Title :
Mechanical-state estimator for doubly-fed induction generators — Application to encoder-fault tolerance and sensorless control
Author :
Mollet, Yves ; Gyselinck, J.
Author_Institution :
Ecole Polytech. de Bruxelles, Bio-, Electr.- & Mech. Syst. (BEAMS) Dept., Univ. Libre de Bruxelles, Brussels, Belgium
Abstract :
In this paper a new model reference adaptive system observer for estimating the mechanical quantities of a doubly-fed induction generator, namely position, speed and electromagnetic torque and power, is presented. It is based on the computation the air gap power vector, considering saturation by using a variable magnetizing inductance. This algorithm can be used as a back-up in case of encoder fault thanks to an associated fault detection and isolation technique, combining a cumulative-sum function and a state-space adaptive model. Experimental tests on a 3 kW 4-pole doubly-fed induction generator show that a varying magnetizing inductance is recommended if the reactive power flow at stator terminals has to be varied in a large range. The encoder-fault detection time does not vary significantly with speed, torque or stator reactive power flow.
Keywords :
adaptive estimation; air gaps; angular velocity control; asynchronous generators; fault diagnosis; fault tolerance; machine insulation; magnetisation; model reference adaptive control systems; observers; position control; power control; sensorless machine control; torque control; air gap power vector; cumulative-sum function; doubly-fed induction generator; electromagnetic torque control; encoder-fault detection; encoder-fault tolerance; fault isolation technique; mechanical-state estimator; model reference adaptive system observer; position control; power 3 kW; power control; reactive power flow; sensorless control; speed control; state-space adaptive model; stator terminal; variable magnetizing inductance; Adaptation models; Current measurement; Reactive power; Rotors; Saturation magnetization; Stators; Voltage measurement; Adaptive model; Doubly-fed induction generator; Fault detection; Fault tolerance; Power systems reliability; Redundancy; Sensorless control; State-space methods; Variable speed drive; Wind energy generation;
Conference_Titel :
Electrical Machines (ICEM), 2014 International Conference on
Conference_Location :
Berlin
DOI :
10.1109/ICELMACH.2014.6960424