随着人类建设活动的日益频繁，天然边坡人工扰动诱发的滑坡灾害逐渐增多，已经严重威胁到人员和工程构筑物的安全与稳定，滑坡监测变得越来越重要。介绍了一种经过改进后的滑坡监测预警系统，该装置基于北斗卫星和通用分组无线服务技术（GPRS）双路通讯平台，采用zigbee技术构建多跳自组织无线传感器网络，利用具有能量吸收特性的恒阻大变形锚索替换原来的常规锚索作为传力装置，保证边坡岩体变形2 m锚索不被拉断，实现了对大变形滑坡灾害的全过程监测预警目标。另外，按照红、橙、黄、蓝4个颜色等级，建立了临滑、近滑、次稳定和稳定4级预警准则，并开发出3D自动搜索和处理软件系统，嵌入监测区地形地貌遥感图、地质断面图、监测点分布图等矢量地图，具有自动搜索、监测数据智能处理、监测曲线自动显示和监测信息快速查询等功能。该系统已在南芬露天铁矿推广应用，根据预警准则成功对2011年10月5日发生的滑坡灾害提前5 d发出预警信息，并结合降雨量和累计采矿量监测数据，耦合对比分析滑坡前、后各参数的演变特征，证明滑动力监测参数是滑坡超前预警的惟一有效参数。该系统的成功应用为其他类似边坡的超前监测预警提供了理论和实践依据。

Due to the rapid development of modern construction activities, the landslide disaster induced by artificial disturbance is increasing gradually. Therefore, landslide monitoring is becoming more and more important to ensure the safety of human and stability of engineering structures. An improved system is introduced to monitor and to pre-warn the whole process of large deformation landslide disaster. This system based on two-way communication platform includes Beidou satellite and general packet radio service (GPRS). The zigbee technology is applied to build a wireless multi-hop self-organizing sensor network. The new anchor cable with constant resistance and large deformation functions, which can absorb energy, is used to replace the common anchor cable as the transmission device. Thus two-meter deformation of slope rock mass is allowed without tensile failure of energy absorption anchor cable. In addition, four level warning criteria are established according to four color grades (i.e. red, orange, yellow and blue). 3D automatic searching and processing software system is also developed. Vector maps, such as topography remote sensing map, geologic profile map and distribution map of monitoring points, are embedded in the monitoring area. Therefore, this new system has benefits in: searching area automatically, processing data intelligently, displaying monitoring curve automatically, and querying monitoring information quickly. The system has been used in Nanfen open pit iron mine. According to the warning criteria, the warning information of the landslide on October 5, 2011 was sent out 5 days in advance. Combined with the monitoring data of rainfall and the cumulative amount of mining, it is proved that sliding force is the only effective parameter for landslide pre-warning, which results from the coupling analysis of the evolution of the various parameters before and after the landslide. The successful application of this system provides theoretical and practical basis for monitoring and pre-warning other similar slopes.