Q-learning and Dynamic Fuzzy Q-learning Based Intelligent Controllers for Wind Energy Conversion Systems

This paper investigates the performance of two different intelligent controllers for rotor side converter of doubly-fed induction generator (DFIG) used in wind energy conversion systems. The design of these controllers are based on the concept of Q-learning and dynamic fuzzy Q-learning (DFQL). The Q-learning based controller essentially provides discrete control actions unlike DFQL based controller which generates continuous control actions. However, the learning time in DFQL based controller is significantly more compared to the conventional Q-learning. The effectiveness of these controllers are studied considering a test wind energy conversion system consisting of 20 wind turbines which are driven by 20 DFIGs. The results of the simulation illustrate that Q-learning based intelligent controller performs better compared to DFQL controller.