容错神经网络仿真器的研制

Study on the simulator of Fault tolerant Neural Networks

这篇文章中，考虑到现阶段使用神经网络时的困难，提出用容错神经网络仿真器来开发神经网络。文中首先给出了整体设计方案，然后详细地规划了各个模块及其相互调用关系，给出了基于ＶｉｓｕａｌＣ＋＋４．２的软件实现方法。文末还提供了程序运行过程中的几个界面，指出了其广阔的应用前景。

In this paper, considering the inconvenience of using neural networks,an simulator used to fault tolerant neural networks is introduced. First,the total design frame is given,then the modules and the relationships with each other are described in detail.According to this idea,the realization based on Visual C ++ 4.2 is discussed.At the end,several run time interfaces are provided,and the application perspective is pointed out.