Abstract
Recently, wind power generation systems have seen significant developments aimed at improving performance and efficiency. Permanent magnet synchronous generators (PMSG) are essential for wind power production systems because of their exceptional power density, high efficiency, and dependable operation. These properties enable PMSGs to effectively convert wind energy into electrical energy with minimal losses and high accuracy. This study proposes an enhanced control system for wind power generation using permanent magnet synchronous generators (PMSG), integrating artificial neural networks (ANN) and predictive current control (PCC) techniques to optimize efficiency and dynamic performance. Traditional control systems often require wind speed measurement and utilize separate loops for speed and current regulation. In contrast, the proposed approach eliminates the need for wind speed sensing by employing an ANN to estimate optimal reference currents based solely on rotational speed measurements. PCC is then applied to regulate the generator currents, replacing conventional PI controllers. This configuration enables the removal of the speed regulation loop while maintaining maximum power point tracking capability. Simulation results demonstrate that the ANN-PCC control system achieves superior dynamic response and higher overall efficiency compared to conventional methods. The system exhibits faster tracking of maximum power points, reduced total harmonic distortion, and enhanced stability under variable wind conditions. Additionally, the proposed controller simplifies the overall system architecture by reducing sensor requirements and control loops. The paper presents a comprehensive comparison between the proposed ANN-PCC system and two conventional control methods: a perturbation and observation (P&O) algorithm with PI control, and a tip speed ratio (TSR) technique with PI control. Results show that the ANN-PCC approach outperforms both conventional methods in terms of response time, power output stability, and grid current quality. These improvements make the ANN-PCC approach a promising solution for increasing the performance and reliability of PMSG-based wind energy conversion systems in grid-connected applications, potentially contributing to the broader adoption of wind power technology.

Key words: Predictive Current Control (PCC), Artificial Neural Networks (ANN), Wind Energy Conversion System (WECS), Maximum Power Point Tracking (MPPT), Permanent Magnet Synchronous Generator (PMSG), Grid Integration.