基于数据驱动的装备故障预测技术研究

Research on Data-driven Fault Prediction Technologies for Equipments

基于数据驱动的故障预测技术是目前应用最广泛,研究最深入的一类故障预测方法;进一步梳理总结了故障预测技术的分类,从可利用数据特点以及不同预测思路两个角度对基于数据驱动的故障预测技术进行了分类研究,着重分析了不同方法的预测思想并阐述了其优缺点及适用范围,最后讨论并指出了基于多源信息融合的系统级故障预测技术是这一领域的重要发展方向。

Data--driven fault prediction technologies are the most popular methods in the field of fault prediction. The categories of fault prediction technologies are summarized further, the data--driven fault prediction technologies are classified and studied according to the char- acteristics of data which could be used for predicting and the different thoughts of prediction. Then, the thoughts of these different methods are analyzed and their advantages and shortcomings are expounded. Finally, it＇ s pointed out that the system--level fault prediction technol- ogy based on multi--source information fusion is an important development tendency in this field.