﻿ 基于数据挖掘的船舶通信网络失效节点自动识别方法
 舰船科学技术  2022, Vol. 44 Issue (19): 146-149    DOI: 10.3404/j.issn.1672-7649.2022.19.029 PDF

1. 江苏师范大学 计算机科学与技术学院，江苏 徐州 221000;
2. 中国劳动关系学院 应用技术学院，北京 100048

Automatic identification method of failure node of ship communication network based on data mining
CHEN Wen-qing1,2
1. School of Computer Science and Technology, Jiangsu Normal University, Xuzhou 221000, China;
2. School of Applied Technology, China University of Labor Relations, Beijing 100048, China
Abstract: The current proposed method for automatic identification of failed nodes in ship communication network is time-consuming, and redundant nodes have a low shutdown rate. In order to solve the above problems, a new method for automatic identification of failed nodes in ship communication network is studied based on data mining. Determine the target data set, obtain the feature extraction optimal solution set and the overall optimal solution set, identify and extract the abnormal data features, establish the abnormal data clustering set, and introduce the data density coefficient to calculate the optimal value of the node data clustering result. The automatic identification function of the failed node is constructed, the adaptive objective function of the retrieval radius is obtained, the data mining is carried out, and the automatic identification of the network failure node is realized through unified processing. The experimental results show that the time-consuming time of the automatic identification method of the failure node of the ship communication network based on data mining is less than 4 s.
Key words: data mining     ship communication     communication network     failed node     automatic identification     identification method
0 引　言

1 船舶通信网络失效节点自动识别方法具体设计 1.1 节点特征提取

 ${e_i} = \alpha * {r_1}\left( {{u_{mm}} - {e_{im}}} \right) + \beta * {r_2}\left( {{u_{am}} - {e_{im}}} \right)。$ (1)

1.2 节点分类检测

 $L\left( 1 \right) = \sum\limits_{i = 1}^K {{\mu _{ij}}{L_j} + } \sum\limits_{j = 1}^{K'} {{\mu _{ij}}{L_i}'} 。$ (2)

 ${X_i} = - \delta \frac{{{{\left\| {{x_1} - {x_2}} \right\|}^2}}}{{({r_b}/2)}}。$ (3)

 图 1 聚类后的通信网络节点 Fig. 1 Communication network nodes after clustering

2 船舶通信网络失效节点数据挖掘

 图 2 船舶通信网络失效节点数据挖掘流程 Fig. 2 The data mining process of the failure node of the ship communication network
2.1 失效节点自动识别函数

 $F\left( {x\left( t \right)} \right) = \sum\limits_{i = 1}^n {\gamma {S^2}{x_i}\left( t \right) + \lambda S{x_i}\left( t \right)}。$ (4)

 $f\left( {R,\phi } \right) = \frac{1}{{1 + {e^{ - F\left( {{x_i}\left( t \right)} \right)}}}}。$ (5)

3 实验结果与分析

 图 3 通信网络原始节点分布状态 Fig. 3 Distribution of original nodes of communication network

 图 4 通信网络原始节点识别结果 Fig. 4 Identification results of original nodes of communication network

 图 5 识别耗时实验结果 Fig. 5 Recognition time-consuming experiment results
4 结　语

1）通过聚类算法对船舶通信网络中的海量数据进行识别提取，根据时间序列划分数据节点，引入优化系数提高特征提取的相关性，再通过聚类运算对节点特征数据进行分类检测，多次迭代和密度系数检验也有利于进一步提高聚类分析的效果。

2）以聚类结果为初始数据构建熵目标函数，在自动检索半径内对异常数据进行自动识别检测，具有良好的灵活性。

3）采用离散序列算法对通信网络失效节点进行数据挖掘，能够提高数据处理效率和精准度。

 [1] 高宇鹏, 吕淑芳. 舰船通信网络脆弱节点智能定位方法[J]. 舰船科学技术, 2021(3): 103-105. GAO Yupeng, LV Shufang. Intelligent location method of vulnerable nodes in ship communication network[J]. Ship Science and Technology, 2021(3): 103-105. [2] 王盟. 舰船无线通信网络故障节点自动定位方法[J]. 舰船科学技术, 2020, 42(20): 98–100. WANG Meng. Automatic location method of fault node in ship wireless communication network[J]. Ship Science and Technology, 2020, 42(20): 98–100 [3] 刘国强, 林叶锦, 张志政, 等. 基于粗糙集和优化DAG-SVM的船舶主机故障诊断方法[J]. 中国舰船研究, 2020, 15(1): 68–73. LIU Guoqiang, LIN Yejin, ZHANG Zhizheng, et al. Fault diagnosis method of marine main engine based on rough set and optimized DAG-SVM[J]. China Ship Research, 2020, 15(1): 68–73 [4] 游兰, 韩雪薇, 何正伟, 等. 基于改进Seq2Seq的短时AIS轨迹序列预测模型[J]. 计算机科学, 2020, 47(9): 169–174. YOU Lan, HAN Xuewei, HE Zhengwei, et al. Short time AIS trajectory sequence prediction model based on improved Seq2Seq[J]. Computer Science, 2020, 47(9): 169–174 [5] 魏国东, 朱石坚, 罗忠, 等. 基于数据挖掘的装备维修性定量分析与研究[J]. 海军工程大学学报, 2020, 32(2): 87–91. WEI Guodong, ZHU Shijian, LUO Zhong, et al. Quantitative analysis and research of equipment maintainability based on data mining[J]. Journal of Naval University of Engineering, 2020, 32(2): 87–91 [6] 李忠, 靳小龙, 王亚杰, 等. 属性网络中基于变分图自编码器的异常节点检测方法[J]. 模式识别与人工智能, 2022, 35(1): 17–25. [7] 刘明. 基于数据挖掘技术的网络控制系统建模与分析[J]. 沈阳工业大学学报, 2022(4): 446–450. LIU Ming. Modeling and Analysis of Networked Control System Based on Data Mining Technology[J]. Journal of Shenyang University of Technology, 2022(4): 446–450 [8] 张磊, 吴昊, 赵强, 等. 基于数据挖掘技术的地下工程目标毁伤效应计算方法[J]. 爆炸与冲击, 2021, 41(3): 1–10. [9] 任伟, 徐子立, 宋晓林, 等. 基于数据挖掘的配网数字化计量系统运行特性监测和评价方法研究[J]. 高压电器, 2020, 56(8): 183–191. [10] 韩建富, 肖春, 宋小兵, 等. 基于GA-BP神经网络的能源互联网窃电行为识别方法[J]. 电气传动, 2022, 52(14): 38–44.