基于PMC模型的神经网络

在计算机系统中有许多诊断算法,其中针对PMC模型有0-1规划的诊断算法,这个诊断算法是NP完全的,是在多项式时间内不可解的。然而利用神经网络的高度并行性这个特点,就可以在很短的时间内解出结果。主要分析了系统诊断模型PMC的症候特征,得到其充要条件,提出了新的算法,该算法具有高度的并行性,神经网络的发展为此提供了用模拟电路来实现算法的可能性,并进行了模拟验证。

The Neural Network of the PMC Model

In the computer system, there are many diagnosis algorithms. The diagnosis algorithm of 0-1 program is proposed for the PMC model. The diagnosis algorithm is NP complete and could not been solved in the multinomial time. However, by use of the characteristics of the high parallelism neural network, it may be solved in the very short time. This article mainly analyzes the characteristic symptoms the system diagnosis model PMC, obtains its satisfied condition and proposes the new algorithm. This algorithm has a high parallelism and the development of neural network has provided the algorithm possibility with the analogous circuit. Smulation confirmation has been carried out.