基于1DCNN-GRU的启闭机液压系统故障诊断
Fault Diagnosis of Hoist Hydraulic System Based on 1DCNN-GRU
刘英杰 1董詠依 2刘鹏鹏 2葛孟伟2
作者信息
- 1. 华北水利水电大学 水利学院,郑州 450046;水资源高效利用与保障工程河南省协同创新中心,郑州 450046;河南省水环境模拟与治理重点实验室,郑州 450046
- 2. 华北水利水电大学 水利学院,郑州 450046
- 折叠
摘要
由于启闭机液压系统内部结构复杂,故障信号不易采集,使用AMESim软件搭建启闭机液压系统仿真模型,构建 6 种典型故障数据集.基于这些数据集,提出一维卷积神经网络(1 Dimensional Convolutional Neural Network,1DCNN)与门控循环单元(Gated Recurrent Unit,GRU)相结合的故障诊断方法,利用1DCNN提取信号数据的空间特征和GRU提取信号数据的时间特征,实现对信号数据空间及时间特征的融合,并对融合特征进行分类识别.
Abstract
As the internal structure of the hoist hydraulic system is complex and fault signals are not easy to collect,AMESim software is used to build a simulation model of the hoist hydraulic system and construct 6 typical fault datasets.Based on these dataset,a fault diagnosis method combining 1 Dimensional Convolutional Neural Network(1DCNN)and Gated Recurrent Unit(GRU)is proposed.1DCNN is used to extract the spatial features of signal data and GRU is used to extract the temporal features of signal data to realize the fusion of spatial and temporal features of signal data,and the fusion features are classified and recognized.
关键词
启闭机/液压系统/一维卷积神经网络(1DCNN)/门控循环单元(GRU)/特征融合/故障诊断Key words
hoist/hydraulic system/1 Dimensional Convolutional Neural Network(1DCNN)/Gated Recurrent Unit(GRU)/feature fusion/fault diagnosis引用本文复制引用
基金项目
国家自然科学基金(51709116)
水利部重大科技项目(SKS-2022142)
出版年
2024
NETL
NSTL
维普
万方数据