Section Division and Multi-model Method for Early Detection of Icing on Wind Turbine Blades

Icing on wind turbine blade is commonly seen in bad running environment. And this may cause deformation of blade, lower wind turbine efficiency, and affect the stability of the power grid. The working state of the wind turbine changes dynamically under different wind speeds and powers, which makes it difficult to establish a global model for blade icing detection. In this paper, an automatical section division algorithm was proposed. It takes multi-working-conditions into account, partitions the whole samples into different sections by exploring the changes of underlying characteristics, and then establishes different submodels for each section. For online application, the method can quickly assign each new sample to its own section by using the indication feature. In comparison with global classification method, it not only reveals better performance in detection accuracy and achieves earlier detection of icing tendency, but also provides enhanced process understanding and interpretability.