| 前馈人工神经网络的变步长直接搜索训练算法 |
| |
| 引用本文： | 邢济收,王信义.前馈人工神经网络的变步长直接搜索训练算法[J].北京理工大学学报,1997,17(2):189-193. |
| |
| 作者姓名： | 邢济收 王信义 |
| |
| 作者单位： | 北京理工大学机械工程学院 |
| |
| 摘 要： | 提出了一种用于前馈人工神经网络的变步长直接搜索训练算法，在网络学习时，以多样本在网络的各输出单元上的误差平方和作为目标函数，因它是权重系数的多元函数，这样将问题转化为多元函数的极值问题求解，对误差采取“批处理”的 权重系数的变步长直接修正方法，使该算法具有比较快的收敛速度和较高的收敛精度。
|
| 关 键 词： | 人工神经网络 故障诊断 变家长 直接搜索法 |
Directly Searching Algorithm for Multi-Layer Feed-forward Artificial Neural Networks |
| |
| Xing Jishou,Wang Xinyi,Zhang Weimin,Xu Chunguang,Wang Keyong School of Mechanical Engineering,Beijing Institute of Technology,Bejing.Directly Searching Algorithm for Multi-Layer Feed-forward Artificial Neural Networks[J].Journal of Beijing Institute of Technology(Natural Science Edition),1997,17(2):189-193. |
| |
| Authors: | Xing Jishou Wang Xinyi Zhang Weimin Xu Chunguang Wang Keyong School of Mechanical Engineering Beijing Institute of Technology Bejing |
| |
| Institution: | Xing Jishou,Wang Xinyi,Zhang Weimin,Xu Chunguang,Wang Keyong School of Mechanical Engineering,Beijing Institute of Technology,Bejing 100081 |
| |
| Abstract: | This paper presents a new learning algorithm for multi-layer feed-for-ward artificial neural networks. The subject transforms into an optimal value ofthe multivariate function through taking the error square sum of all output unit onall samples as objective function, that is multivariate function of weighting coeffi-cient.Using batch method of the output error and step-changing directly adjustingmethod of the weight,the algorithm has fast learning speed and high precision. |
| |
| Keywords: | artificial neural networks optimum algorithm fault diagnosis monitoring |
| 本文献已被 CNKI 维普 等数据库收录！ |
| 点击此处可从《北京理工大学学报》浏览原始摘要信息 |
| 点击此处可从《北京理工大学学报》下载免费的PDF全文 |
