一种利用低秩矩阵填充技术恢复气象数据的方法

摘要

矩阵填充技术是继研究压缩感知技术之后又一备受关注的新技术,并已被应用于科学与工程等诸多领域.矩阵填充技术主要研究如何通过已知的部分矩阵数据来达到恢复整个矩阵的目的.气象数据为天气预报和开展各种气象服务提供依据,具有空间相关性和时间稳定性,加上气象数据矩阵的低秩性,为研究如何利用低秩矩阵填充技术对气象数据的连续获取提供了条件.最后通过真实气象数据进行了仿真实验,证明了该方法的有效性.%Matrix filling is another attractive new research field after compressive sensing.It has been applied in the fields of science and engineering.Matrix filling principally studies how to restore the whole matrix elements by the known partial matrix elements.The meteorological data have the spatial correlation,time stability,and low rank of meteorological data matrix.All these advantages provide the continuous acquisition of meteorological data for researching how to use the low rank matrix filling technique.Finally,the simulation experiments are performed by real weather data,and the effectiveness of the method is proved.