摘要
针对2D位移激光传感器在高速和宽波段的钢轨波磨动态测量过程中受到点头振动和噪声干扰的问题,提出一种基于经验模态分解(EMD)的分趋势式数据拼接方法。首先,将采集到的波磨数据进行EMD,并将其分为长趋势与短趋势;然后,通过盲识别方法提取出数据采集中产生的偏移角度,从而准确提取出相邻采样数据之间的相干数据集;最后,通过不同趋势段的数据配准完成相邻采样数据集的拼接。实验结果表明:所提方法可稳定、精确地在真实铁路测量环境中对钢轨波磨进行测量。
Aiming at the problems of 2D displacement laser sensor disturbed by nodding vibration and noise interference in process of high-speed and broadband rail corrugation dynamic measurement,a sub-trend data splicing method based on empirical mode decomposition(EMD)is proposed.Firstly,EMD is used to decompose the collected data,which is divided into long trend and short trend.Then,the blind recognition method is used to extract the offset angle generated in the data acquisition,so as to accurately extract the coherent datasets between the adjacent sampling data.Finally,data registration of different trend segments is used to complete the splicing of adjacent sampling datasets.The experimental results show that the proposed method can measure rail corrugation stably and accurately in real railway measurement environment.
作者
滕云
刘宏立
马子骥
刘建伟
TENG Yun;LIU Hongli;MA Ziji;LIU Jianwei(College of Electrical and Information Engineering,Hunan University,Changsha 410082,China)
出处
《传感器与微系统》
CSCD
北大核心
2023年第5期121-125,共5页
Transducer and Microsystem Technologies
基金
国家自然科学基金资助项目(61771191,61971182)
湖南省自然科学基金资助项目(2020JJ4213)
长沙市科技计划资助项目(KQ1801194)。
关键词
钢轨波磨
数据拼接
点头振动
动态测量
2D激光位移传感器
rail corrugation
data splicing
pitching vibration
dynamic measurement
2D laser displacement sensor