Research Of The Fault Diagnosis Of Analog Circuits Based On LabVIEW And BP Neural Network

The applications of large-scale integrated circuits, the vulnerable of components, and the technology of analog circuits fault diagnosis's slowly development make lots of countries do research in area of the analog circuits fault diagnosis constantly. And neural network has many advantages in diagnosis, such as study by himself, the ability of information processing. LabVIEW is a graphic program software, and it advantages also make it used in fault diagnosis. This paper is based on that three theories, and designed a fault diagnosis of analog circuits system based on LabVIEW and BP neural network. What we do and get in this paper include fields shown as following:Firstly, we research the Algorithm of BP neural network and the improved Algorithms, and provide the normal methods and steps to realize standard BP Algorithm on the platform of LabVIEW. This method uses fully the advantage of LabVIEW's graphic programming, realize the BP Algorithm with the LabVIEW's controls and functions, and prove its effect with the use of BP neural network in analog circuits fault diagnosis.Secondly, we mainly discuss how to use hybrid programming to realize improved BP Algorithms, and prove its effect with its use in the area of function approach.This method uses the interfaces provided by the LabVIEW to communicate with lots of program languages and applications interfaces. The interface used in this paper is Matlab Script node, which realize the communication of LabVIEW and Matlab. With the use of Matlab Script node, we can not only benefit from the Matlab's huge Matrix computing power, but also we can use Matlab's neural network toolbox, which can save programming time and make our work efficiency.Last, in this paper we design the analog circuits fault diagnosis system based on LabVIEW and BP neural network. And use the system to diagnose the multiple soft fault and single hard fault in two circuits to test its efficiency. In this system, we use the sub VI to realize the fault's classify show. Compared with the old output result, this system analyze the result and classified the fault sorts, and most important we make the result shown directly in the computer's interface. In this way, even though you are not a expert in fault diagnosis, you can known the fault, which ease the workers'mission greatly.