摘要
提出了一种基于模拟退火算法的广义交叉验证(Generalized cross validation,GCV)小波阈值理论,并应用到轨道不平顺参数的处理,成功地获取了轨道静态功率谱密度。利用小波变换将轨道不平顺信号和所得谱图信号分解到互不重叠的频带上,结合GCV阈值准则和模拟退火算法阈值寻优,对小波高频系数进行阈值量化处理,滤掉高频段内的干扰与噪声信号,然后对信号进行重构,从而实现了不平顺信号的去噪和对应谱图的平滑处理。结果表明:这种处理方法能有效提高轨道静态参数信噪比,降低均方根误差,同时能够有效提高轨道静态功率谱密度图的可视性,从而为轨道维护和列车运行品质研究提供准确有效的数据。
A generalized cross validation wavelet threshold theory,which is based on simulated annealing algorithm,was proposed.This theory was applied to process the parameters of the track irregularity and its static power spectrum density was derived successfully.First,the track irregularity signal and its power spectrum density signal were decomposed into non-overlapping frequency bands by using wavelet transformation.Then,the wavelet thresholds were optimized by combining GCV threshold with simulated annealing,the high-frequency coefficients of the signal were processed by using wavelet threshold and the interference and noise of the high frequency part of the signal were filtered off.Finally,the signal was reconstructed and thus realizing de-noise of track irregularity data and smoothing of its power spectrum density.The results show that this wavelet threshold data processing method based on the GCV simulated annealing algorithm can increase the signal-noise ratio（SNR）,reduce the root mean square error（RMSE）of the track static parameters,and improve the visibility of the track static power spectrum density map,and hence provide accurate and effective data for track maintaining and research of train operating quality.
出处
《中南林业科技大学学报》
CAS
CSCD
北大核心
2010年第12期177-183,共7页
Journal of Central South University of Forestry & Technology
基金
中南林业科技大学青年基金资助项目(项目编号:200853B)
关键词
铁道工程
轨道分析
模拟退火
GCV
小波阈值
功率谱密度
去噪
信噪比
railway engineering
railroad track analysis
simulated annealing
generalized cross validation
wavelet threshold
power spectrum density
de-noise
signal-noise ratio
