一种进化模糊逻辑控制器的新方法

结合进化学习分类器的密歇根和匹兹堡方法的优点,首次将对单条控制规则的评价引入了模糊逻辑控制器（ＦＬＣ）的进化过程中,解决了匹兹堡类型的学习分类器系统“强化信息的带宽窄”的问题,实现了ＦＬＣ在控制器级和规则级的同时进化,控制器的控制规则数目也可以自由变化,实验结果表明新方法有较高的效率,优化的模糊控制器的结构简单,性能良好。     

ELEMENTARY APPROACH TO THEORY OF FUNCTIONAL ORGANIZATION

The starting point of this paper is the well-known fact that the organization of a system in statistical physics can be expressed in terms of entropies, that is, as the difference between the entropies of the elements and the entropy of the whole. An analogous definition for goal-directed systems is proposed, in which a system is functionally organized if the success of the system toward attaining its goal exceeds the sum of the successes of the parts. This type of organization of goal-directed complex systems, referred to as functional, can be described within the framework of general systems theory by means of some basic concepts such as functional connection, cooperative dependence, and organizational function. A functional organization is defined as the actual manner in which the elements of a goal-directed complex system are functionally connected. A goal-directed complex system is characterized by an assembly of data from which certain quantitative measures of relevant properties of the functionally organized systems (e.g. their complexity and degree of organization) can be formed.     

模糊数学在卷烟直径测量中的应用

图像边缘检测的模糊方法能有效地将物体从背景中区分开来，本文介绍应用图像边缘检测的模糊数学方法测量卷烟直径的基本原理，并介绍了如何利用计算机实现整个测量过程，该系统具有较高的精确度。     

一种FCMAC及在Wiener模型辨识中的应用研究

本文将模糊算法和小脑模型神经网络有机地结合在一起，提出了一种单输入单输出(S ISO)的模糊小脑模型神经网络(FCMAC)．它在对输入进行分级量化的同时进行模糊量化，利 用Takagi Sugeno模糊算法进行推理，并将模糊算法引入CMAC的权值训练，具有输入量化级 数少、函数逼近精度高等特点．这种FCMAC用于Wiener模型辨识具有结构确定、计算量小、 训练速度快、辩识效果好等特点．