摘要
基于组合电路测试生成的离散Hopfield神经网络模型,将混沌搜索与Hopfield网络的梯度算法相结合,利用混沌搜索的内随机性及遍历性来克服梯度算法易于陷于局部极小的缺点,形成一种具有全局搜索能力的测试生成有效算法。该算法综合了随机性和确定性算法的优点,其性能优于一般的随机性算法。实验结果验证了该测试生成算法的有效性。
Based on a Hopfield neural network model for combinational circuit test generation, a test generation algorithm with global searching ability is proposed. This algorithm is the combination of chaotic searching with gradient algorithm. By means of the inherent stochastic and ergodic property of chaotic searching, the problem of being trapped in local minimum of gradient algorithm can be solved. The proposed algorithm has the advantages of both stochastic and deterministic algorithms. the performance of proposed algorithm is superior to ordinary stochastic algorithm. The experimental results confirm the effectiveness of the algorithm.
出处
《电路与系统学报》
CSCD
2002年第4期108-111,共4页
Journal of Circuits and Systems
基金
国防科技预研基金资助项目
