根据灰色模型建立和检验理论,针对信息量少的变形数据进行预测这一特性,采取建立GM模型进行预测,同时考虑卡尔曼滤波的优点,提出了基于卡尔曼滤波的GM模型的建立及相应的精度评定,结合实例来说明并对其进行分析预测。数据处理结果显示,本模型有效地剔除观测数据粗差,精度较高,为变形观测研究提供了更为可靠的观测数据。 According to the gray model establishment and test theory, the deformation data with small amount of information is predicted. The GM model is established for prediction. At the same time, the advantages of Kalman filter are considered. The establishment of GM model based on Kalman filter and corresponding Accuracy assessment, combined with examples to illustrate and analyze the forecast. The data processing results show that this model effectively eliminates the gross errors of the observed data and has higher accuracy, which provides more reliable observation data for the deformation observation research.