摘要
介绍了在微机上开发的一种基于神经网络的测试生成系统结构,详细讨论了系统中各模块的实现方案。从提高效率的角度。
A test pattern generation system implemented with neural network method is introduced in this paper.The system consist of circuit description,preprocessing,the build of neural network,test pattern generation and test informations module.A lot of strategies are discussed to improve the performances of the method,such as the selection of constrained network that accords with the circuit under test,evaluating the values of neurons,circuit partition,the selection of optimization algorithms.Some problems in the test pattern generation method based on neural network is presented and discussed.The experimental results indicate the test system is effective.
出处
《电子科技大学学报》
EI
CAS
CSCD
北大核心
1997年第3期324-327,共4页
Journal of University of Electronic Science and Technology of China
基金
国家"八五"重点科技攻关项目
关键词
数字电路
测试图形产生
神经网络
结构分析
test pattern generation
neural network
constrained network
optimization algorithms