基于最优神经网络的电路测试方法研究

The Circuits Test Generation Method Based on Optimal Neural Network Model

在数字电路最优神经网络模型的基础上,研究基于该模型的电路测试生成方法.首先获得了多输入基本门电路的最优神经网络能量函数的一般表达式,然后对这种测试方法的原理、实现步骤、以及加速测试的措施等进行了详细研究.结果表明以最优神经网络模型为基础的电路测试方法在测试生成的速度方面快于其它类似方法,如基于Hopfield神经网络的电路测试生成,具有较好的应用潜力.

An optimal neural network model for digital circuits is studied.The circuit is characterized by the energy function of optimal network A new method based on the optimal network model for digital circuit test generation is presented The test generation method consists of the following two steps: building a network model for circuit under test, and finding the minimum of energy function which accords with the circuit It is illustrated that the model presented enhances the computing efficiency of the test method based on Hopfield neural network