A Robust Real-Time Maximum Power Points Tracking (MPPT) Method for Wind Power Systems Yu Zou yz31@zips.uakron.edu Malik Elbuluk melbuluk@uakron.edu Yilmaz Sozer ys@uakron.edu Electrical and Computer Engineering Department The University of Akron Akron, Ohio Abstract—Maximum power point tracking (MPPT) is the key to notably improve efficiency of variable-speed wind power systems. A power variation rate checking (PVRC) and reference power curve (RPC) methods have been developed and implemented earlier. The former requires considerable computation along the whole operation while the latter highly depends on the system model and thus leading to weak robustness. This paper combines the power variation rate checking and reference characteristic power curve methods to eliminate their disadvantages. The proposed MPPT simply updates the reference power curve by tuning the curve coefficients. It is simpler and less computation compared to PVRC. Also, compared to RPC, this method is immune to the potential system variations and thus robust. To verify its performance, the proposed algorithm is simulated in Matlab/Simulink environment and the system operation under proposed MPPT is tested experimentally. I. INTRODUCTION Due to the unpredictable wind speed and highly non- linear power outputs in wind energy conversion, complicated MPPT algorithms are employed to control the power processing [1]. The studied MPPT methods can classified in three strategies. These are (1) the methods relying on wind speed, (2) the methods relying on output power measurement and calculation, like PVRC and (3) the methods relying on characteristic power curve, like RPC. Firstly, since the optimal generator speed depends on wind speed, many MPPT method research require either measurement or estimation of wind speed [2-3]. Therefore, such systems suffer from additional cost of sensors and complexity. In addition, tracking the maximum power could also be implemented through PVRC [4]. The maximum power operating point can be reached when the rate of change of power (P) with respect to speed (ω) is zero (i.e. dP/dω=0). Besides of the obstacle of fast varied wind, such method requires a lot of real-time calculations throughout the whole operation, which is an extra cost. Lastly, the characteristic curve of the wind power versus generator speed could be used as reference for control [5-6]. Such RPC method is simple and fast, but the reference curve relies on the accurate mathematical model of the wind turbine. That is, once the reference power curve is determined based on turbine model, there is no on-line tracking if any changes happen in the operating point. In an effort to solve the disadvantages of the above MPPT methods, this paper proposes a simple real-time robust algorithm, combining the PVRC and RPC. Instead of targeting the optimal power outputs, the coefficients that characterize the optimal power curve are tuned in real-time. By incrementally changing the coefficients of the reference power curve, the output power can be measured and the optimal coefficient value will be returned to update the existed reference power curve. Therefore, this method combines the advantages of the PVRC and RPC together with the absence of their weaknesses. The wind turbine and the doubly-fed induction generator (DFIG) models are discussed in Section II. The proposed novel MPPT method is presented in Section III, where the stability and sensitivity are discussed as well. Section IV gives the simulation and experimental setup and results, respectively. Summary and conclusion are provided in Section V. II. WIND TURBINE AND DFIG MODEL The kinetic energy ( ܧ ௞ ) for the wind with particular wind speed  ௪ is described as: E ୩ ൌ ଵ ଶ mV ୵ ଶ (1) where  represents the mass of the wind. The power contained in the wind then can be written as: P ୵ ൌ பE ౡ ப୲ ൌ ଵ ଶ ப୫ ப୲ V ୵ ଶ ൌ ଵ ଶ ሺρAV ୵ ሻV ୵ ଶ ൌ ଵ ଶ ρAV ୵ ଷ (2) 978-1-4673-0803-8/12/$31.00 ©2012 IEEE 3148