面向山区公路地质灾害监测的三维无线传感器网络层次化数据融合模型研究

Data aggregation is the key technology in geological hazard monitoring system of mountainous highway. Aiming at the defects of current data aggregation mechanism, based on the multi-source information fusion model with Wireless Sensor Networks (WSNs), this project will explore 3D space model for wireless sensor networks, which is suitable for the complex mountain terrain & landform of mountainous highway; and then establishe efficient hierarchical topology. .Over the few recent years, distributed compressed sensing (DCS) theory has received increasing attention in wireless sensor networks. However, due to the intra- and inter-signal correlation of geological hazard monitoring sensor data, this project will construct a novel hierarchical and distributed data aggregation model using distributed compressed sensing approach, which includes sparse representation, designing of measurement matrix, joint sparse model and exact signal reconstruction, where time-spatial correlation and hierarchical topology will be exploited between sensor nodes by non-cooperative way. Undoubtedly, the success of such a DCS approach, especially for dynamic real-time applications, such as geological monitoring of mountainous highway, and will solve many problems including accuracy, robustness, network delay and data congestion, etc. .This project will provide a novel idea for geological hazard data fusion. Moreover, the broader impact of the proposed research has extremely important theoretical and application values for assessing geological hazard risk, predicting time-space forecast and disaster prevention reduction along the highway.

数据融合是山区公路地质灾害监测体系中的关键技术。本项目针对现有数据融合机制中存在的缺陷，基于无线传感器网络的多源信息感知模式，首先根据复杂的山区地形地貌，研究面向山区公路部署的三维无线传感器网络空间模型，建立高效的层次化网络拓扑结构；在此基础上结合新兴的分布式压缩感知理论，利用地质灾害监测感知数据的时空相关性，围绕感知数据的稀疏化表示、观测矩阵的建立、联合稀疏模型和重构恢复算法四个关键问题，构建层次化的分布式数据融合模型，实现网内监测节点以非协作的方式进行数据感知与融合，建立面向山区公路地质监测的新型时空数据融合与处理模型，解决已有数据融合模型在精确度、鲁棒性、网络延迟和通信拥塞等方面存在的问题。通过开展本项目的研究，不仅可以为地质灾害数据融合机制的研究提供新的学术思路，而且对于公路沿线地质灾害的危险性评价、时间与空间预测预报和岩土工程防灾减灾等方面同样具有十分重要的理论和应用价值。

数据融合是山区公路地质灾害监测体系中的关键技术。本项目针对现有数据融合机制中存在的缺陷，基于无线传感器网络的多源信息感知模式，首先根据复杂的山区地形地貌，研究面向山区公路部署的无线传感器网络空间组织和演化模型，设计高效的层次化网络拓扑结构；在此基础上结合新兴的分布式压缩感知理论，利用地质灾害监测感知数据的时空相关性，围绕感知数据的稀疏化表示、观测矩阵的建立、联合稀疏模型和重构恢复算法四个关键问题，构建层次化的分布式数据融合模型，实现网内监测节点以非协作的方式进行数据感知与融合，建立面向山区公路地质监测的新型时空数据融合与处理模型，解决已有数据融合模型在精确度、鲁棒性、网络延迟和通信拥塞等方面存在的问题。通过开展本项目的研究，不仅可以为地质灾害数据融合机制的研究提供新的学术思路，而且对于公路沿线地质灾害的危险性评价、时间与空间预测预报和岩土工程防灾减灾等方面同样具有十分重要的理论和应用价值。