LabVIEW及BP神经网络在模拟电路故障诊断中的应用

Application of LabVIEW and BP neural netword in anolog circuits fault diagnose

LabVIEW是一种图形化编程语言,它不光设计界面直观,而且提供了许多与多种编程语言的通信接口,为虚拟仪器设计者提供了一个便捷、轻松的设计环境,被工业界及研究实验室所广泛接受。Matlab有非常强大的矩阵计算能力且有非常成熟的工具箱,适合设计及训练神经网络。若将两者结合起来,即用Matlab实现神经网络的设计训练,而用LabVIEW实现故障的诊断及显示,充分发挥了两者的优势。实验结果表明,两者结合运用于故障诊断系统中,是非常有效的。

LabVIEW is a graphical programming language, it is not only has an an intuitive design interface, but alse provides a lot of communication interfaces with a variety of programming languages. It provides a convenient and relaxed design environment for virtual instrument designer, so it is widely accepted by the industrial sector and research laboratory. Matlab has a very powerful matrix computing power and has very sophisticated toolbox, and is suitable for designing and training neural netword. If we combine this two softwares,that is,using Matlab to realize neural network＇s designing and training, and LabVIEW to realize fault diagnosis and display. So we can play the advantages of both fully. The experiment results show that a combination of both applied to fault diagnosis system is very effective.