Posters

Abstract

There is a broad wind energy industry effort to improve lidar-measured turbulence intensity (TI). One pathway for improvement is the use of machine learning on existing lidar datasets, without developing new hardware, scan strategies, or signal processing. If machine learning correction techniques are effective, the current generation of lidar devices can easily be used for more wind energy applications, such as Site Suitability turbine selection, and for additional aspects of Energy Yield Assessment that today require use of anemometer-measured TI. These new lidar applications will accelerate wind development by reducing development time and increasing safety through reduced usage of meteorological masts. In this presentation, we benchmark the effectiveness of a machine-learning based correction to WindCube v2.1 TI measurements (MLTI) using the KPIs developed by the CFARS Site Suitability Sub-group.