Search

This paper focuses on the improvement of an air-gap-power-vector-based observer for sensorless doubly fed induction generators. After showing how a mismatch in the machine inductances can worsen the performance of the basic observer scheme, an adaptive tuning of the stator inductance is proposed. The adaptation is achieved by minimizing the error between the squared modulus of two suitable air-gap complex power space vectors. A criterion to settle the dynamics of the adaptation mechanism is provided, and a sensitivity analysis in the steady-state operation is performed. Some experimental results are reported and discussed in order to validate the soundness of the proposed adaptive tuning algorithm. [ABSTRACT FROM PUBLISHER]

This paper deals with the performance of some numerical integration (NI) methods for ordinary differential equations in real-time applications involving stiffness. After summarizing the main results about accuracy and stability of some well-known schemes, the point of view of the control is emphasized, by showing the impact of the NI on the dynamics of the discretized model. Moreover, some peculiarities of the NI in hybrid systems are touched by introducing a simple hybrid model for testing the performances of some NI schemes. Finally, some experimental tests are shown, dealing with a control of a sensorless drive whose model is characterized by stiffness. [ABSTRACT FROM AUTHOR]

This paper deals with the adaptive tuning of the stator inductance in a rotor-current-based Model Reference Adaptive System observer for the sensorless control of doubly fed induction machines. At first, the effect of mismatched parameters in this observer is discussed in order to show the considerable influence of the stator inductance on the accuracy of the estimated rotor position. Then, an adaptive tuning of the stator inductance is proposed and a small signal model is deduced in order to design the tuning loop. Moreover, a theoretical sensitivity and stability analysis is performed. Finally, the performances of the proposed scheme are experimentally investigated and validated. [ABSTRACT FROM AUTHOR]