| 基于BP神经网络的引线键合模型 |
| |
| 引用本文： | 罗智芸,高健,陈新.基于BP神经网络的引线键合模型[J].半导体技术,2008,33(11). |
| |
| 作者姓名： | 罗智芸 高健 陈新 |
| |
| 作者单位： | 广东工业大学,机电工程学院,广州,510090 |
| |
| 摘 要： | 提出了一种基于BP神经网络的新建模方法.利用MATLAB神经网络函数建立网络,通过组合不同参数,将50组训练样本输入网络多次训练,比较结果误差及训练步数、时间,确定了最佳网络结构及参数设置值,建立了引线键合模型.采用训练样本外的10组实验数据对模型进行验证,分析多种误差.验证结果表明,该方法对建立引线键合模型是有效的、可行的,有着较高的精确度.
|
| 关 键 词： | 引线键合 BP神经网络 MATLAB神经网络函数 |
Wire Bonding Model Based on BP Neural Network |
| |
| Luo Zhiyun,Gao Jian,Chen Xin.Wire Bonding Model Based on BP Neural Network[J].Semiconductor Technology,2008,33(11). |
| |
| Authors: | Luo Zhiyun Gao Jian Chen Xin |
| |
| Affiliation: | Luo Zhiyun,Gao Jian,Chen Xin(Faculty of Electromechanical Engineering,Guangdong University of Technology,Guangzhou 510090,China) |
| |
| Abstract: | A new modeling method was proposed based on BP neural network.Using MATLAB neural network functions,a network was set up,50 groups of samples were put in the network to train for several times,through combining with different parameters,then the errors,train epochs and train time were analyzed.Based on it,the best network structure and the wire bonding model were established.The model was verified with 10 groups of experimental data out of the train sample and the error was analyzed.The test result shows th... |
| |
| Keywords: | wire bonding BP neural network MATLAB neural network function |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
