基于小波方法的空间风速场模拟

Spatial Wind Field Simulation Based on Wavelet Method

摘要:

为克服随时间变化模型基于适应过滤器和窗口估计方法中窗口长度不宜过大,产生的短时时间序列模拟准确性不够高的缺点,应用小波分解方法,结合线性滤波器法的向量过程自回归(AR)模型,给出了模拟空间脉动风场的一种新方法.该方法对于AR模型自回归系数在空间上进行小波扩展,采用最小二乘法来估计AR模型的自回归系数,并给出了该方法模拟空间风速场的实现步骤.将该方法应用于一空间结构的风速场模拟,并给出了模拟结果与目标值的对比,以及与向量过程AR模型模拟结果的对比,结果证实该方法可以减少风速时程分析在频域上的信息损失,对短时时间序列模拟具有较高的准确性,并具有较高的计算效率.

Abstract:

To overcome inadequacy of timevarying models based on adaptive filters and windowed estimation for short time series,a wavelet decomposition method is employed to present a new method for simulating fluctuating wind field,combined with vector AR model of linear filter method.Wavelet expansion is performed for autoregressive coefficients of AR model in adequate space,adopting least square method to estimate the above autoregressive coefficients.Simulating procedure for spatial wind field based on the method is given.The method is applied to simulating the fluctuating wind field of a spatial structure.And comparison between the simulating results and target values,results obtained from vector AR model are made.The results show that the proposed method diminishes information loss in frequency domain,and can accurately simulate short time series with a fairly high calculating efficiency.