Decoupling Network Design for Inner Current Loops of Stand-Alone Brushless Doubly Fed Induction Generation Power System

Brushless doubly fed induction generator (BDFIG) is of high reliability due to its brushless structure. However, brushless structure also results in the complicated d-q vector model, in which the d and q channels are seriously coupled. Traditionally many feedforward (FF) terms were added during control for decoupling purpose. However the FF method features several drawbacks: 1) it requires extra sensors; 2) the rotor position information needs to be known for d-q transformation; and 3) the decoupling effect highly depends on the parameter accuracy. To overcome these problems, this letter proposes a decoupling method based on the decoupling network (DN). With DN method, the BDFIG and the load are first modeled as a dual-input-dual-output system, and then the DN is designed. From control point of view, the DN is in series with the BDFIG and the load, and thus, form a new control plant with highly decoupling feature. As a result, the controllers in d and q channels can be easily designed. The proposed DN method requires neither extra sensors nor rotor position information, and the robustness is enhanced since parameter variations of BDFIG and load can be fully considered during DN design. In this letter, such a DN method is presented and applied on the inner current loops of a stand-alone BDFIG system to obtain the good decoupling feature in the overall operational range, and the design results are verified by the experiments from a stand-alone BDFIG system platform.