基于自回归法的滑坡监测数据Kalman滤波的应用

Application of Kalman filtering to the landslide monitoring data procession based on auto-regression method

介绍了卡尔曼滤波的基本模型，针对运动学方法以监测点位置参数、速率参数为状态向量，以加速度为噪声向量建立观测方程和状态方程的过程比较复杂。把自回归法引入状态方程和观测方程的建立中，并结合某露天矿滑坡动态监测数据分析了模型的应用。结果表明，滤波值和观测值之间差值保持在1～3cm之间，滤波后图像较原观测值图像更为光滑，表明所建模型是合适的，能够反映滑坡动态变化过程，从而为矿滑坡监测预报提供数学工具。

It introduces a basic model of Kalman filtering. The kinematic method establishes the observation equation and state equation based on location parameters and rate as the state vector, acceleration as the noise vector. The auto-regression method is introduced into the establishment of the state equation and observation equation, simply compared with the kinematic method. Taking one opencast mine landslide dynamic monitoring data as an example, it analyzes the model application. The results show that the deviation of filtering value and observed value is between 1 to 3cm, and the filtering image is smoother than the original observation image. It shows that the model is appropriate, and it can reflect the dynamic change process of landslide. So the method provides a powerful mathematical tool {or this opencast mine landslide monitoring and prediction.