电梯曳引系统故障监测和诊断技术研究及实现  

Research and implementation of fault monitoring and diagnosis technology for elevator traction system

作　　者：周奇才[1] 朱梦田 康振扩 冯双昌[2] Zhou Qicai;Zhu Mengtian;Kang Zhenkuo;Feng Shuangchang

机构地区：[1]同济大学机械与能源工程学院,上海201800 [2]上海市特种设备监督检验技术研究院,上海201800

出　　处：《起重运输机械》2025年第3期94-100,共7页Hoisting and Conveying Machinery

摘　　要：曳引系统是电梯的核心部件之一,若其在运行过程中产生故障可能会导致重大生命财产损失,文中针对该问题提出了融合感知层、边缘处理层和云服务层的监测系统和基于长短期记忆网络(LSTM)的故障诊断方法,实现了电梯曳引系统的在线故障监测和诊断。通过对某服役电梯进行现场测试,结果显示该技术方法对电梯曳引系统的诊断响应时间平均约0.03 s,准确率高达95.89%,较传统的人工巡检方式具有更好的及时性和诊断准确率,为电梯这类特种设备的实时在线健康管理提供了技术基础。The traction system is a critical component of an elevator,and its failure during operation can result in significant loss of life and property.To address this,a comprehensive monitoring system that integrates a perception layer,edge processing layer,and cloud service layer,is proposed,along with a fault diagnosis method based on the Long Short-Term Memory network(LSTM).This system enables real-time fault monitoring and diagnosis of the elevator traction system.Field test results from an in-service elevator demonstrate that the average diagnostic response time for the traction system is approximately 0.03 seconds,with an accuracy rate of up to 95.89%.These figures exceed the efficiency and accuracy of traditional manual inspection methods,providing a technical basis for the real-time online health management of critical equipment like elevators.

关 键 词：电梯 曳引系统 故障监测 故障诊断 长短期记忆网络 

分 类 号：TU857[建筑科学]