机械设备性能退化评估与预测研究综述  

REVIEW ON MACHINERY PERFORMANCE DEGRADATION ASSESSMENT AND PROGNOSTICS

作　　者：王恒[1] 马海波[1] 徐海黎[1] 朱龙彪[1] 贾民平[2] 

机构地区：[1]南通大学机械工程学院,江苏南通226019 [2]东南大学机械工程学院,江苏南京211189

出　　处：《机械强度》2013年第6期716-723,共8页Journal of Mechanical Strength

基　　金：国家自然科学基金(50775035);江苏省自然科学基金(BK2011391);南通市应用研究计划(BK2011036)资助~~

摘　　要：机械设备性能退化评估技术在提高设备的可靠性和利用率、缩短停机维修时间、实现设备主动维护和智能维护方面有着重要的意义,是故障诊断技术的重要组成部分。综述国内外机械设备性能退化评估与预测的理论和方法,将这些方法分为两类,基于物理模型的方法和基于数据驱动的方法,重点论述基于数据驱动的方法及取得的进展,主要包括基于统计理论方法和人工智能方法等,总结比较每种方法的优缺点,并对机械设备性能退化评估与预测技术的发展方向进行展望。Machinery performance degradation assessment technology plays a significant role in proactive maintenance and intelligent maintenance, it is able to improve reliability and utilization, reduce downtime and maintenance time, and it is also an important component of fault diagnosis technology. It makes attempt to summary and review the methods of machinery performance degradation assessment and prognostics, and set these methods into two categories： physics-based approach and data-driven approach. It takes an emphasis on data-driven approach which contents statistic-based methods and artificial intelligence methods and so on. The advantages and disadvantages of each method are compared. Finally, it points out a helpful direction on machinery performance degradation assessment and prognostics for future research.

关 键 词：退化评估与预测 数据驱动 信息融合 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置] TH165.3[自动化与计算机技术—控制科学与工程]