Control and Information Processing Functions using Chaotic Dynamics in Neural Network Model and Cellular Automata

结果表明，包含混沌的复杂动力学可以用简单规则实现复杂函数和复杂控制。这两个计算机实验已经完成。一种是使用神经网络模型，另一种是使用细胞自动机，前者使我们能够理解大脑记忆动力学和生物控制中观察到的优秀控制的可能机制。后者使我们不仅能够分析声音信号、音乐和声音所包含的复杂性，而且能够以完全可逆的方式压缩数据。利用神经网络模型，对二维迷宫的典型病态问题进行了计算机实验。用数值方法对其性能进行了评价。提出了硬件实现方案。所有的工作都在会议上报告，并提交给几个科学或工程期刊。

It is shown that complex dynamics including chaos can be useful in realizing of complex functions and complex control by means of simple rule (s) . The two computer experiments have been done. One is to use a neural network model and the other is to use cellular automata The former enables us to understand a possible mechanism of memory dynamics in brain and excellent control observed in biological control. The latter enables us not only to analyze complexity contained in sound signal, music and voices, but also to compress data in completely reversible way. Using the neural network model, computer experiment to solve two-dimensional maze which is set as a typical ill-posed problem was done. The performance is evaluated by numerical methods. A plan of hardware implementation is proposed. All works were reported in conference meetings and were submitted to several scientific or engineering journals.