Abstract

Abstract

Abstract

Abstract

The accurate forecasting of wind power plays a veritable part in integrating renewable energy from wind turbines into power grids. Wind power, being a highly volatile mode of energy generation owing to temporal variations and complex weather patterns, renders reliable predictions essential for energy management and grid stability. In order to tackle this, we propose a hybrid Multi-Head Attention and 1D-Convolutional Neural Network (MHA-CNN) architecture that combines attention mechanisms and convolutional layers to capture both long-term dependencies and localized features in time-series data from a Supervisory Control and Data Acquisition (SCADA) system. The model effectively improves forecasting performance by attaining an R2score of 99.42 for hour-ahead and 96.52 for day-ahead predictions on a 50,540-sample, 10-min SCADA dataset using 5-fold chronological cross-validation, outperforming traditional methods without any manual feature engineering. The proposed method is also evaluated across multiple scenarios to assess the robustness of the proposed approach.OPEN ACCESS Received: 01/10/2025 Accepted: 10/11/2025

Abstract

The accurate forecasting of wind power plays a veritable part in integrating renewable energy from wind turbines into power grids. Wind power, being a highly volatile mode of energy generation owing to temporal variations and complex weather patterns, renders reliable predictions essential for energy management and grid stability. In order to tackle this, we propose a hybrid Multi-Head Attention and 1D-Convolutional Neural Network (MHA-CNN) architecture that combines attention mechanisms and convolutional layers to capture both long-term dependencies and localized features in time-series data from a Supervisory Control and Data Acquisition (SCADA) system. The model effectively improves forecasting performance by attaining an R2score of 99.42 for hour-ahead and 96.52 for day-ahead predictions on a 50,540-sample, 10-min SCADA dataset using 5-fold chronological cross-validation, outperforming traditional methods without any manual feature engineering. The proposed method is also evaluated across multiple scenarios to assess the robustness of the proposed approach.OPEN ACCESS Received: 01/10/2025 Accepted: 10/11/2025 Published: 23/01/2026

Abstract

Wind turbines in cold climates are likely to suffer from icing events, deteriorating the aerodynamic performances of the blades and decreasing their power output. In this work, a 3-hour rime ice accretion event is numerically simulated on five significant sections of a wind turbine blade operating in steady wind using a high-fidelity procedure based on the Blade Element Momentum Theory. The onshore NREL 5MW reference wind turbine is studied. Ice accretion is simulated through a fine multi-step process, adding ice layers approximately 0.5 mm thick; each step consists of the successive coupling of a CFD simulation, a Lagrangian particle-tracking of the cloud droplets, an ice accretion step, and re-meshing of the new geometry. Ice roughness is modelled with an equivalent sand-grain approach. After computing the aerodynamic coefficients of ice-contaminated airfoils, power losses are obtained considering the aeroelastic response of the wind turbine in turbulent winds as defined by the Design Load Case 1.1. The effect of the extension of roughness on the surface of the blade is also assessed. In the considered operating conditions and accretion times, a strong dependence between the decrease in power output and the tip-speed ratio and a small dependence on surface roughness are found.