15 th International Power Electronics and Motion Control Conference, EPE-PEMC 2012 ECCE Europe, Novi Sad, Serbia Wind turbine with doubly fed induction generator operating at limited power point V.D. Lazarov 1 , L.S. Stoyanov 1 , Z.J. Zarkov 1 , G. Notton 2 1 Technical University of Sofia, Sofia, Bulgaria, E-mail vl_lazarov@tu-sofia.bg, ludiss@tu-sofia.bg, zzza@tu-sofia.bg 2 University of Corsica “Pascal Paoli”, Ajaccio, France, E-mail notton@univ-corse.fr Abstract — The paper presents the modelling of a wind energy conversion system with doubly fed induction generator with control systems operating in two modes – maximum and limited power point tracking. The simulation results illustrate the system’s correct operation. Keywords — Wind energy, doubly fed induction generator, control strategies. I. INTRODUCTION Over the last ten years the number of installed wind power generators has seen annual increases due to the depletion of traditional energy sources and due to legislation imposing a cost on greenhouse gas emission [1]. Installation owners seek to operate with maximum power point tracking in order to obtain return on their initial investment as soon as possible. This scenario is not always acceptable to the system operator because of the need to preserve the “production–consumption” balance or because of the capacity limitations of the electrical grid. The owner may then be obliged to switch off the generator to maintain the balance of the power system. This is not an optimal solution from an economic point of view. A preferable approach is a reduction of the power of wind generators, rather than their complete turn-off. This paper presents the modeling of a full wind energy conversion system (WECS) and its control, which ensures the operation of the generator with either maximal or limited output power using electrical ways without any additional mechanical mechanisms. The proposed control strategy for output power limitation has an advantage over the standard pitch angle variation because of the simpler wind turbine (WT) design, i.e. a construction without the drives of the WT pitch control. Moreover, the electrical control is faster because the smaller time constant. An indicator for the correct operation of the control is selected and the system simulation is presented. II. STUDIED SYSTEM MODELLING The studied wind energy conversion system (Fig. 1) is composed of a wind turbine, a gear box (GB) and a doubly fed induction generator (DFIG). The configuration DFIG uses a wound rotor induction generator (WRIG) with a back-to-back power converters structure. This structure allows the machine operation in generator mode in the largest possible speed range – from sub-synchronous to super-synchronous [2]. The models of different system elements are presented bellow. WT GB WRIG stator rotor Back-to-back Grid WT GB WRIG stator rotor Back-to-back Grid Fig. 1. Studied WECS. The wind turbine model determines the available mechanical power (P mec ) for the generator entrainment by (1) [3-4], where c p is the power coefficient with a theoretical limit 0.593, called a Betz limit [5], ρ а =1.225 kg/m 3 is the air density, A r is the blades sweep area and V is the wind speed. 3 a r p mec V A c 2 1 P ρ = (1) The power coefficient is the only unknown variable for a given wind speed and rotor speed. It is modeled by an analytical expression (2) [4], where c 1 -c 6 and x T are empirical coefficients, ϑ is the pitch angle and the ratio 1/Λ is determined by (3) where λ is the tip speed ratio. Λ ϑ ϑ Λ 1 c 5 x 4 3 2 1 p 6 T e c c c 1 c c c -       - - - = (2) 3 1 035 . 0 08 . 0 1 1 ϑ ϑ λ Λ + - + = (3) The determined mechanical power is used for the mechanical torque on the generator shaft calculation (4). r mec m P T ω = (4) where ω r is the generator rotor speed The gear box is considered only taking into account a speed ratio. Thus the generator speed is the product of the gear box ratio and the turbine speed. Using ω r in (4) the mechanical torque of the turbine is referred to the generator electromagnetic torque. The electromagnetic processes in the induction generator are modeled in the dq reference frame with 2 systems with 4 equations - the system with Kirchhoff’s law application on each generator winding, and the system with flux equations [6]. In Matlab/Simulink the machine model has to use state variables. They could be currents or fluxes. The second option is used in this paper. The currents in the system with flux equations are expressed in (5). Those current equations are replaced in