Research On Test And Diagnosis Method Of Printed Circuit Board Based On Artificial Intelligent Technology

With the development of electric technology and military projects, printed circuit boards become more and more complicated and many of them are working at very hard environment. Sometimes, printed circuit boards have to work all the time until the mission is accomplished, then some components on the printed circuit boards lapse easily. So we have to find the new way to improve the reliability of printed circuit boards and when they are broken, we could find broken components quickly and accurately.The project"research on test and diagnosis method of printed circuit board based on artificial intelligent technology"is assigned to find a method to test and diagnosis printed circuit board with comprehend of artificial intelligent technology including wavelet analysis,data mining,neural network and information fusion.The project studies on some artificial intelligent technologies, and combines them in the software. This software has six parts which are training data-input,wavelet analysis,data -train,diagnosis data-input,wavelet analysis,data-diagnose.Some special points on printed circuit board are measured, then their voltage are analyzed by wavelet analysis, so this paper could use data mining and neural network to diagnose faults on the boards. Owing to many printed circuit boards are very complex, this paper imports information fusion to improve the software's veracity.According to lots of experiments, this method's accurateness reaches to 90% and it is very useful and effective.