TD3 Actor‐Critic Reinforcement Learning for Maximum Power Point Tracking in Wind Turbines Coupled With PID Controller

ABSTRACT

Rising energy demand, depletion of fossil fuels and environmental challenges such as global warming have significantly increased the use of renewable energy sources. Wind energy, being easily accessible and economically viable, has garnered considerable attention from society. Two crucial factors in wind turbine development are cost‐effective energy production and maximising the extracted energy. This paper aims to develop a maximum power point tracking (MPPT) algorithm for wind turbines using an actor‐critic reinforcement learning (RL) approach, specifically the TD3 algorithm, which is cascaded with a proportional‐integral‐derivative (PID) controller for wind turbine torque control. The implemented algorithm achieves MPPT by controlling the rotational speed of the wind turbine rotor. The TD3 RL algorithm determines the reference rotor speed for the PID controller, which then regulates the turbine torque to track this reference. The proposed algorithm has been simulated in the MATLAB/SIMULINK environment and compared with several existing methods.