Hybrid Sliding Mode Control and RNN-Based Strategy for Maximizing Power Extraction in Small Wind Turbines for Electric Vehicles

Advanced technologies like Ram Air Turbines (RATs) are being investigated because of the aviation industry's need for fuel-efficient and alternative renewable energy sources. In situations where power generation is necessary in the event of an emergency involving unmanned aerial vehicles (UAVs), RATs are essential. Optimising the RATs' performance—including power output and operational stability—under variable and unexpected wind conditions is the main obstacle, though. Conventional control techniques frequently don't adjust to these changing conditions. In order to monitor the ideal turbine rotation speed, a sliding mode control rule is developed in the proposed controller. This article emphasises the need of using a recurrent neural network (RNN) to identify unpredictable wind turbine dynamics. Control over maximum power extraction is then made possible by the development of an online update mechanism that provides real-time weight changes for the RNN. Simulation findings show that, even in the presence of significant nonlinearities and system uncertainties, the proposed controller performs 13 times better than a conventional control strategy in monitoring the ideal turbine rotation speed and obtaining the maximum wind output from RATs.