Intelligent control of a class of wind energy conversion systems

This paper discusses the control problem for a class of wind energy convers ion systems (WECS). It first develops a detailed model and then compares am ong four control algorithms based on conventional and intelligent control t heories. A simple PI conventional controller for the exciter loop is carrie d out by using a first order model. When the system operating points change , the PI controller fails to provide sufficient damping or acceptable perfo rmance. Therefore both a fuzzy voltage and a fuzzy power regulators are int roduced. Also a conventional adaptive pitch controller is proposed to adjus t the pitch angle of the rotor blades in order to maximize the energy captu re and reduce the mechanical loads. As an alternative to this controller, a neural network controller is also designed. Using the existing nonlinear w ind model and the different control algorithms, the dynamic behavior of the controlled (WECS) is simulated. By selecting convenient wind data, the sys tem characteristics such as its tracking performance, its robustness, and i ts ability to recover from large disturbances are studied and discussed.