相似性聚类在直升机传动轴健康管理中的研究

Research on Similarity Propagation Clustering in Helicopter Transmission Shaft Health Management

直升机传动系统作为直升机的三大核心部件,其安全余度小,直接影响着直升机的安全性能。传动轴作为传动系统中的核心部件,其安全性能关系直升机安全。利用智能算法对传动系统数据进行分析和提取,用于直升机异常状态监测与健康管理,是智能航空和物联航空发展的热点核心技术。以直升机传动轴为研究对象将提出一种加权的相似性聚类的方法,将其用于传动轴故障特征提取,实验表明其具有独特的优越性,很好地解决了传动轴故障特征提取和数据降维处理等问题。与传统的相似性聚类方法相比,其在传动轴故障特征提取方面有了很大的进步,提高了传动轴故障早期和细微故障的分辨率和准确率。用先进智能算法解决了传动系统中故障诊断依靠工人经验的弊端,提高了效率,保证了安全,为未来智能航空中智能故障诊断提供了思路。

Helicopter transmission system is one of the three core components of the helicopter,it has small safety redundancy,which directly affects the safety performance of the helicopter.As the core component of transmission system,the safety performance of transmission shaft is related to helicopter safety.The intelligent flight calculation algorithm is used to analyze and extract the aviation flight data,and is applied to helicopter abnormal condition monitoring and health management,which is the core technologies of aviation and IOT aviation development.A weighted similarity propagation clustering method is proposed to extract fault features of helicopter transmission shaft.The experiment shows that it has unique advantages and can well solve the fault characteristics of the transmission shaft extraction and data dimensionality reduction and other issues.Compared with the traditional similarity clustering method,the proposed method has make great progress in fault feature extraction of transmission shaft,which can improve the resolution and accuracy of early and minor faults of transmission shaft.Advanced intelligent algorithm solves the problem that fault diagnosis in transmission system depends on workers'experience,improves efficiency,ensures safety,and provides ideas for intelligent fault diagnosis in future intelligent aviation.