基于本征正交分解的平流层风场建模与预测

平流层风场环境对临近空间低动态飞行器设计和轨迹控制具有重要影响。针对平流层风场建模，提出一种基于本征正交分解(POD)的风场数据降阶方法，在此基础上，提出一种可以对平流层风场进行预测的Fourier模型。以长沙地区2005—2009年风场为例，采用提出的POD方法与Fourier预测模型对风场进行建模与预测，并对Fourier预测精度进行分析。研究结果表明，采用POD方法可以对东西方向风场进行高效率高精度降阶建模；通过Fourier预测模型可以对东西方向风场进行准确预测，预测精度与实际风场随时间变化的规律性有关，风场数据越紧凑，周期性越明显，预测精度越高。 

Modeling and prediction of stratospheric wind field based on proper orthogonal decomposition

The stratospheric wind field has an important influence on the design and trajectory control of the near-space low dynamic aerostat. Focused on the modeling of the stratospheric wind field, this paper proposes a reduced order analysis method of the wind field data based on the proper orthogonal decomposition (POD) method. On this basis, a Fourier model which can predict the stratospheric wind field is proposed in this paper. This paper takes the wind field in Changsha from 2005 to 2009 as an example, uses the proposed POD method and Fourier prediction model to model and predict the wind field, and analyzes the accuracy of Fourier prediction model. The results show that the POD method can be used to model the east-west wind field efficiently and accurately. The Fourier prediction model can be used to predict the east-west wind field accurately and the prediction accuracy is closely related to the regularity of the actual wind field. The more compact the wind field data are, the more obvious the periodicity is, the higher the prediction accuracy is.