|
| 摘要: |
| 本文以神经网络为工具，探讨了Ni基喷涂材料Si、Cr、B等元素对其涂层硬度的影响规律，在工程和实验的实际样本的基础上，以均匀设计方法优化了网络结构和学习对参数，得到能正确反映元素-硬度相关性的人工神经网络，计算结果表明：神经网络法不失为正确复现这一复杂相关性有力的新工具，文中指出：以各无素偏相关指数绝对值的大小判别其对Ni基喷涂材料硬度影响程度，效果亦佳。 |
| 关键词: 人工神经网络,喷涂材料,均匀设计,偏相关指数 |
| DOI： |
| 分类号:TG17 |
| 基金项目: |
|
| Investigation Regularities of Ni-based Spraying and Coating Materials'''' Elements'''' Influence on Hardness Via neural Networks |
|
Lu Wenjiang Tang Fuling Liu Shuqun Yn Weiyuan
|
| Abstract: |
| Artificil neural network has been used to investigate the regularities of elements influence on sparying and coating materials' hardness, and partial correlation index to find the relative importance of various elements affecting the hardness,The results of computation indicated that neural networks are new effective assistant technical to rehabilitate this complex relativity exactly, and partial correlation index's affect was good. |
| Key words: artificial neural networks uniform design crossover validarion partial correlation index nickel-based spraying and coating materials |