基于三值神经网络的组合电路测试生成算法

针对集成电路的规模和复杂度不断增加而相应的测试却越来越困难问题，提出了一种基于三值神经网络的组合电路测试生成算法．该算法不需要传播，也不需要回退，而是利用三值神经网络把组合电路表示成双向的神经网络，并构造网络的能量函数，用遗传算法求解能量函数的最小值点来求得测试矢量，这样就把组合电路的测试生成问题转化为数学问题．在一些基准电路上的实验结果表明，本算法具有较高的故障覆盖率和较短的测试时间．

Test Generation Algorithm Based Three-valued Neural Networks for Combinational Circuits

With the growth in size and complexity of integration circuits, test generation for them is becoming increasingly difficult. A combinational circuits test generation algorithm based three-valued neural networks is proposed in this paper. This algorithm does not need propagation and backtracks, but represents the combinational circuits as a bidirectional network of neurons using the three-valued neural networks, and constructs the energy function for the network. A genetic algorithm was used to find the global minimal as the test vectors. So we translate the problem of the combinational test generation to the math problem. The experimental results on some standard circuits demonstrate that the algorithm has high fault coverage and short test time.