基于粗集的遗传挖掘在故障诊断中的应用研究

Applied research of data mining in failure diagnosis based on rough sets and genetic algorithms

介绍应用粗集理论和遗传算法相结合进行数据挖掘的方法。利用目前企业采集到的关键设备运行状态的大量数据,首先运用粗集理论的属性约简消去冗余的属性,然后以约简后的数据作为样本训练集,应用优化改进的遗传算法建立分类模型。根据构建的分类模型,可以发现故障设备运行的内在规律,快速对未知故障设备进行归类,从而为故障诊断与故障预测提供决策依据。

The combined method of rough sets theory and genetic algorithms is applied to data mining. By making use of operating data of the essential equipment in enterprises, rough sets theory is adopted to cut down redundant attributes, and the treated data act as a sample training set. Then the classified model is built by optimized genetic algorithms. On the basis of the classified model, the intrinsic regularity how the machine runs can be found, and the unknown breakdown equipment may be classified quickly, which will provide a powerful backing for failure diagnosis and failure prediction.