基于BP神经网络的液压缸内泄漏诊断  

Internal Leakage Detection of Hydraulic Cylinder Based on BP Neural Network

作　　者：李琳[1,2] 陶建峰[1] 黄亦翔[1,2] 刘成良[1,2] 

机构地区：[1]上海交通大学机械与动力工程学院,上海200240 [2]机械系统与振动国家重点实验室,上海200240

出　　处：《液压与气动》2017年第7期11-15,共5页Chinese Hydraulics & Pneumatics

基　　金：国家自然科学基金(51305258);国家重点基础研究发展规划项目计划(973计划)(2013CB035403);上海市科委项目(1411104600)

摘　　要：液压缸是工程机械中常用的执行元件,内泄漏是其常见的故障模式,将严重影响机械系统的工作效率和安全性,及时识别液压缸内泄漏能够保证液压缸的安全正常工作。通过小波分解提取液压缸进口压力信号特征,利用BP神经网络建立分类器,实现了对液压缸内泄漏的智能识别,分类准确率高,提高了液压缸内泄漏故障诊断的效率,为实现液压缸智能状态监控提供了基础。Hydraulic cylinders are common execute components of the engineering machinery, and the internal leakage is a common fault mode of hydraulic cylinder that can affect the efficiency and safety of the machinery, so the intime detection of internal leakage can assure the normal work of the hydraulic cylinder. Based on the wavelet decomposition, the signal features of the inlet pressure of hydraulic cylinder could be extracted and the internal leakage could be intelligently detected through a BP neural network classifier. The test results show that the accuracy of the classification is acceptable. Consequently, the efficiency of the fault diagnosis of the internal leakage of hydraulic cylinder is improved and this study provides a basement of the intelligent condition based monitoring for hydraulic cylinder.

关 键 词：液压缸内泄漏 故障诊断 BP神经网络 状态监控 

分 类 号：TH137[机械工程—机械制造及自动化] TP182[自动化与计算机技术—控制理论与控制工程]