| 基于BP训练数据选取的压气机特性曲线计算 |
| |
| 引用本文： | 梁超,吴新跃.基于BP训练数据选取的压气机特性曲线计算[J].计算机与数字工程,2013,41(10):1593-1595. |
| |
| 作者姓名： | 梁超 吴新跃 |
| |
| 作者单位： | 海军工程大学机械工程系 武汉430033 |
| |
| 摘 要： | 基于BP神经网络方法，研究了压气机特性曲线的拟合及BP神经网络结构设计。对不同训练数据选取方法和比例情况下的BP神经网络进行了计算，分析BP神经网络数据选取与网络泛化能力之间的关系。分析结果表明，均匀顺序选择的BP神经网络在压气机特性曲线仿真上具有更优的泛化能力。
|
| 关 键 词： | 压气机 BP神经网络 泛化能力 数据选取 |
Calculation of Compressor Characteristics Based on Sample Selection in BP Neural Network |
| |
| LIANG Chao
,
WU Xinyue.Calculation of Compressor Characteristics Based on Sample Selection in BP Neural Network[J].Computer and Digital Engineering,2013,41(10):1593-1595. |
| |
| Authors: | LIANG Chao WU Xinyue |
| |
| Affiliation: | (Department of Mechanical Engineering, Naval University of Engineering, Wuhan 430033) |
| |
| Abstract: | Based on the method of BP neural network function, a calculation of compressor characteristics is made. The relation be- tween samples selection and generalization ability of BP neural network is analyzed, the BP neural network in the cases of different sample se- lection methods and data proportion are calculatated. The result shows that the neural network utilizing the method of uniformly order of se- lection, has better generalization. |
| |
| Keywords: | compressor BP neural network generalization ability sample selection |
| 本文献已被 维普 万方数据 等数据库收录！ |
