海量船舶故障数据挖掘中的模糊运算聚类算法分析  

Analysis of fuzzy operation clustering algorithm in massive ship fault data mining

作　　者：魏亚清 夏少芳 胡秀源 WEI Ya-qing;XIA Shao-fang;HU Xiu-yuan(University of the Cordilleras,Baguio 006374,Philippines;School of Mathematics and Information Technology,Xingtai University,Xingtai 054001,China)

机构地区：[1]凯迪雷拉大学,菲律宾碧瑶006374 [2]邢台学院数学与信息技术学院,河北邢台054001

出　　处：《舰船科学技术》2022年第13期182-185,共4页Ship Science and Technology

基　　金：邢台市科技局项目(2019ZC009)。

摘　　要：为能获取可以准确描述船舶运行状态的故障数据,提出海量船舶故障数据挖掘的模糊运算聚类算法。使用局部切空间排列算法从船舶运行数据提取船舶故障征兆变量,利用离散化算法完成离散化处理运用模糊运算聚类算法挖掘与各故障征兆变量离散化结果相匹配的船舶故障数据。实验结果表明,该方法具有较优良的船舶故障征兆变量降维效果,海量船舶故障数据挖掘性能较为理想。In order to obtain the fault data that can accurately describe the ship operation status, a fuzzy operation clustering algorithm for massive ship fault data mining is proposed. The local tangent space permutation algorithm is used to extract the ship fault symptom variables from the ship operation data, the discretization algorithm is used to complete the discretization processing, and the fuzzy clustering algorithm is used to mine the ship fault data matching the discretization results of each fault symptom variable. The experimental results show that this method has a better effect of reducing the dimension of ship fault symptom variables, and the performance of massive ship fault data mining is ideal.

关 键 词：海量船舶故障数据挖掘 故障数据挖掘 模糊运算聚类 局部切空间 故障征兆变量 可辨识矩阵 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]