Optimized deep learning approaches for six phase dfig based wind energy conversion system

Abstract

Renewable energy generation has gained significant importance in recent years due to the increasing global concern about climate change and environmental sustainability. Among renewable energy sources, wind energy stands out as a promising option for clean power generation. Doubly Fed Induction Generator (DFIG)-based wind energy conversion systems have emerged as a popular choice for harnessing wind energy efficiently. However, the conventional Proportional-Integral (PI) controller used in DFIGs exhibits limitations in maintaining stability and efficiency when subjected to dynamic variations and faults in the system. Considering the limitations of conventional control approaches, this study introduces two innovative approaches: the Flower Pollination Optimization-Based Deep Convolutional Neural Network-based Adaptive Control Scheme (FPA-DCNN) and Flower Pollination Optimized deep reinforcement learning algorithm (FPA-RLA) for DFIG-based wind energy conversion systems. The primary motivation for this research is to address the instability and inefficiency of the PI controller under diverse operating conditions. The proposed FPA-DCNN and FPA-RLA aims to provide self-adaptation capabilities to the DFIG system, ensuring optimal performance under various scenarios, such as changes in wind speed, fluctuations in generator parameters, and asymmetrical grid faults. newline