| 基于改进人工神经网络的SMT质量智能鉴别 |
| |
| 引用本文： | 徐湖,潘尔顺. 基于改进人工神经网络的SMT质量智能鉴别[J]. 工业工程与管理, 2009, 14(6) |
| |
| 作者姓名： | 徐湖 潘尔顺 |
| |
| 作者单位： | 上海交通大学,机械与动力工程学院,上海,200240 |
| |
| 基金项目： | 国家自然科学基金项目 |
| |
| 摘 要： | 采用了一种改进的BP神经网络,针对BP神经网络的不足进行了改进:采用变学习率法减少网络训练时间、采用高斯惩罚函数避免局部最小值,并使整个网络能自主调整其隐层节点的数量.运用改进的BP神经网络对于样本进行训练,训练后的神经网络能够较为精确的预测SMT产品质量问题.
|
| 关 键 词： | 人工神经网络 BP算法 |
Intelligently Predict the Quality of the SMT Products based on the Artificial Neural Network |
| |
| XU Hu,PAN Er-shun. Intelligently Predict the Quality of the SMT Products based on the Artificial Neural Network[J]. Industrial Engineering and Management, 2009, 14(6) |
| |
| Authors: | XU Hu PAN Er-shun |
| |
| Abstract: | An improved BP neural network is adopted in this paper.To solve the Defect of the BP network,several methodologies are used,which include taking changeable learning rate to shorten training time,taking Gauss Punishment function to avoid local minimum and automatically adjusting the number of the hidden layer nerve cells.After training with the sample,the improved BP neural network can predict the quality problem of the SMT products accurately. |
| |
| Keywords: | SMT SMT artificial neural network BP algorithm |
| 本文献已被 万方数据 等数据库收录！ |
|
