中文摘要：

在露天矿及排土场的边坡监测中由于放炮震动等因素影响使得测量机器人监测基础数据中总是含有一定量的粗差。针对如何快速而准确的剔除监测数据中的粗差，文中提出利用小波阈值降噪的方法剔除数据中的粗差，该方法通过有用信号最小频率确定分解层数，根据确定的分解层数对含有粗差的监测基础数据进行软阈值消噪处理。该方法取得良好效果，为基于测量机器人监测基础数据的粗差剔除提供了一种有效的方法。

英文摘要：

There is always a certain amount of gross error contained in surveying robot monitoring basic data in openpit and dump slope monitoring. The wavelet threshold de- noising method is proposed to eliminate gross error con-tained in monitoring data quickly and accurately in this paper. The method determines the decomposition level by theminimum frequency of the useful signals, and then the monitoring basic data with gross error contained in is sup-pressed with soft threshold de - noising method according to the determined decomposition level. This method isproved to be effective for the elimination of monitoring basic data based on the surveying robot.