limited range variable speed sensorless brushless doubly-fed reluctance machine drive

Abstract

PurposeThis paper aims to suggest a parameter independent and simple speed estimator for primary field-oriented control of a promising electro-mechanical energy conversion device in the form of brushless doubly-fed reluctance machine (BDFRM) drive.Design/methodology/approachThe speed estimation algorithm, in this context, is formulated using a modified secondary winding active power (mPs)-based model reference adaptive system (MRAS). The performance of the proposed estimator is verified through computer aided MATLAB simulation study, compared with conventional active power-based MRAS and further supported with experimental validation using a 1.6 kW BDFRM prototype run by a dSPACE-1103 controller.FindingsThe formulation of mPs-MRAS is insensitive to any machine parameters and does not involve any integration/differentiation terms. Thus, any deviation therein does not hinder the performance of the mPs-MRAS-based speed estimator. The proposed speed estimator shows stable behavior for variable speed-constant load torque operation in all the four quadrants.Originality/valueThe formulation of mPs-MRAS is insensitive to any machine parameter and does not involve any integration/differentiation terms.