摘要
牵引电机、联轴节及齿轮等传动机械广泛应用于轨道交通车辆,是高速列车动力链的重要组成部分。这些动力链部件长期工作在复杂恶劣的环境下,在宽速域、大负载工况及轮轨冲击振动等因素影响下容易发生故障,进而影响列车的安全运行与行车秩序。因此及时预警潜在故障对于确保轨道交通车辆的正常运行与行车秩序具有重要的意义。由于基于电信号的诊断技术具有信号易于获取、信号可靠性和准确性高、可实现对象部件的非嵌入式监测等优点,逐渐成为轨道交通故障诊断方向的研究热点。文章阐述了轨道交通车辆动力链关键部件的故障原理,以基于电信号的诊断方法为切入点,对该领域的现有诊断方法与研究成果进行整理与分析,然后基于多特征融合与机器学习理论,提出了一种全新的基于电信号的多变量解析诊断法。该方法首先获取各电信号数据,进行小波降噪,然后通过信号的分解与重构提高信噪比,基于重构信号提取不同的故障特征,最后利用决策树统合各故障特征进行诊断。验证试验与实际应用效果表明,本研究提出的电信号诊断法能够有效检测并识别动力链故障,可以实现早期故障预警,保障高速列车的运行安全。
Transmission machineries such as traction motors, couplings and gears are widely used in rail transit vehicles, which are regarded as important parts of the power chain of high-speed trains. Due to the long-term operation in complex and harsh environment, the influence of wide speed range, heavy load conditions and rail surface impact, the transmission components are easy to cause faults, which will affect the normal operation of vehicles and traffic order. Therefore, warning potential faults timely is of great significance to guarantee the normal operation and traffic order of rail transit vehicles. Because the diagnosis method based on electrical signal has the advantages of easy signal acquisition, high signal reliability and accuracy, non-embedded monitoring of object equipment, it has gradually become a research hotspot in the field of rail transit diagnosis. The fault mechanism of key components in rail transit vehicles power chain was first described. Taking the diagnosis method based on electrical signal as the breakthrough point, the existing diagnosis methods and research findings in this field was sorted out and analyzed. Then, based on the theory of multi-feature fusion and machine learning, a new electrical signal diagnosis method was proposed. The data of each electrical signal was obtained, the wavelet denoising was performed first, and the signal-to-noise ratio was improved through signal decomposition and reconstruction, then different fault features were extracted based on the reconstructed signals, and finally the decision tree to integrate the fault features for diagnosis was used. The verification test and actual application results showed that the electrical signal diagnosis method proposed in this research can effectively detect and identify power chain faults, realize early fault warning, which can ensure the driving safety of vehicles.
作者
冯江华
FENG Jianghua(CRRC Zhuzhou Institute Co.,Ltd,Zhuzhou,Hunan 412001,China)
出处
《机车电传动》
北大核心
2021年第1期1-9,共9页
Electric Drive for Locomotives
基金
“十三五”国家重点研发计划项目(2016YFB1200401)。
关键词
高速列车
滚动轴承
牵引电机
齿轮
信号处理
机器学习
故障诊断
轨道交通
high-speed train
rolling bearings
traction motors
gear
signal processing
machine learning
fault diagnosis
rail transit
作者简介
冯江华(1964-),男,博士,教授级高级工程师,长期从事牵引传动与控制技术研究,E-mail:fengjh@csrzic.com。
