基于提升格式小波的磁浮间隙信号降噪处理  

Denoising of Maglev Gap Signal Based on Lifting Scheme Wavelet

作　　者：郑树彬[1] 林建辉[1] 林国斌 

机构地区：[1]西南交通大学 牵引动力国家重点实验室,四川成都610031 [2]上海磁浮交通工程技术研究中心,上海201204

出　　处：《计算机测量与控制》2008年第5期711-714,共4页Computer Measurement &Control

基　　金：国家863磁浮专项创新课题(2005AA505010-511)

摘　　要：针对高速磁浮轨道不平顺检测中间隙信号的降噪问题,提出基于提升格式的小波滤波方法;通过Lagrange插值法设计预测算子,并根据小波消失矩的特性建立线性方程组求解出更新算子,该方法也适用于求解带有边界问题的预测算子和更新算子;仿真实验比较了采用不同小波分解层数时硬阈值和软阈值的降噪效果,结果表明对信号进行4层分解并采用硬阈值法对小波系数进行处理,可以更有效提高信噪比,并更好保留了间隙信号的脉冲尖峰特征;将其仿真结论应用于实测间隙信号的降噪处理中,结果表明该方法满足了轨道不平顺实时检测的要求。For the problem of denoising high--speed maglev＇s gap signal used for inspecting track irregularity, a filtering method based on wavelet lifting scheme was presented. Predict operator was designed using Lagrange interpolation, and update operator was worked out by solving linear equations established by the vanishing moment property of wavelet. And this method is also capable of getting the predict operator and update operator for the boundary problem of wavelet transform. Simulation was conducted to check the denoising performances of hard--thresholding and soft--thresholding with various wavelet decomposition levels. And Simulation results show that SNR （signal--to--noise ratio） is more effectively improved, and the impulse spike characteristic of gap signal is also more preserved while gap signal is decomposed into 4 levels combined with hard--thresholding than soft--thresholding to modify the wavelet coefficients. According to simulation conclusions, actual gap signal was denoised, and the result indicates that this filtering method meets the requirement of track irregularity inspection in real--time.

关 键 词：轨道检测 高速磁浮 小波 提升格式降噪 

分 类 号：TN911.72[电子电信—通信与信息系统] U237[电子电信—信息与通信工程]