结构健康监测中基于压缩传输与仿生修复的WSN故障诊断研究

Structural health monitoring (SHM) based on wireless sensor networks (WSNs) is a current research focus. As one of the key technique involved in WSNs, timely and accurate diagnosis of nodes and network failures is an important guarantee to improve the system reliability. For the application characteristics of WSNs in the areas of composite SHM, we will propose the fault diagnosis and self-repair mechanisms which are suitable for this particular field. Since distinguishing between sensor fault and structural damage can utilize the fact that the sensor faults are local, while structural damage is global. A distributed fault-tolerant algorithm Based on Compressive sampling and Multiflow transmission, which can improve the detection accuracy of the structural damage, can be applied for failure diagnosis. In wireless sensor network, since the sensor node's hardwares are fixed, the node can not be re-configured. Using the bio-inspired hardware mechanisms, we can deal with the above problem. The self-repairing theories and methods are studied acorrding to the node failure, part failure of the network and nodes invasion. The content research are include: the Research of efficient and stable multiflow transmission mechanism, compressive sampling in sensor nodes, fault model of the sensing components, bio-inspired hardware mechanism design of wireless sensor nodes and self-reparing technology, the fault diagnosis method of the node and network. Through the research of this project, we can improve the robustness and security of WSNs in SHM, and can enhance the self-repairing and fault tolerance of the system.

基于无线传感网络的结构健康监测技术是当前研究的重要方向，及时准确的对节点和网络的故障进行诊断是提高系统可靠性的重要保证。本项目针对无线传感器网络在复合材料结构健康监测领域中的应用特点，提出一种基于压缩感知和多流传输相结合的无线传感网的故障诊断和仿生修复机制。在可靠多流传输机制下，分析压缩采样数据，识别故障节点的异常数据和结构损伤采集样本的本质区别，实现节点软硬件故障的容错算法，从而提高材料损伤的识别精度。针对现有无线传感器网络中节点一旦失效只能舍弃的问题，尝试将仿生硬件应用于无线传感节网络中，使节点和网络在发生故障时具有自修复和自重构的能力。项目主要研究内容包括：高效稳定的多流传输机制、节点压缩采样机制、传感部件故障模型的研究、自修复节点的设计、节点和网络故障诊断和自修复机制。通过课题的研究，能提高无线传感器网络应对等异常事件的自修复和容错能力，提高无线传感器网络的鲁棒性和安全性。

无线传感网络中及时准确的对节点和网络的故障进行诊断是提高系统可靠性的重要保证。本项目设计了一种基于压缩感知和多流传输相结合的无线传感网的故障诊断和仿生修复机制。本项目的主要研究内容可概括为四点：1）SHM中高效实时路由机制的研究；2）SHM中故障节点的异常数据对结构损伤测量值的干扰处理；3）SHM中基于压缩感知的无线传感器节点的软故障诊断机制；4）SHM中无线传感器节点的硬故障诊断机制及自修复节点的设计。在本项目的研究过程中，通过建立节点的故障模型，并针对无线传感网络中节点故障模型展开了深入的研究，从而分析得出了相应的故障诊断方法。随后，通过分析压缩采样数据，识别故障节点的异常数据和结构损伤数据的本质区别，实现节点软硬件故障的容错算法，从而提高材料损伤的识别精度。然后，对无线传感网络中的压缩感知和可靠传输协议进行了详细地研究，设计了基于仿生硬件的节点的自修复机制，并通过实验成功模拟了无线传感网络中节点损伤的实时监测。最后，项目组建立了自修复节点及自修复无线传感网络的性能评价方法，实现了无线传感网中故障诊断方法的性能的评估。在本项目的研究中，项目组已在国内外学术期刊上发表SCI(E)收录论文14篇，EI收录的论文3篇，中文论文3篇，会议论文1篇，申请发明专利6件，获批软件著作权5件，培养已毕业研究生8名。本项目的研究成果可以弥补现有的无线传感网络通信协议的诸多缺陷，保障协议的安全性、可靠性已经高效性，为相关领域的理论研究和实际应用提供重要的参考价值。.