多参量状态下网络节点异常故障诊断仿真

发布时间：2018-09-09 19:01

【摘要】：多参量状态下对网络节点异常故障进行诊断,可以有效保障网络的运行质量。对网络节点异常故障的诊断,需要提取异常节点特征参数,进而设置故障检测阈值,完成网络节点异常故障诊断。传统方法构造一系列具有表征状态能力的特征向量,对不同的特征向量进行分类,但忽略了设置故障检测阈值,导致诊断误差较大。提出基于信息熵的多参量状态下网络节点异常故障诊断方法。该方法先融合于信息熵理论选择受网络状态变化影响较小的信息熵作为故障特征参数提取,按照地理位置进行分簇,在能量不受限的汇节点处进行簇内故障特征再聚类,在此基础上设置故障检测阈值,利用该阈值完成对多参量状态下网络节点异常故障诊断。实验结果表明,所提方法故障诊断精度高,为保障网络的稳定运行质量奠定有力的基础。
[Abstract]:The network node abnormal fault diagnosis in multi-parameter state can effectively guarantee the network operation quality. To diagnose the abnormal faults of network nodes, it is necessary to extract the characteristic parameters of the abnormal nodes, and then set the threshold of fault detection to complete the abnormal fault diagnosis of network nodes. The traditional method constructs a series of feature vectors with the ability to represent the state and classifies the different eigenvectors, but neglects to set the threshold of fault detection, which leads to a large diagnosis error. A method of network node anomaly diagnosis based on information entropy is proposed in this paper. Firstly, the information entropy theory is fused into the information entropy theory to select the information entropy which is less affected by the network state change as the fault feature parameter extraction, then cluster according to the geographical position, and cluster the fault feature in the cluster at the meeting node where the energy is not limited. On this basis, the threshold of fault detection is set, and the threshold is used to complete the fault diagnosis of network nodes in multi-parameter state. The experimental results show that the proposed method has high fault diagnosis accuracy and lays a strong foundation for ensuring the stable operation quality of the network.
【作者单位】：商丘工学院信息与电子工程学院;
【基金】：2016年度河南省科技攻关项目(162102210361)
【分类号】：TP393.06

【相似文献】