顾及未知系统误差的变形监测滤波算法

来源期刊：中南大学学报(自然科学版)2011年第3期

论文作者：左廷英 曾磊

文章页码：738 - 743

关键词：系统误差；边坡；变形监测；卡尔曼滤波

Key words：systematic errors; slope; deformation monitoring; Kalman filtering

摘    要：通过建立一种带有未知系统误差的滤波模型，研究变形监测滤波解算中一些未知边坡物理信息的处理问题，给出一种基于移动窗口的系统误差自适应拟合法。研究结果表明：该算法充分融合了包含在观测中的统计信息、边坡滑坡的有关力学状态和地质条件所提供的信息，减弱了观测异常所带来的影响, 提高了形变参数解算精度。

Abstract: By constructing a filter models with unknown systematic error, slope’s physical information disposal was studied in the course of deformation monitoring filter calculation, and an adaptive fittings algorithm of systematic errors was given using moving windows. The results show that the method is fully fused with statistic information in the course of observation, dynamical state of slope slippage and geophysical information, it can eliminate the influence of abnormal observation and improve the precision of deformation parameter estimation.