Scholarlink

Accurate forecasting of wind power generation is closely linked to how well wind speed variability is represented, particularly through the estimation of Weibull distribution parameters. In this study, four parameter estimation approaches AMLM, MM, GM, and MMLM—are systematically evaluated using real operational data from the Tétouan wind farm located in northern Morocco. The analysis is based on wind speed records measured at 80 m above ground level throughout the year 2019, with the objective of assessing each method’s capability to predict actual energy production. The comparative results indicate that the Method of Moments (MM) provides the most reliable performance, yielding a minimal corrected deviation of +1.09% between predicted and measured energy output. Furthermore, the integration of turbine technical availability information derived from the SCADA system enhances the robustness and realism of the estimation process. These findings highlight the importance of combining statistical wind modeling with real operational data to achieve more accurate energy forecasts and to support the effective operation and management of wind farms under practical conditions.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Performance Assessment of AMLM, MM, GM and MMLM for Weibull-Based Wind Power Prediction in Real Operating Conditions

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Mohamed Bousla,
  • Ali Haddi,
  • Youness El Mourabit,
  • Badre Bossoufi

摘要

Accurate forecasting of wind power generation is closely linked to how well wind speed variability is represented, particularly through the estimation of Weibull distribution parameters. In this study, four parameter estimation approaches AMLM, MM, GM, and MMLM—are systematically evaluated using real operational data from the Tétouan wind farm located in northern Morocco. The analysis is based on wind speed records measured at 80 m above ground level throughout the year 2019, with the objective of assessing each method’s capability to predict actual energy production. The comparative results indicate that the Method of Moments (MM) provides the most reliable performance, yielding a minimal corrected deviation of +1.09% between predicted and measured energy output. Furthermore, the integration of turbine technical availability information derived from the SCADA system enhances the robustness and realism of the estimation process. These findings highlight the importance of combining statistical wind modeling with real operational data to achieve more accurate energy forecasts and to support the effective operation and management of wind farms under practical conditions.