多层传感器故障数据的挖掘模型仿真  

Multiple Sensor Fault Data Mining Model Simulation

作　　者：徐凯华[1] 赵雪琴[1] 

机构地区：[1]郑州大学西亚斯国际学院,河南新郑451150

出　　处：《计算机仿真》2014年第12期393-396,共4页Computer Simulation

基　　金：河南省科技厅重点科技攻关项目(142102210501)

摘　　要：多层传感器的故障准确定位对保证各自应用安全至关重要。多层传感器不同于传统的传感器网络,其不同层次的传感器故障的特征差异较大,不同层次传感器之间存在故障特征"断层"问题。传统的基于流数据异常特征识别的多层传感器故障数据的挖掘模型需要明确层次网络故障之间的关联特征,若传感器层次之间的故障特征关联性不强,故障挖掘的阀值就无法固定,产生故障特征无法定位问题,导致误警率较高。提出了一种基于贝叶斯信念网络的多层传感器故障数据的挖掘模型,针对多层传感器故障数据属性多样性的问题,分析了贝叶斯信念网络的结构,搜索一个最匹配待分类故障数据样本的贝叶斯信念网络,通过评估函数评估各个可能的网络结构与样本多层传感器故障数据间的契合度,采集一个最佳样本多层传感器故障数据解,通过"压缩侯选"的贝叶斯信念网络算法,计算样本多层传感器故障数据间的依赖关系,集中扫描最可能是待挖掘数据的变量集,实现故障数据的挖掘。实验结果表明,利用所提模型能够有效提高多层传感器故障数据的挖掘的准确性。The accurate positioning of multilayer sensor faults is very important to ensure the safety of their appli- cation. The paper presented a muhilayer sensor based on Bayesian belief network data mining model. Aiming at the problem of multilayer sensor fault data attributes diversity, on the analysis of the structure of bayesian belief network, the paper searched the best match for classification of fault data samples of Bayesian belief network. Through the e- valuation function, the possible network structure was evaluated and the sample fitted between multilayer sensor fault data A best sample of muhilayer sensor data was collected, the sample muhilayer sensor fault data were computed through the ＂ compression candidate＂ algorithm of Bayesian belief network. The dependency relationship between concentration of scanning is most likely to be mining data set of variables, which realizes the fault data mining. The experimental results show that using the proposed model can effectively improve the accuracy of the muhilayer sensor failure data mining.

关 键 词：多层次 传感器 挖据模型 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]