Simulation Studies for Doubly-Fed Wind Generator Based on Neural Network Internal Model Control
Wind power is a clean and renewable energy, and its exploitation is developing rapidly across the world. The Variable Speed Constant Frequency (VSCF) wind power generation system requires the doubly-fed wind generator (DFIG) to have high response as well as good robustness. DFIG will be affected both by internal disturbance of parametic variations and external disturbance of the load torque oscillation. The dynamics of the current response of the present PI controlled rotor may be in bad condition. In addition, the above disturbances reduce efficiency and stability of the wind turbine. Internal model control (IMC) is a kind of control strategy based on mathematical model to design the controller, it has advantages such as simply to design, robustness, convenient to research. Compared with PI control, IMC has a faster response, and is insensitive to the parameter variation and disturbances, but the IMC highly depends on accurate model of the controlled object. To solve this problem, combine the neural networ and IMC to design the neural network internal model controller .The simulation indicates that the dynamic performance and the anti-interference ability of DFIG are improved.
J. Liu and J. R. Chen, "Simulation Studies for Doubly-Fed Wind Generator Based on Neural Network Internal Model Control", Advanced Materials Research, Vols. 383-390, pp. 3571-3577, 2012