当观测数据含有粗差时使用单一的GM(1,1)模型不能取得很好的精度,本文在卡尔曼GM(1,1)基础上提出基于卡尔曼GM(1,1)滤波的稳健估计模型。通过对大坝变形监测数据的处理结果可以看出卡尔曼GM(1,1)稳健估计具有更高的精度和效果。
Estimate accuracy can' t be enough by using a single GM ( 1, 1 ) model when the observation has gross errors. In this paper, for the purpose of establishing a steady model, a robust estimation model based on the Kalman-GM (1, 1) filter is presented. The robust estimation model is verified by using a set of dam deformation monitoring dada. The results show that the new robust estimation method has higher precision and better effect.