WRBF network based control strategy for PMSG on smart grid

Abstract

The rotation speed of turbines can be adjusted in the real time according to wind speed for maximum power point tracking (MPPT) in power generation systems on smart grid. In this paper, a Wilcoxon radial basis function neural (WRBFN) network based MPPT strategy is proposed for permanent magnet synchronous generator (PMSG) on variable speed wind turbine generation systems. The proposed MPPT strategy adopts a hill climbing searching (HCS) method, and thus is independent of the turbine and generator characteristics. The design of a high-performance on-line training WRBFN is used for a PMSG with back-propagation learning algorithm regulating controller.