Application research of self-adapting output feedback controller based on RBF neural network on WECS

This paper presented an adaptive control method for WECS with variable speed and variable pitch. To gain zero steady tracking error and reject the gust disturbance, the linearized state equation was extended by introducing a extra state variable. The eigenvalues of the extended system equation were shifted to a pre-specified vertical strip to achieve satisfied the dynamic performance (damping). For practical implementation, the design method only uses partial output feedback. RBF neural network was used to approximate the nonlinear mapping relation between stable state operating point and the output feedback gain. For demonstrating the effectiveness of the system performance enhance, simulation results are used to show that the proposed controller gives significant improvement in the steady performance, dynamic performance and rejecting disturbance in WECS.