Scholarlink

According to the characteristics of escalator in public transportation, such as low speed, heavy load and sudden change, and combined with the monitoring requirements of common faults of key components such as escalator drive host, step chain, and brake, the multi-dimensional sensor monitoring scheme for escalator vibration, displacement, speed, and temperature is constructed. After obtaining the signal collected by the sensor, the instantaneous characteristics and frequency of the vibration signal are obtained by using high-precision time–frequency analysis. The non-stationary signals in escalator operation are analyzed with the method of order analysis, so as to eliminate the non-fault redundancy information caused by the change of working condition and load to a large extent. Therefore, the decoupling and separation of fault trend characteristics and variable load state characteristics are realized. Based on MRDenseNet (Multipath Residual Densely Connected Convolutional Networks), the time domain signal of escalator vibration is converted into time–frequency diagram by AWSET (Adaptive Window Synchronous Extraction Transformation), and the fault diagnosis of inner ring of main drive wheel bearing under varying working conditions and load is realized by using tachometer step analysis. It can provide reliable technical support for intelligent fault diagnosis of escalators in practical applications.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Research on Condition Monitoring and Fault Diagnosis of Escalator Under Complex Working Conditions

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Qianfei Zhou,
  • Shuqing Ding,
  • Yuegui Feng,
  • Guangwei Qing,
  • Huifang Wang

摘要

According to the characteristics of escalator in public transportation, such as low speed, heavy load and sudden change, and combined with the monitoring requirements of common faults of key components such as escalator drive host, step chain, and brake, the multi-dimensional sensor monitoring scheme for escalator vibration, displacement, speed, and temperature is constructed. After obtaining the signal collected by the sensor, the instantaneous characteristics and frequency of the vibration signal are obtained by using high-precision time–frequency analysis. The non-stationary signals in escalator operation are analyzed with the method of order analysis, so as to eliminate the non-fault redundancy information caused by the change of working condition and load to a large extent. Therefore, the decoupling and separation of fault trend characteristics and variable load state characteristics are realized. Based on MRDenseNet (Multipath Residual Densely Connected Convolutional Networks), the time domain signal of escalator vibration is converted into time–frequency diagram by AWSET (Adaptive Window Synchronous Extraction Transformation), and the fault diagnosis of inner ring of main drive wheel bearing under varying working conditions and load is realized by using tachometer step analysis. It can provide reliable technical support for intelligent fault diagnosis of escalators in practical applications.