摘要
文章介绍了一种基于遗传算法的自适应测试生成方法。首先讨论了用遗传算法进行测试生成时构造评价函数的一些方法,然后应用组合电路的Hopfield神经网络模型,提出了基于遗传算法的自适应测试生成算法,该方法不同于传统的方法,它不需要故障传播、回退等过程。实验结果表明了本算法的可行性。
This pape r mainly proposes a self -ad aptive test generation algorithm using the genetic algorithm.Firstly,several evaluation functionsof the self -ad aptive test ge neration algorithm using the genetic algorithm are intr oduced.Secondly,the co m-binatory circuits is represented as a bidirectional network of neurons using the Hopfield nets,and a self -adaptive genetic algorithm is used for the test genera tion.This method is radically different from the conventional methods,a nd it doesn' t need the process of propagation and backtracks.Some experimental results on the combinational circuits demon strate the feasibility of this algo rithm.
出处
《微电子学与计算机》
CSCD
北大核心
2002年第3期14-16,共3页
Microelectronics & Computer
