中文摘要：

着重介绍数据驱动故障预测和健康管理（PHM）方法的研究现状.通过对数据驱动PHM方法的分类阐述,逐步说明面向复杂系统数据驱动PHM的方法体系和流程,并重点对构成数据驱动PHM方法体系的核心环节进行分析和总结.在此基础上,采用一个锂离子电池循环寿命预测实例综合分析了数据驱动PHM的实现过程.最后,分析了数据驱动PHM方法的发展趋势和研究挑战.

英文摘要：

The data-driven prognostics and health management （PHM） approaches are focused in this review.The methodologies and categories for data-driven PHM approaches are firstly introduced.Then,the data-driven PHM framework and system architecture for complex system are discussed in detail.The health state monitoring,feature identification and extraction,data-driven prediction algorithms,prognostic uncertainty and hybrid prognostic approach in data-driven PHM framework are systematically described.Based on above,the life cycle prediction of a lithium-ion battery is taken as the example to synthetically analyze the implementation process of data-driven prognostics and health management.Finally,with summarizing the research hot issues,the challenges and the developing trend of data-driven PHM are analyzed.