利用本征正交分解(Proper orthogonal decomposition,POD)原理设计了一种针对风力机翼型动态失速的时变过程的辨识方法。首先对周期俯仰运动的风力机翼型流场的动态失速过程进行数值模拟,然后用上述方法进行了有效的辨识并从中提取了关于动态失速过程的主要模态信息。在给定的误差阈值下,分别针对浅失速和深失速的情况,将该降阶模型的辨识结果与数值计算原始结果进行了对比并对误差进行了相应的分析。结果表明,该降阶模型方法能够以明显降低的计算量精确辨识翼型的浅失速情况;对深失速的辨识会由于湍流模型的精度影响有所降低。
A reduced-order modeling method is developed based on the proper orthogonal decomposition(POD) strategy.The low-speed incompressible air flow around a periodically pitching wind turbine airfoil is identified with the above method.The information from dominant modes in the dynamic stall procedure is then extracted from the results of 2-D Reynolds-averaged Navier-Stokes(RANS) computation.The identification results are compared with the RANS data under the pre-specified threshold both for light stall and deep stall cases,and the error analysis is discussed as well.It is shown that the POD methods accurately identified the stall features of light pitching airfoil flow.However,the performance of the above methods may be affected for the deep stall case,due to the inaccuracy of turbulence models in the RANS solver.