中文摘要：

在基于无线传感器网络的结构健康监测中，故障节点产生并传输的错误数据将消耗节点的能量和带宽，同时会形成结构健康监测的错误决策。针对该问题，提出一种节点故障诊断算法，利用节点感知传感数据的空间相似性，通过对邻节点所感知的传感数据进行比较，确定检测节点的状态，将测试状态向网络中其他相邻节点进行扩散，实现网络中全部节点的故障检测。同时介绍在不同平均节点度下改善故障诊断率的方法。实验结果表明，该算法能够获得较好的性能。

英文摘要：

In structural health monitoring of Wireless Sensor Network（WSN）, a faulty sensor may produce incorrect data and transmit them to the others. It consumes the limited energy and bandwidth of WSN. The base station may make inappropriate decision when it receives the incorrect data. Aiming at this problem, a distributed fault detection algorithm for WSN is presented. Faulty sensor nodes are identified based on comparisons between neighboring nodes and dissemination of the decision made at each node. The whole node detection is implemented. The method to improve fault diagnosis rate in different average node degree is proposed. Experimental results show that sensor nodes with permanent faults are identified with high accuracy for a wide range of fault rates.