Hybrid sliding mode-ANN control for UPQC in wind turbine systems under real variable speeds in Adrar, Algeria

This paper explores the design and performance evaluation of a three-phase Wind Turbine System-Unified Power Quality Conditioner (WTS-UPQC). The system integrates both series and shunt voltage compensators, connected through a common DC link, capitalizing on the benefits of distributed generation and active power filtering. The shunt compensator performs dual functions: extracting energy from the wind turbine system and compensating for harmonic currents and reactive power caused by nonlinear loads. Harmonic detection and reactive power compensation are achieved using the well-known p- q theory. The study is based on real wind speed measurements from the Adrar region of Algeria. Additionally, the Shunt Active Power Filter (ShAPF) not only corrects harmonic currents and reactive power but also improves energy extraction from the Wind Turbine System by utilizing maximum power point tracking (MPPT). This research addresses the drawbacks of conventional sliding mode control (SMC) for three-phase ShAPF, particularly the ripple effect produced by the sign function in the control law, which can degrade power quality and impact the materials used in wind conversion systems. To overcome these issues, an Artificial Neural Network-Sliding Mode Control (SMC-ANN) method is proposed. This approach enhances wind energy capture, mitigates the ripple effect, and improves the overall quality of power supply, especially in the face of variable wind speeds and voltage fluctuations. Simulation results and comparative analysis demonstrate that SMC-ANN significantly outperforms traditional SMC, proving its efficiency in practical applications.