Enhanced radial fuzzy wavelet neural network with sliding mode control for a switched reluctance wind turbine distributed generation system

摘要

In this article, an Enhanced Radial Fuzzy Wavelet Neural Network with Sliding Mode (RFWNNSM) and hill-climb searching (HCS) maximum power point tracking (MPPT) strategy are proposed for a switched reluctance generator (SRG) in a variable-speed wind energy conversion system (WECS). The implemented MPPT control is an HCS algorithm that does not require the knowledge of turbine or generator characteristics. However, this study proposes a simple optimal DC-Link voltage search type HCS control, and therefore a faster convergence to MPPT is achieved. A high-performance online training Enhanced Radial Fuzzy Wavelet Neural Network (RFWNN) using a back-propagation learning algorithm with a sliding mode regulating controller is designed for an SRG. The MPPT strategy locates the system operation points along the maximum power curves based on the DC-Link voltage of the inverter, thus avoiding the need to detect the generator speed.