Comparative study of selected speed estimation methods for asynchronous electric drives.

Speed estimation problem stands viral for the field of electric drive control theory. This paper presents a study of two angular speed estimation methods of AC motor drives. On the example of an induction motor drive with direct stator field vector control speed feedback and with speed monitors, a comparative analysis is carried out. A comparative analysis of classical model-reference adaptive system and a neural network speed estimator is given. The study analyzes the effectiveness of well-known artificial neural network-based techniques and shows how they can be strengthened using some expert knowledge. Whereas many researchers denote that using black-box concept in electrical drives parameters estimation does not give the best solutions. Better results could be achieved by involving known relations between the estimated parameter and the other known parameters of the drive. Here come the so-called grey-box techniques, where expert knowledge supports the estimation algorithms. [ABSTRACT FROM AUTHOR]