It is well known, that to achieve the maximum power in wind power generation systems, the maximum power point tracking (MPPT) algorithm should be employed as the main strategy to extract the wind power. However, due to the intermittent and unpredictabe nature of the wind speed, then the usage of the MPPT without using any compensation, will practically generate severe power fluctuation. The high fluctuation of the wind power basically come from to the cubic curve relation between the extracted wind power and wind speed in the MPPT method. To minimize the high fluctuation of the extracted wind power, in this paper we propose a simple linear power or a constant torque reference strategy. Based on simulation result run at Matlab/Simulink, it is shown that the power extracted by using a linear power or a constant torque reference is smoother compared with the usage of the maximum power point tracking strategy. However, by using this strategy the power average is lower than the MPPT. ; As electrical energy demand has been increasing rapidly, and due to the depletion of fossil fuels and also boosted by global warming issues, the utilization of green power generation systems such as wind power generation system is very demanding [1-2]. To achieve a high wind power efficiency, an algorithm that best known as Maximum Power Point Tracking (MPPT) is usually employed as the main strategy to extract the wind power. In this strategy, the rotor of the wind turbine generator (WTG) is varied to track the maximum of the wind power, either by controlling output power directly or by regulation of the rotor speed at the outer control loop of the rotor side-converter [3-6]. Although the MPPT is the best strategy in the sense of the maximum power generation, however the wind turbine power usually will fluctuate severely. The huge fluctuation of the wind turbine power is come from the fact that the relation of the wind power to the wind speed is a cubic function, while the wind speed is actually an uncertain turbulent variable that fluctuate rapidly and stochastically and very difficult to predict precisely [7]. In grid connected wind turbine systems, this fluctuating power could make serious problem mainly. In this case, the utility grid frequency could deviate from its nominal. Due to high expectations in the penetration of more wind turbine generation systems in the future time, the research of the wind power smoothing is very important and demanding to be done. There are several methods found in literatures to cure the fluctuation in wind power harvesting. One of the solutions to smooth the output power is by employing an energy storage system, including supercapacitor [8-9], the superconducting magnetic energy storage-SMES [10-11], high speed flywheel [12-13] and battery energy storage system [14]. The operation principle of the energy storage systems in the wind turbine system are almost the same: the storage will be charged if there is an excess energy caused by a high wind speed, and the stored energy will be released if the wind power output is lower than desired. The other techniques to smooth the wind power fluctuation is by employing a pitch control of the turbine blade [15] and by a constant power reference strategy [16]. Although the pitch control technique can give a promising result in the wind power smoothing, however, as point out by [17] the pitch control will make the stress on the blades increased. While the later, to be work properly this strategy practically need a very accurate prediction system of the very-short term wind speed (in the second time scale) The main objective of this work is to smooth wind power fluctuation of the generator system output without using energy storage devices and without an accurate wind speed prediction system. To fulfil this objective, we here propose a linear power or a constant torque reference strategy to mitigate power fluctuation of the wind turbine system output. For simulation purposes, the WTG used in this study is DFIG-double fed induction generator. The remainder of the paper is organized as follows. Section 2 describes the general model of the wind turbine generator based on DFIG model. In Section 3, the wind turbine aerodynamic and MPPT algorithm will be discussed, next, Section 4 present a linear power or a constant torque reference design. Next, the simulation result will be presented in Section 5. Finally, the conclusions are drawn at Section 6.