基于自适应神经网络的高速数字电路测试的研究  

Research on Field Theory Based Adaptive Resonance Theory High Speed Digital Circuit Test

作　　者：周跃夫[1] 姜海勋[1] 袁钢 王彤威[3] 

机构地区：[1]海军驻北京地区航空军事代表室,北京100041 [2]海军航空军械修理所,上海200436 [3]北京航天测控技术开发公司,北京100041

出　　处：《计算机测量与控制》2009年第6期1092-1094,共3页Computer Measurement &Control

摘　　要：介绍了神经网络的模型建立。针对高速数字电路故障诊断的特点,提出了基于自适应神经网络诊断方法;与传统的神经网络相比,基于自适应神经网络具有学习速度快、网络能够收敛到最优权值、收敛精度很高、网络的分类能力强等特点,从而很好地提高故障诊断的效率;通过一个测试用例验证自适应神经网络在高速数字电路测试诊断中的应用,可以大大提高故障诊断的效率。The modeling of neural network is introduced in this paper. In order to improve the efficiency of fault diagnosis , The method of Field Theory based Adaptive Resonance Theory is presented with the characteristic of High Speed Digital Circuit fault diagnosis. Comparing with the traditional neural network, Field Theory based Adaptive Resonance Theory have the characteristic of the fast speed and the network convergence etc. The test moulded board is verified with the Field Theory based Adaptive Resonance Theory on the application of the high speed digital circuit test.

关 键 词：神经网络 自适应神经网络 故障诊断 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]