摘要
研究了一种基于BP神经网络的水果气调包装机械故障诊断方法。以水果包装容器中5种气体作为神经网络的输入,建立了可对水果气调包装机械一氧化碳充入设备故障、二氧化碳充入设备故障、氧气充入设备故障、氮气设备充入设备故障、抽真空设备故障5种单故障及混合故障进行诊断的BP神经网络模型,并通过MATLAB实现了仿真验证。通过对监测数据进行测试,测试结果表明BP神经网络用于包装机械故障诊断所建立模型的各项性能指标均处于较优水平。
A fault diagnosis method for fruit MAP machinery based on BP neural network was researched. With 5 kinds of gas in the fruit packaging container as the input of neural network,the BP neural network model that can diagnose 5 kinds of single faults such as the faults of carbon monoxide filling equipment,carbon dioxide filling equipment,nitrogen filling equipment,oxygen filling equipment and vacuum pumping equipment and their combination of the fruit MAP machinery was established,and verified through MATLAB simulation. By testing the monitoring data,the test results show that the performance indexes of the fault diagnosis model established based on BP neural network and used for fault diagnosis of packaging machinery were all at better level.
作者
文周
林伟健
WEN Zhou;LIN Wei-Jian(Dongguan Polytechnic,Dongguan 523808,China;School of Mechanical,Xi＇an University of Science and Technology,Xi＇an 710054,China)
出处
《包装与食品机械》
CAS
北大核心
2018年第5期69-72,共4页
Packaging and Food Machinery
基金
东莞职业技术学院2017年政校行企合作开展科研与服务项目(政201725)
关键词
BP神经网络
包装机械
故障诊断
仿真
BP neural network
packaging machinery
fault diagnosis
simulation
