摘要
提取天然气压缩机故障状态下振动信号的故障特征是设计机械自动化检测系统的核心技术。提出一种基于非线性关联维特征提取的机械自动化监测系统设计方法,在故障诊断原理基础上,进行故障振动信号时间序列分析,设计故障振动信号相空间重构方法,改进了相空间重构最佳时延和嵌入维数参数计算的关键技术,通过提取关联维故障特征,在Simulink平台上设计了自动化监测系统。系统实验结果表明,该算法和系统能使各类故障状态下提取的关联维特征的标准差显著性降低,特征分布聚类能力明显提高,系统能有效检测各类故障,实现了机械设备的自动化监测,在自动化故障诊断仪表设计等领域具有较好的工程实践价值。
The fault feature extraction of the vibration signal of natural gas compressor in fault state is the core technology in design of mechanical automation detection system.A mechanical automatic monitoring system is proposed based on nonlinear correlation dimension feature extraction,the principle of fault diagnosis is analyzed,and the time series analysis of fault vibration signals is obtained.Phase space reconstruction method of fault vibration signal is designed,and key technology of computing of optimal time delay and embedding dimension parameters for phase space reconstruction is improved.The correlation dimension fault feature is extracted,and the automatic monitoring system is obtained based on Simulink platform.Experimental results show that the system can make the standard deviation of correlation dimension extraction decreased significantly,and the clustering ability is enhanced,the system can effectively detect all kinds of faults,monitoring of mechanical equipment is obtained.It has better engineering practice value in the field of automatic fault diagnosis and instrument design.
出处
《计算机与数字工程》
2014年第12期2311-2315,共5页
Computer & Digital Engineering
关键词
非线性
关联维
特征提取
自动化设备
nonlinear
correlation dimension
feature extraction
automation equipment
