含有扰动的无线传感器网络故障检测  

Fault Detection of Wireless Sensor Network with Disturbance

作　　者：张鹏[1] 杜文祥 邸希元 ZHANG Peng;DU Wenxiang;DI Xiyuan(Harbin University of Science and Technology, Harbin 150080, China)

机构地区：[1]哈尔滨理工大学自动化学院,黑龙江哈尔滨150080

出　　处：《探测与控制学报》2020年第4期54-58,共5页Journal of Detection & Control

摘　　要：针对使用定性或定量方法对含有扰动的无线传感器网络(WSN)进行故障检测时无法区分数据波动是由故障还是扰动引起的问题,提出了含有扰动的无线传感器故障检测方法:置信规则库(BRB-r)。该方法既包含定性又包含定量知识,融合了属性可靠性以及属性权重,在噪声干扰下能够有效判断数据波动是由噪声还是故障导致的,可以更加完美地解决外界不可靠因素对我们进行故障检测等研究的影响。仿真实验结果表明,BRB-r模型能够在WSN处于扰动状态下对WSN是否出现故障进行判断,相较于BRB模型检测准确率更高,稳定性实用性更强。Aiming at the problem of using qualitative or quantitative methods for fault detection of wireless sensor networks(WSN)with disturbances to distinguish whether data fluctuations were caused by faults or disturbances,a fault detection method for wireless sensors with disturbances was proposed,which as the confidence rule base(BRB-r).This method contained both qualitative and quantitative knowledge.It combined attribute reliability and attribute weights.Under noise interference,it could effectively determine whether data fluctuations were caused by noise or faults.It could better solve external unreliable factors and perform fault detection for us.Simulation results showed that the BRB-r model could judge whether the WSN was faulty when the WSN was in a disturbance state.Compared with the BRB model,the detection accuracy was higher and the stability was more practical.

关 键 词：无线传感器网络 故障检测 置信规则库 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]