Fully Connected Recurrent Neural Network MPPT Control Design For DFIG Wind Energy Conversion Systems

This paper is proposing a new maximum-power-point-tracking (MPPT) control design based on recurrent neural networks with real-time recurrent learning (RTRL) algorithm for getting optimal efficiency from doubly-fed-induction-generator (DFIG) wind energy conversion systems. Chosen Recurrent neural network (RNN) is a fully connected RNN with RTRL unsupervised learning algorithm. The inputs to the network are the rotor speed and wind-turbine torque, and the output is the rotor speed command signal for the wind turbine. Simulation results verify the performance of the proposed algorithm.Keywords— maximum power point tracking ,doubly-fed induction generator,recurrent neural network,real-time recurrent

Keywords:

maximum power point tracking , doubly-fed induction generator , recurrent neural network , real-time recurrent learning , wind energy conversion system