摘要
针对产品故障分析与维修过程过于依赖专家知识,而没有充分利用历史故障数据的问题,为了提高产品故障处理效率和水平,文中从产品故障处理过程的流程决策和知识建模视角出发,利用数据挖掘技术,提出一种基于关联矩阵的产品故障处理知识服务模型,用以辅助维修人员进行产品故障处理决策,以提高产品故障处理效率与水平。首先,利用故障处理流程、故障处理内容及人工经验知识,构建包含故障件、故障现象、故障原因、技术方案的故障处理知识模型。针对产品故障文本的特性,综合使用FP-Growth算法及K-means聚类算法进行知识提取。基于词袋模型,采用词频统计方法构建故障件与故障现象之间的关联矩阵,运用模糊分析法构建产品故障现象与故障原因之间的关联矩阵,并在此基础上构建故障原因与技术方案之间的关联矩阵,向故障处理人员提供相应的技术方案组合,进而对整个故障处理过程提供相应的知识服务,最后采用汽车轮胎故障处理案例验证了该模型的有效性。
Since the process of product-fault analysis and maintenance relies too much on knowledge instead of making full use of historical fault-based data,in this article,in order to improve the efficiency and level of handling product faults,a model of knowledge service is set up based on the incidence matrix,from the perspective of process decision and knowledge modeling,It is used to assist maintenance personnel to make decisions on handling product faults,so as to improve the efficiency and level in this aspect.Firstly,the knowledge-service model is constructed by means of fault-handling process,fault-handling content and artificial expertise,including fault parts,fault phenomena,fault causes and technical solutions.The analysis is conducted on the characteristics of those tests with product faults,and the FP-Growth and K-means clustering algorithms are used to extract knowledge.Based on the bag of word model,the correlation matrix between the fault parts and the fault phenomena is constructed by means of the statistical method of word frequency;the correlation matrix between the fault phenomena and the fault causes is constructed with the help of the fuzzy analysis method.On this basis,the correlation matrix between the fault causes and the technical solutions is worked out,so as to provide technical solutions for the fault-handling personnel.Then,the knowledge service is provided for the whole process of fault handling.Finally,Some cases of fault handling for auto tires has verified that the model is effective.
作者
钟武昌
战洪飞
余军合
王瑞
邓慧君
ZHONG Wu-chang;ZHAN Hong-fei;YU Jun-he;WANG Rui;DENG Hui-jun(Faculty of Mechanical Engineering&Mechanics,Ningbo University,Ningbo 315211;Faculty of Electrical Engineering and Computer Science,Ningbo University,Ningho 315211)
出处
《机械设计》
CSCD
北大核心
2022年第3期74-82,共9页
Journal of Machine Design
基金
国家重点研发计划项目课题(2019YFB1707101,2019YFB1707103)
国家自然科学基金(71671097)
浙江省公益技术应用研究计划项目(LGG20E050010,LGG18E050002)
宁波市自然科学基金项目(2018A610131)。
关键词
产品故障
数据挖掘
关联矩阵
知识服务
product fault
data mining
incidence matrix
knowledge service
作者简介
钟武昌(1993-),男,硕士,研究方向:知识管理、数据分析。E-mail:1819823393@qq.com;通信作者:战洪飞(1970-),男,教授,博士,研究方向:知识管理、企业信息化。E-mail:zhanhongfei@nbu.edu.cn。
