基于小波包能量谱的钢轨扣件松脱检测研究  

Study on rail fastener loosening identification based on wavelet packet energy spectrum

作　　者：周文静[1] 刘冲[2] 韦佳宏[1] 张凯[1] 张学海[1] 

机构地区：[1]大连理工大学机械工程学院,辽宁大连116024 [2]大连理工大学辽宁省微纳米技术及系统重点实验室,辽宁大连116024

出　　处：《传感器与微系统》2017年第5期40-43,共4页Transducer and Microsystem Technologies

基　　金：国家"十二五"科技支撑计划资助项目(2011BAG05B02-03;2011BAG05B02-02)

摘　　要：提出了基于小波包能量谱的钢轨扣件松脱识别方法,对获得的振动信号进行小波包分解,结合扣件松脱前后振动信号频域内能量变化,构造损伤指标(DI)。在此基础上,开展了5种工况下的轨道结构动力响应测试,进行了轨道结构状态识别分析。实验结果表明:DI能够有效地识别轨道结构状态,且随着松脱程度的增加而增加;同一工况下得到的损伤指标DI变化波动均在±0.5以内,验证了该方法具有很好的鲁棒性和抗干扰能力。A method based on wavelet packet energy spectrum is put forward to detect rail fastener loosening.Vibration signal is decomposed with wavelet packets,combined with energy variation in frequency domain before and after fastener loosening to build damage index （DI）.On the above basis,conduct dynamic response tests of track structure under five conditions and identify and analyze on track structure state.Experimental results show that DI can effectively identify track structure state,which is significantly increasing with the increased degree of loosening;fluctuation range of DI under the same condition is within ±0.5,which proves that the proposed method has strong robust and capacity of resisting disturbance.

关 键 词：钢轨扣件 振动信号 小波包能量谱 结构状态识别 

分 类 号：U213.2[交通运输工程—道路与铁道工程]