علي قره خاني

Nowadays, due to the increasing need for energy, the limitation of fossil resources, an‎d environmental problems such as global warming, the use of renewable energy sources has increased. Wind energy is one of the renewable energy sources that has attracted the attention of societies due to its easy access an‎d economic nature. The optimality of energy production in terms of economy an‎d the maximization of the amount of energy extracted are two important factors in the development of wind turbines an‎d their energy extraction methods. The aim of this project is to develop an algorithm for extracting maximum power an‎d tracking the maximum power point in wind turbines using the actor-critic reinforcement learning method, which is serially coupled with a PID controller for controlling the torque of the wind turbine. The implemented algorithm tracks the maximum power point by controlling the rotation speed of the wind turbine rotor an‎d using the PID controller, whose reference point is the rotor speed controlled by the reinforcement learning algorithm. The implemented algorithm is simulated in the MATLAB/SIMULINK environment an‎d compared with a number of existing methods.

انرژي تجديدپذير , توربين بادي , نقطه توان بيشينه , رديابي نقطه توان بيشينه , يادگيري تقويتي , الگوريتم انجام دهنده - منتقد

renewable energy , wind turbine , maximum power point , maximum power point tracking , reinforcement learning , actor-critic algorithm

Ali Gharekhani

Soheil Ganjefar