基于BP神经网络的吸运风机故障诊断  

Fault diagnosis of the fan with air suction based on BP neural network

作　　者：郝志勇[1] 刘伟[1] 夏玮[1] 闫闯[1] 

机构地区：[1]辽宁工程技术大学机械工程学院,辽宁阜新123000

出　　处：《工程设计学报》2012年第1期57-60,66,共5页Chinese Journal of Engineering Design

基　　金：国家自然科学基金资助项目(59774033);辽宁省教育厅创新团队项目(2008T068)

摘　　要：吸运风机是农业现代化生产机械联合收割机的重要组成设备.针对目前吸运风机经常出现的故障,收集了吸运风机故障征兆和其对应的故障类型.将故障样本数据和模糊神经网络相结合,并根据BP神经网络确定网络的输入和输出向量,对风机进行故障诊断,诊断结果与实际情况比较吻合.运用MATLAB实现神经网络故障诊断仿真,仿真结果表明诊断误差较小,输出向量与实际故障矩阵结果接近.The air suction fan is an important component device of agricultural modern production equipment.Aiming at the faults occurring currently in the fan,the fan fault symptoms and corresponding fault type were collected.Combined fault sample data and fuzzy neural network,confirmed input and output vector of network according to the BP neural network and diagnosed the fault of fan.The diagnosis is consistent with the reality.Use MATLAB to simulate neural network fault diagnosis and the simulation results show that the diagnosis error is small,output vector and the actual fault matrix results are approximate.

关 键 词：风机 人工神经网络 BP算法 故障诊断 

分 类 号：TH165.3[机械工程—机械制造及自动化]