CAJ | 학술논문

针对电路板故障征兆获取困难的问题，提出了基于 BP 神经网络的故障征兆获取方法；BP 神经网络获取故障征兆是指利用隶属函数将症状空间的向量模糊化，然后再通过 BP 神经网络映射到故障空间的一种方法；实验结果表明：BP 神经网络获取电路板故障征兆不仅较为准确，而且精度较高，适合复杂电路故障的诊断。
침대전로판고장정조획취곤난적문제，제출료기우 BP 신경망락적고장정조획취방법；BP 신경망락획취고장정조시지이용대속함수장증상공간적향량모호화，연후재통과 BP 신경망락영사도고장공간적일충방법；실험결과표명：BP 신경망락획취전로판고장정조불부교위준학，이차정도교고，괄합복잡전로고장적진단。
In order to overcome the difficult problem of circuit board fault diagnosis,this paper proposed a method of circuit board fault based on the BP neural network. This method used membership function to fuzz the vector of fault space and maps the vector through BP neural network. Experimental results show that the BP neural network is accurate as well as precise and it can be used to diagnose complicated circuit board.