New M5P model tree‐based control for doubly fed induction generator in wind energy conversion system

Abstract

Abstract This paper presents a new algorithm based on the combination of fuzzy logic control with M5 decision tree algorithm for doubly fed induction generator active and reactive power control. The M5 model tree is a machine learning algorithm that uses the extracted dataset from fuzzy logic for the aim of control performance enhancement. Because the fuzzy logic provides high design and implementation complexity, the new control approach is proposed to reduce its complexity and to achieve fast dynamic control by translating the fuzzy logic algorithm into simple if‐then instructions using M5 model tree. The obtained model shows good agreement with fuzzy logic in rotor side control. The effectiveness of the proposed control approach is investigated through Matlab/Simulink software. Furthermore, a cosimulated through processor in the loop testing is performed as an experimental verification using a low‐cost STM32F407 discovery board.