Kalman滤波用于大坝位移模拟与预报

Simulation and Prediction of Dam Displacement Based on Kalman Filter

为降低随机干扰因素对大坝位移观测数据的影响,针对大坝位移特性并基于Kalman滤波理论和统计模型建立了大坝水平位移的状态方程和观测方程。在Kalman滤波方程组未知参数的确定中以误差方差估计确定白噪声方差,利用多元线性模型求得状态转移矩阵,这样既反映了真实误差也描述了各因素间的复杂关系。应用实例表明,Kalman滤波可显著降低大坝监测中的噪声影响,提高位移统计模型的模拟和预报精度。

Aimed at the displacement characteristic of the dam,based on the theory of Kalman filter,the state and observed equation of displacement on the basis of statistical model is established.The variance of flat noise is confirmed by the error variance estimation.The state transition matrix is defined by the multiple linear model.By application of this method to the example the noise effect during monitoring the dam is evidently reduced and the analogue and forecast precision of statistical displacement model is raised.