悬臂式掘进机截割机构状态诊断策略研究  

State Diagnosis Strategy for Boring Machine Cutting Institutions

作　　者：王福忠[1] 田晓盈 张丽[2] 

机构地区：[1]河南理工大学电气工程与自动化学院,河南焦作454003 [2]河南理工大学计算机科学与技术学院,河南焦作454003

出　　处：《计算机仿真》2015年第5期390-394,共5页Computer Simulation

基　　金：国家自然科学基金(61340014)

摘　　要：悬臂式掘进机工作中常会遇到煤层硬度突然变大,使截割机构发生异常,甚至导致截齿断裂、减速器齿轮受损、截割电机转子不平衡等故障。为了实现对掘进机截割机构运行状态的远程监控,在分析掘进机截割机构工作原理及重要部件频率特征的基础上,采用Hilbert-Huang分析方法,在MATLAB中仿真得到了四种不同状态下振动信号的频谱特征和能量特征,基于支持向量机算法设计了系统的多故障分类器,通过仿真测试验证了上述分类器能有效地诊断出系统的故障部件,为实现掘进机截割机构运行状态的远程监控奠定了基础。When the cantilever machine is working, it often encountered the situation that the coal seam hardness changes seriously, an abnormal phenomenon occurs in cutting mechanism, even the cutter tooth fractures, gear tooth breaks and motor rotor is imbalance. In order to achieve the remote running state monitoring of the cutting institu- tions, we analyzed the boring machine cutting mechanism principle and frequency characteristics of the important components, and then used Hilbert-Huang technology to resolve the vibration signal to get the signal spectrum signature and energy characteristics in four different states with MATLAB. After that, we designed a multi-fanlt classifier for boring machine cutting mechanism using support vector machine （SVM） algorithm. The simulation test proved that the method can effectively diagnose the fault components of the system. All these laid a foundation for realizing running status remote monitoring systerm of machine cutting mechanism.

关 键 词：掘进机截割机构 经验模态分解 能量特征 支持向量机 多故障诊断 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]