| BP神经网络在摩擦学设计计算中的应用 |
| |
| 引用本文： | 徐建生,赵源. BP神经网络在摩擦学设计计算中的应用[J]. 机械设计, 2000, 17(10): 16-18 |
| |
| 作者姓名： | 徐建生 赵源 |
| |
| 作者单位： | 1. 武汉化工学院,机械系,湖北,武汉,430073
2. 武汉材料保护研究所 |
| |
| 基金项目： | 国家自然科学基金!资助项目 (595750 34) |
| |
| 摘 要： | 用双隐层ＢＰ人工神经网络，建立了丝杆螺母副的磨损率与滑动速度关系的数学模型。该模型可用于准确地计算丝杆螺母副和蜗轮蜗杆副的磨损率，可十分方便地用于摩擦学程序设计。采用Ｌ－Ｍ规则进行神经网络学习训练可使网络收敛快，误差小。风络输出结果与实验结果比较有极好的吻合性。该神经网络为工程设计人员，在摩擦学设计时提供有效的计算工具。
|
| 关 键 词： | 摩擦学设计 神经网络 L－M训练规则 滑动速度 |
The application of BP neural network in the calc ulation of tribology design |
| |
| XU Jian sheng et al. The application of BP neural network in the calc ulation of tribology design[J]. Journal of Machine Design, 2000, 17(10): 16-18 |
| |
| Authors: | XU Jian sheng et al |
| |
| Affiliation: | Wuhan Institute of Chemical Technology |
| |
| Abstract: | By the use of double hidden layered BP artificial neural network a mathematical model of relationship between wear rate of screw-nut pair and sliding speed is established.This model can be used in the accurate calculation of wear rate of screw-nut pair and worm-worm wheel pair,and can be applied quite conveniently to tribology programming as well.By adopting the L-M rule,the learning and training of neural network carried out in this paper enable the network a quicker convergence and a less error.The output result of network possesses extreme conincidence compared with the result of experiment.This neural network can provide an effective calculation means for personnels of engineering design while doing their tribology design. |
| |
| Keywords: | Tribology design Neural network L-M training rule Sliding speed |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
