|
|||||
|
|
| 一种用于布尔逻辑设计的神经网络学习算法 | |
| 引用本文： | 刘新平,李宗民,纪友芳,孙士明. 一种用于布尔逻辑设计的神经网络学习算法[J]. 中国石油大学学报(自然科学版), 2001, 25(3) |
| 作者姓名： | 刘新平 李宗民 纪友芳 孙士明 |
| 作者单位： | 石油大学计算机科学系, |
| 摘 要： | 提出了一种实现任意布尔函数的前向三层神经网络学习算法 ,简称HKML算法。HKML算法引入汉明距离扩展、卡诺图化简及最小项抑制等思想训练神经网络 ,以尽量少的神经元实现布尔函数的输入输出映射。给出了设计原理及算法流程图。与其他二进神经网络学习算法相比较 ,此方案结构简单可靠 ,学习速度快 ,易于硬件实现 |
| 关 键 词： | 布尔函数 神经网络 学习算法 数字逻辑 |
INVESTIGATION AND ACTUALIZATION OF A NEW MULTI VARIABLE FUZZY CONTROL METHOD |
|
| LIU Xin ping,et al.. INVESTIGATION AND ACTUALIZATION OF A NEW MULTI VARIABLE FUZZY CONTROL METHOD[J]. Journal of China University of Petroleum (Edition of Natural Sciences), 2001, 25(3) | |
| Authors: | LIU Xin ping et al. |
| Abstract: | In order to overcome the deficiency in high precision control and the complication of multi variable system of fuzzy controller, a new method to design a multi variable fuzzy controller is given. The method greatly simplifies the design program of such complex multi variable system, and can also resolve the contradiction between precision and complexity by drawing a dynamic compensate model into the common controller. A water tank temperature controller with high precision is designed. Testing result shows that this method has better control performances. |
| Keywords: | multi variable system fuzzy control dynamic compensate model controller |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |