摘要
液压系统作为控制和动力传输设备的核心部件,在现代工业生产机械中被广泛应用,准确诊断其故障具有提高生产效率和保障工作安全性等重要的工程意义。液压系统一旦发生故障往往是多故障同时出现,传统BP神经网络故障诊断算法往往不能满足多故障诊断准确率,提出一种基于遗传算法改进BP神经网络(GA-BP)的液压系统故障诊断方法,针对不同采样频率下多传感器信息融合的液压系统3种典型的故障模式进行对比分析。结果表明:GA-BP故障诊断算法相对于传统的BP神经网络具有更好的诊断性能。
Hydraulic system is the core component of control and power transmission equipment.It is widely used in modern industrial production machinery.Improving the accuracy of hydraulic system fault diagnosis has important engineering significance such as improving engineering efficiency and ensuring work safety.Once faults occur,multiple faults occur at the same time,the traditional BP neural network fault diagnosis system cannot meet the diagnostic accuracy.A method of hydraulic system diagnosis was proposed based on genetic algorithm optimizing the BP neural network(GA-BP).Aiming at three typical fault modes of multi-sensor information fusion hydraulic system under different sampling frequencies,the comparative analysis was made.The results show that the GA-BP fault diagnosis algorithm has better diagnostic performance than the traditional BP neural network.
作者
李文华
牛国波
刘羽佳
LI Wenhua;NIU Guobo;LIU Yujia(National Center for International Research of Subsea Engineering Technology and Equipment,College of Marine Engineering,Dalian Maritime University,Dalian Liaoning 116026,China)
出处
《机床与液压》
北大核心
2023年第8期159-164,共6页
Machine Tool & Hydraulics
基金
国家自然科学基金面上项目(51779026)
工信部高技术船舶科研计划(工信部装函﹝2018﹞473号)
国家重点研发计划资助项目(2018YFC0309003)
辽宁省科学技术计划(2020-HYLH-35)
辽宁省“兴辽英才计划”资助项目(XLYC2007092)
高等学校学科创新引智计划(111计划)(B18009)。
