复杂装备故障预测方法研究综述  

A research review on fault prognostic techniques for complex equipments

作　　者：徐兆平 郭波[1] XU Zhaoping;GUO Bo(College of Systems Engineering,National University of Defense Technology,Changsha 410073,China)

机构地区：[1]国防科技大学系统工程学院,湖南长沙410073

出　　处：《长沙理工大学学报（自然科学版）》2023年第2期10-26,共17页Journal of Changsha University of Science and Technology:Natural Science

基　　金：国家自然科学基金资助项目(72071208)。

摘　　要：【目的】阐明复杂装备故障预测内涵,指导装备主动性维修。【方法】对复杂装备故障预测研究内容、国内外研究现状以及方法体系进行调研、归纳和分析,划分并评述现有方法的适用条件和优缺点。【结果】基于知识的故障预测方法可充分利用来自相关领域专家的经验知识,但知识的获取是瓶颈问题;基于模型的故障预测方法可深入理解对象系统本质,但实际复杂装备的精确模型很难构建;数据驱动的故障预测方法依赖于大量数据,而实际应用中一些复杂装备的典型数据的获取代价很大;混合方法能克服单个预测方法的局限性,但有效的模型设计是一个难点。【结论】混合方法能更好地提高预测系统的智能性和预测性能,是复杂装备故障预测的重要发展趋势。[Purposes]This study aims at directing the condition-based maintenance of equipment by expounding the connotation of complex equipment fault prognostic.[Methods]In this study,the relevant research contents,status,and methods were investigated,summarized,and analyzed.The existing fault prognostic methods were divided into different categories and the corresponding application conditions,advantages,and drawbacks were discussed.[Findings]The knowledge-based methods can take full advantage of the experiential knowledge from experts,but the knowledge acquisition was a bottleneck problem.The modelbased methods had the advantages of in-depth understanding of the nature of the target systems,but it was difficult to establish accurate models for complex equipments.The data-driven methods relied on a large amount of data.However,the cost of acquiring typical data of some complex equipments was very high.The hybrid methods can overcome the limitation of a single method,but designing an effective hybrid model was challenging.[Conclusions]The hybrid methods can improve the intelligence and performance of the fault prognostic system,which is an important development trend of complex equipment falt prognostic.

关 键 词：复杂装备 故障预测 主动性维修 混合方法 基于知识的方法 基于模型的方法 数据驱动的方法 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]