摘要
传统船舶的故障数据自动分类方法,存在故障数据类型定义不准确、分类时间过长等弊端。为有效解决上述问题,设计基于关联规则的新型船舶故障数据自动分类方法。通过船舶故障数据的采集及预处理、数据的进一步挖掘两大步骤,完成关联规则下的船舶故障数据感知。通过BP自动分类神经网络设计、船舶故障数据的归一化处理、HIWO自动分类算法设计三大步骤,完成新型船舶故障数据自动分类方法的搭建。设计对比实验结果表明,新型船舶故障数据自动分类方法,与传统方法相比,可以在提升故障数据类型定义准确性的同时,有效控制分类时间。
The automatic classification method of traditional ship′s fault data has some disadvantages,such as the inaccurate definition of fault data type and the long classification time. In order to solve the above problems, a new automatic classification method of ship fault data based on association rules is designed. By the two steps of collecting and preprocessing the data of the ship′s fault and mining the data, the ship fault data perception under the association rules is completed. Through BP automatic classification neural network design, ship fault data normalization processing, HIWO automatic classification algorithm design three steps,we completed the new ship fault data automatic classification method. The design comparison experiment results show that compared with the traditional methods, the new automatic classification method of ship fault data can improve the accuracy of fault data type and control the classification time effectively.
出处
《舰船科学技术》
北大核心
2018年第6X期55-57,共3页
Ship Science and Technology
关键词
关联规则
故障数据
自动分类
数据预处理
数据挖掘
神经网络
归一化
HIWO
association rules
fault data
automatic classification
data preprocessing
data mining
neural network
normalization
HIWO
