摘要
在浆体管道输送系统中,受设备成本、安装难度等因素的限制,往往安装的测量仪表较少,从而导致可测变量偏少,且变量之间存在非线性关系,难以建立精确的数学模型,给故障诊断带来很大困难。通过MATLAB神经网络工具箱构建BP神经网络,利用离线数据对网络进行训练,将训练好的BP网络应用到浆体管道输送系统故障诊断中,有效地实现了两种典型故障的诊断。
Due to the limit of the equipment cost and the install difficulty,the number of measuring instruments installed in the slurry pipeline transportation systems is very small,so the number of measurable variables obtained is very small.In addition,due to the fact that there exists nonlinear relationship among the measurable variables,it is hard to establish an accurate mathematical model,which causes great difficulty in the fault diagnosis.The BP neural network was established by using MATLAB neural network toolbox,and the BP network was trained by using off-line data.The application of the BP network to the fault diagnosis of slurry pipeline transportation systems shows that this method is effective to diagnose the two classical faults.
出处
《机床与液压》
北大核心
2012年第7期194-196,203,共4页
Machine Tool & Hydraulics
关键词
神经网络
浆体管道输送
故障诊断
BP算法
Neural network
Slurry pipeline transportation
Fault diagnosis
BP algorithm
