论文部分内容阅读
传统的传感器节点故障诊断模型通常采用分布式模型或集中式模型,因此具有诊断效率低和扩展性差的缺点,为此,提出了一种基于分簇路由协议并结合集中式故障诊断和分布式故障诊断的混合式故障诊断模型。首先,在监测区域建立层次分簇路由协议和WSN节点故障诊断模型,然后,采用ICA独立成分分析法(Independent component analysis,ICA)对特征向量数据属性提取独立成分,以降低数据维数,从而获得最小属性集样本数据。最后,通过赋予各属性权值对朴素贝叶斯分类模型进行改进,得到加权依赖贝叶斯分类模型,并采用此模型实现节点故障诊断。仿真实验证明该模型能有效地进行故障诊断,与其他方法相比,具有故障诊断效率高和诊断精度高的优点,具有很强的可行性。
Traditional sensor node fault diagnosis model usually adopts distributed model or centralized model, so it has the disadvantage of low diagnostic efficiency and poor scalability. To solve this problem, a clustering routing protocol is proposed combining with centralized fault diagnosis and distributed fault Diagnostic Hybrid Fault Diagnosis Model. Firstly, the Hierarchical Clustering Routing Protocol (WS-Hierarchy Routing Protocol) and WSN node fault diagnosis model are established in the monitoring area. Independent component analysis (ICA) is then used to extract the independent components of the eigenvector data attributes to reduce the data dimension to obtain Minimum attribute set sample data. Finally, we give a weight-dependent Bayesian classification model by improving the naive Bayesian classification model, and use this model to realize the node fault diagnosis. Simulation results show that this model can effectively diagnose faults. Compared with other methods, this model has the advantages of high efficiency of fault diagnosis and high accuracy of diagnosis, which is very feasible.