| 摘要: |
| 针对离散化方法在焊接智能建模和控制中的重要性和目前研究存在的不足,以粗糙集智能建模方法为例,采用脉冲钨极氩弧焊的一个预测焊接熔池背面宽度的模型为背景,对离散化方法进行研究。确立了离散化方法在焊接建模领域的主要评价标准。据此对比了常用的适合焊接过程的离散化方法,并得出基于熵的离散化算法更好。最后,在基于熵算法的基础上提出了一种改进的离散化方法。结果表明,该算法稳定且更加适合焊接领域的建模。 |
| 关键词: 离散化 焊接过程 粗糙集 智能建模 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金资助项目(60474036) |
|
| An new discretization algorithm for intelligent modeling in welding |
|
LI Wen-hang, CHEN Shan-ben, LIN Tao, DU Quan-ying
|
|
Institute of Welding Engineering, Shanghai Jiaotong University, Shanghai 200030, China
|
| Abstract: |
| Discretization method is usually important in intelligentmodeling or controling for welding process.The evaluation criterion for discretiztion methods inwelding field were discussed.Based on the selected evaluation criterion, common discretization methods were compared, where a welding data set was obtain in pulsed GTAW experiment as modeling data and rough set modeling method was used to obtained a model that can predict backside width of melten pool.It showed that discretizationmethod based entropy performes better.Based on the entropy algorithm, a modified discretization algorithm was put forward, and the last result showed that this method was better and suited for the welding field. |
| Key words: discretization welding process rough set intelligent modeling |
