Applied Systems Theory: Wind Turbine Output Power Prediction based on Wind Energy Utilization Coefficient

Abstract

The output power of a wind turbine is the most critical variable reflecting the operating status of the turbine. To improve the interpretability of the prediction model, a segmented output power method based on wind energy utilization coefficient is established. First, the wind energy conversion system of the wind turbine is given, and the SCADA data of a wind turbine is visually analyzed. Then it is proposed to separate the data into three groups according to different operating regions of wind turbines: the Maximum Power Point Tracking region, the rotator speed control region, and the power control region. In the Maximum Power Point Tracking region, wind energy utilization coefficient is found by a fitted cubic polynomial of the tip speed ratio. In the rotator speed control region, a modeling method for determining wind energy utilization coefficient through dynamic labels is designed. In the power control region, the output power is kept at the rated value. Finally, the 3 models are connected so that time-series data can be handled. The SCADA data of a 2.1MW wind turbine is used to verify the above models. The performance of these models is given in the form of Root Mean Square Error, indicating that the output power predicted by this method has good accuracy.The segmented output power model based on wind energy utilization coefficient can simulate the operation process of wind turbines, and has good accuracy and interpretability.