BACKSTEPPING INTEGRAL SLIDING MODE CONTROL FOR ENERGY CAPTURE OPTIMIZATION OF WIND TURBINE SYSTEM

This paper reports an efficient nonlinear robust controller design method of a variable-speed wind turbine, in which maximum wind energy can be extracted at below the rated wind speed using torque control. Furthermore, one of the design problems of controllers is related to the uncertainties in the dynamic model system. To overcome this problem, a robust controller is investigated; the designed controller is developed by combining a nonlinear backstepping approach and an integral sliding mode control strategy. The proposed controller is combined with intelligent systems such as a neural network and an evolutionary algorithm to improve the controller performances. However, to predict the uncertain part of the wind turbine model with lower switching gain, a general regression neural network (GRNN) is adopted. Thus, a particle swarm optimization approach with an efficient global search technique is employed by training online the GRNN weights. In addition, the Lyapunov approach is proposed to investigate the system stability with the considered controller. Moreover, a comparison with other strategies such as backstepping sliding mode control, integral sliding mode control and sliding mode control controllers is reported. We noticed from the results of simulation that the studied controller presents good performances in terms of transition response and tracking error level.