Robust Direct Power Control Schemes of DFIGs Driven by Variable-speed Wind Turbines

Abstract

Despite its several advantages, a classic direct power control (DPC) technique of doubly fed induction generators (DFIGs) driven by variable-speed wind turbines has some drawbacks such as high power ripples and variable switching frequency. In this paper, two robust controllers are designed to improve the classical DPC performance without complicating the overall scheme. First, an integral sliding mode controller (ISMC) is designed to regulate the stator active and reactive powers. Two integral switching functions are selected for controlling stator active and reactive powers. The idea of total sliding mode controller is selected to avoid reaching phase stability problem. Second, a diagonal recurrent neural network (DRNN) controller is designed and trained based on DPC. The DRNN has several advantages compared to the classical static neural networks such as recurrence and simple construction. Simple off-line back-propagation algorithm is proposed to train the proposed DRNN. The stability of the proposed ISMC and DRNN controller is proved using the Lyapunov stability theorem. The grid side converter is controlled based on the DPC principle to ensure both constant DC-link voltage and grid side reactive power. The feasibility of the proposed DPC schemes is validated by simulation studies on a 1.5-MW wind power generation system. The performance of the proposed schemes is compared with a conventional DPC scheme under different operating conditions.