基于深度信念网络的自动扶梯制动力智能诊断系统  

The Intelligent Diagnosis System of Braking Force of Escalators Based on Deep Belief Network

作　　者：穆彤[1] 刘凯[1] 周韬[1] MU Tong;LIU Kai;ZHOU Tao

机构地区：[1]天津市特种设备监督检验技术研究院,天津300192

出　　处：《中国电梯》2021年第5期6-10,共5页China Elevator

摘　　要：依据深度信念网络理论,提出建立自动扶梯制动力预警模型和故障诊断模型并构建自动扶梯制动力智能诊断平台。该平台通过传感器数据采集,多维时间序列的数据约减,深度学习故障诊断等过程,对制动力的异常变化情况起到预警作用,为排查自动扶梯制动力存在的隐患提供了在线智能诊断技术。Based on the deep belief network theory,it is proposed to establish an escalator braking force early warning model and a fault diagnosis model and build an escalator braking force intelligent diagnosis platform.It shows that the platform can forecast the abnormal changes of the braking force through the processes of data collected by sensor,data reduced by multivariate time series,fault diagnosis by deep learning.This technology of online intelligent diagnosis can be used to support troubleshooting hidden dangers of the braking force of escalators.

关 键 词：自动扶梯 制动力 深度信念网络 多维时间序列 智能诊断 

分 类 号：TH236[机械工程—机械制造及自动化]