中文摘要：

泥石流在我国广泛存在，每年造成的直接经济损失达20亿元，但目前国内外对其研究较少，且尚无完善的监测系统。为实现泥石流的定量在线实时监测，采集一片区域所有点的参数（场测量），提出了一种采用电学CT技术的新型泥石流传感网络。采用不同成像算法进行了仿真对比，总结出了适合泥石流监测的Landweber算法。采用ECT技术模拟了泥石流的发生过程，分别探究了不同降雨、地形、地质条件对ECT成像结果的影响，并对比了不同条件下的雨水扩散速度和小部分土体分离时间。结果表明，降雨是导致泥石流发生的直接因素，坡度越大，泥石流越易发生；沙石比例为6：4左右时最易发生泥石流。该系统突破了传统的点测量模式，实现了非接触非侵入的场测量方式，并融合了多种信息，弥补了现有监测系统难以定量确定多因素权值的缺陷，为泥石流的监测技术提供了新的方向。

英文摘要：

Debris flow is widespread in our country, resulting in huge direct economic loss to the amount of two bil- lion RMB each year. However, there are few studies at home and abroad and no complete monitoring system. To achieve the on-line real-time monitoring of debris flow quantitatively and collect parameters at all points in a area （field measurement）, an innovative sensor network based on electrical CT technology was put forward in this arti- cle. Firstly, different electrical capacitance tomography（ECT） imaging algorithms were used to compare simulation results, concluding that Landweber algorithm is most suitable to debris flow monitoring. Then, the effects of differ- ent conditions （rainfall, topography and geology） on the results of ECT imaging were explored. In addition, the diffusion rate of rainwater and separation time of small portions of soil were compared respectively under different conditions. The conclusion is that rainfall is the direct factor leading to debris flow. And the greater the slope, the easier debris flow happens. When the ratio of sand and stone is 6：4 approximately, debris flow happens most easily. This system breaks through the traditional point measurement mode and achieves a non-contact and non-invasive field measurement. Multiple information is in fusion to compensate the defect that the existing monitoring system is difficult to determine muhifactor weights and the system presented provides a new direction in debris flow monitoring technology.