Abstract

In this paper, several design algorithms of the adaptive observer for speed estimation in sensorless drives, are viewed from a common framework based on the positive-real property. Analytical stability conditions are derived for each design strategy, using which the general solutions of stabilizing gains are obtained. The closed-form solutions of all feasible gains facilitate the practical implementation of the algorithms for any machine's parameters, and are used in the design of poles and zeros placement. New design rules for the feedback gains and the adaptation gains which preserve the required performances for all operating conditions, are also proposed. © 2007 IEEE.