基于VDM的复杂装备故障数据分析

在历史故障数据基础上，分别掏建故障信息、装备信息、装备使用状况的数据集市，通过数据挖掘模型和可视化交互平台，形成基于可视化数据挖掘（Visual Data Mining，VDM）故障分析架构。以此架构为基础，探讨了故障数据仓库的建立以及故障数据挖掘算法，对实际故障数据进行预处理，并分析故障责任、故障趋势、故障分布等特性，得到图形化的分析结论，为装备的设计、制造、维护提供支持。

Failure Data Analysis of Complex Equipment Based on Visual Data Mining

On the basis of historical Complex Equipment＇s Fault Record, data market of fault information, equipment information and equipment state information are established separately. A fault analysis frame is founded by adding data mining model and visual communion plat- form on these data markets. Failure data warehouse and data mining algorithms are described here. Using the frame, this paper does some pretreatment to the real complex equipment＇s fault record, and analyzes the fault property such as responsibility, trend and distributing. The visual analysis results can provide effective advice on design, manufacture and servicing.