现代电子技术2025，Vol.48Issue(24)：54-60,7.DOI:10.16652/j.issn.1004-373x.2025.24.009
基于1D-CNN的消防水力系统状态监测与故障诊断
1D-CNN-based condition monitoring and fault diagnosis of fire hydraulic system
摘要
Abstract
In order to carry out real-time monitoring and fault diagnosis of fire hydraulic system,a fire hydraulic system fault diagnosis algorithm and monitoring system based on one-dimensional convolutional neural network(1D-CNN)is proposed.In this system,RS 485 bus is used to collect parameters such as fire pump speed,power take-off input/output shaft temperature,inlet and outlet pressure,and pipeline flow by means of ModBusRtu protocol.The host computer built based on Winform is used to display and analyze the collected parameters in real time.Then,the 1D-CNN and LSTM algorithms are used to carry out fault diagnosis experiments.The 1D-CNN model is connected to the monitoring system for application to realize the function of fault diagnosis.The experimental results show that the fault diagnosis accuracy of 1D-CNN is as high as 98.66%,and the model performance is better than LSTM.The monitoring system runs stably and can display various monitoring data in real time.In comparison with the traditional method that requires manual feature extraction,after the fault diagnosis model is connected to the monitoring system,the information collected by multi-sensors can be directly used as input to integrate the feature extraction and classification process,so as to realize end-to-end fault diagnosis of the fire hydraulic system.The monitoring system has a friendly human-computer interface,accurate parameter monitoring,and timely fault feedback,which is suitable for daily use of fire brigades.关键词
一维卷积神经网络/消防水力系统/状态监测/故障诊断/多传感器/WinformKey words
one-dimensional convolutional neural network/fire hydraulic system/condition monitoring/fault diagnosis/multi-sensor/Winform分类
信息技术与安全科学引用本文复制引用
付文泓,胡志华,陶莉莉..基于1D-CNN的消防水力系统状态监测与故障诊断[J].现代电子技术,2025,48(24):54-60,7.基金项目
国家自然科学基金资助项目(62203291) （62203291）
