复杂系统维护策略最新研究进展:从视情维护到预测性维护  

Latest Progress on Maintenance Strategy of Complex System:From Condition-based Maintenance to Predictive Maintenance

作　　者：陆宁云[1] 陈闯 姜斌[1] 邢尹 LU Ning-Yun;CHEN Chuang;JIANG Bin;XING Yin(College of Automation Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 211106;School of Earth Sciences and Engineering,Hohai University,Nanjing 211100)

机构地区：[1]南京航空航天大学自动化学院,南京211106 [2]河海大学地球科学与工程学院,南京211100

出　　处：《自动化学报》2021年第1期1-17,共17页Acta Automatica Sinica

基　　金：国家自然科学基金(61873122);中央高校基本科研业务费(NC2020002)资助。

摘　　要：对于复杂、可修复的工程系统,设备维护是确保系统安全性、可靠性、可用性的重要手段之一.系统维护策略已经历修复性维护、定时维护、视情维护等多种维护策略.其中,视情维护是目前最受关注的维护策略,它通过收集和评估系统的实时状态信息进行维护决策,具有全寿命周期内系统可靠性高、运营维护成本低等优点.近年来,随着物联网技术、信息技术和人工智能的快速发展,一种更新颖的视情维护策略——预测性维护逐渐成为领域研究热点.本文首先简要回顾了系统维护策略的发展历程;然后,重点介绍了视情维护的研究进展,根据决策支持技术的不同,将视情维护划分为基于随机退化模型的视情维护和基于数据驱动的预测性维护,对每类技术的发展分支与研究现状进行了疏理、分析和总结;最后,探讨了当前复杂系统维护策略面临的挑战性问题和可能的未来研究方向.Device maintenance is one of important and effective means to ensure the safety,reliability and availability of complex but repairable engineering systems.Maintenance strategy has experienced various phases including corrective maintenance(CM),time-based maintenance(TBM)and condition-based maintenance(CBM).Conditionbased maintenance is the most attractive one in recent years.It can make in-time maintenance decisions by collecting and evaluating real-time condition monitoring information,hence it can be expected to achieve life-cycle high-reliability and low maintenance cost.Enabled by the internet of thing(IoT),advanced information and artificial intelligent(AI)technologies,a novel CBM strategy,predictive maintenance(PdM),is emerging and gaining increasing attentions.This paper firstly reviews the main development history of maintenance strategy and then focuses on the latest progress of CBM.According to the differences of decision-support approaches,CBM strategies are divided into stochastic deterioration model based CBM and data-driven PdM.It should be noted that,PdM can be regarded as an extension of CBM.After that,development branches and research status of each approach are sorted out and summarized.Finally,the challenging problems and possible future research directions are discussed.

关 键 词：系统维护策略 视情维护 预测性维护 随机退化模型 数据驱动 

分 类 号：N941.4[自然科学总论—系统科学]