TURBULENT WIND SEED SELECTION USING ROBUST STATISTICS FOR DESIGN LOAD ANALYSIS OF WIND BLADE DESIGN

摘要

The rotor size of wind turbines is expected to increase continuously in future. In order to account for the substantial size and severe external loading conditions, load calculation of wind turbine systems should be effective and as well as conservative. However, overly conservative load calculation in blade design would lead to an over-sized design, increasing the cost of the entire system. Therefore, an appropriate design load calculation method needs to be selected carefully. The IEC 61400-1 standard provides the stochastic turbulence wind models to have a specified mean wind speed at hub height and a turbulence intensity, and to be generated with different initial values ("seeds") for producing the turbulent wind field. As different seed values of the turbulent wind model could lead to considerable variations in the results in extreme analysis, seed selection for turbulent wind generation is an important part of the design load case (DLC) setup procedure. In this paper, a turbulent wind seed selection method is analyzed using robust statistics, which provides an alternative approach to standard statistical methods in order to be not affected by outliers or other minor departures from model assumptions. For the design load simulation of turbulent wind at various wind speeds, 15 wind fields with different random seeds were generated for each wind speed. Based on the robust z-score calculated with blade bending moment and tip deflection, the three turbulent wind seeds were chosen for blade design load simulations and compared with the results of the classical statistical method.