CAJ | 학술논문

结合小波分析及有序中值方法（ROM），提出铁路轨道几何数据冲击噪声自动滤波方法。利用传统中值滤波技术滤除冲击噪声会减少信号的有用信息及质量，提出利用小波分析将信号分解成含冲击噪声的高、低频信号；利用ROM自动识别高频信号中冲击噪声位置并滤波。据铁路轨道几何数据特点与阈值指标的内在关系，将ROM两阈值指标合并成一个，并据专家知识确定阈值。新方法具有简单、高效及强健优势，且可在线实现。用新方法对含冲击噪声的模拟信号、实测轨道不平顺信号进行滤波，结果表明该方法不仅能准确识别冲击噪声位置并自动修复，且能完整保留道岔处大轨向、大值超限等有用信息。
결합소파분석급유서중치방법（ROM），제출철로궤도궤하수거충격조성자동려파방법。이용전통중치려파기술려제충격조성회감소신호적유용신식급질량，제출이용소파분석장신호분해성함충격조성적고、저빈신호；이용ROM자동식별고빈신호중충격조성위치병려파。거철로궤도궤하수거특점여역치지표적내재관계，장ROM량역치지표합병성일개，병거전가지식학정역치。신방법구유간단、고효급강건우세，차가재선실현。용신방법대함충격조성적모의신호、실측궤도불평순신호진행려파，결과표명해방법불부능준학식별충격조성위치병자동수복，차능완정보류도차처대궤향、대치초한등유용신식。
An algorithm by combining wavelet decomposition and ranked order mean (ROM)method was presented for detection and removal of impulse spike noise in track geometric data of railway infrastructure.Random events including anomalous reflection of sunlight,sensors and data transmission errors always cause spike noise in track data measured. Being different from the usual random noise,impulsive noise is a local noise.So the direct use of the traditional mean filtering technology to filter out spike noise may reduce the useful features and quality of a signal.Here,a new algorithm was developed.With it,wavelet analysis was applied to decompose a signal into higher frequency components including spike noise and lower frequency ones,and the ROM was proposed to automatically identify the location of impulsive noise in the higher frequency components and to filter it out.Two threshold indicators of ROM were combined to one on basis of their inherent relation,and its threshold was determined based on expert knowledge.It was shown that the new method has advantages of relatively simple,efficient and robust,and can be realized online;the method can be used to remove spike noise in analog signals and the measured track geometric irregularity ones;it is able not only to accurately identify the location of spike noise and repair signals,but also to retain useful features of large exceeding value and large alignment value within railway switches.