基于能量算子的声源定位技术在TADS探测车轮故障中的应用  

Application of Sound Source Localization Technology Based on TED in TADS Wheel Fault Detection

作　　者：贾超 乔森 苗鑫 杨征西 李彤 刘鹏飞[3] JIA Chao;QIAO Sen;MIAO Xin;YANG Zhengxi;LI Tong;LIU Pengfei(Beijing Aeroctro Information Technology Co.,Ltd.,Beijing 100095,China;China Railway Xi’an Group Co.,Ltd.,Xi’an 710054,China;State Key Laboratory of Mechanical Behavior and System Safety of Traffic Engineering Structures,Shijiazhuang Tiedao University,Shijiazhuang 050043,China)

机构地区：[1]北京航天康拓信息技术有限公司,北京100095 [2]中国铁路西安局集团有限公司车辆部,陕西西安710054 [3]石家庄铁道大学省部共建交通工程结构力学行为与系统安全国家重点试验室,河北石家庄050043

出　　处：《铁道车辆》2024年第5期58-65,共8页Rolling Stock

基　　金：中国国家铁路集团有限公司科技研究开发计划(N2022J027-B)。

摘　　要：车轮故障是影响铁道车辆运行安全的重要因素。使用TADS技术可以实现对车轮故障等走行部部件的故障检测。由于同转向架车轮噪声信号的叠加干扰会造成车轮故障等噪声信号区分困难,容易造成车轮故障的误报和误判,因此对故障噪声来源进行定位和区分十分重要。在TADS探测车轮故障进行数据分析时,借鉴了医学领域广泛应用的基于近距离麦克风的双耳结构多运动声源定位的经验。利用TADS的短间距声学阵列,结合TADS探测中声源移动速度快、难捕捉的特点,综合使用谱峭度、Gammatone滤波器、能量算子等技术和手段建立了适应铁道车辆动态检测运用场景的运动声源定位模型,能够实现对快速运动声源的追踪和定位。通过对实际案例进行分析验证了运动声源定位模型方法的有效性,可以给出快速运动声源的运动轨迹。通过运动轨迹与理论轨迹相匹配进行声源性质的判定,可用于辅助诊断车轮故障。Failure of wheels is a significant obstacle to the smooth operation of railway vehicles.TADS technology can be used to detect problems with running components,like wheel failure.Due to the imposed interference with the noise signal of the wheel unit,it becomes challenging to distinguish wheel fault signals,resulting in false alarms and misjudgments of wheel faults.Therefore,it is imperative to locate and distinguish the source of fault noise.TADS perform data analysis by referring to the binaural multi-motion sound source localization technique based on close-range microphone,which is widely used in medical field.Due to the characteristics of fast moving speed and challenging capture of sound sources in TADS detection,a moving sound source localization model is established for railway vehicle dynamic detection using the short-pitch acoustic array of TADS,spectral kurtosis,Gammatone filter,energy operator and other technologies and means.This approach permits the tracking and localization of fast-moving audio signals.The efficacy of the moving sound source localization model has been validated through the analysis of actual cases,and the movement trajectory of a fast-moving sound source is readily available.The sound source’s characteristics can be found by comparing its movement with its theoretical trajectory.This can help diagnose problems with wheels.

关 键 词：能量算子 谱峭度 Gammatone滤波器 TADS 运动声源定位 铁道车辆 车轮故障 

分 类 号：U270.331[机械工程—车辆工程]