| 非线性函数逼近的神经网络方法 |
| |
| 引用本文： | 金炯华,钟秉林,黄仁.非线性函数逼近的神经网络方法[J].东南大学学报(自然科学版),1993(Z1):74-79. |
| |
| 作者姓名： | 金炯华 钟秉林 黄仁 |
| |
| 作者单位： | 东南大学机械工程系 |
| |
| 摘 要： | 针对工程实际中一类典型的非线性函数逼近问题，本文阐述了单层函数型网络及双层网络在应用时各自的特点，并提出了对网络输入节点进行优化选择，可提高网络的逼近精度，简化网络结构．文中给出了具体的选择方法，计算机模拟结果验证了该方法的有效性。
|
| 关 键 词： | 非线性函数 函数联接网络 输入节点优化 |
An Approximation of Nonlinear Function with Neural Network |
| |
| Jin Jioanghua Zhong Binglin Huang Ren.An Approximation of Nonlinear Function with Neural Network[J].Journal of Southeast University(Natural Science Edition),1993(Z1):74-79. |
| |
| Authors: | Jin Jioanghua Zhong Binglin Huang Ren |
| |
| Institution: | Department of Mechanical Engineering |
| |
| Abstract: | A method for the approximation of a typical nonlinear function is presented by neural network theory,in which the characteristics of a kind of function-linked neural network with one or two layers are studied in detail.The results show that the optimum selection of the input nodes is important in getting high accuracy and simplicity structure of neural network.Therefore a selection proceeding of input node is given,the emulation shows its availability. |
| |
| Keywords: | nonlinear function function-linked neural network input node optimization |
| 本文献已被 CNKI 维普 等数据库收录！ |
