基于数据驱动的氢粉碎工艺氢含量的预报

详细信息

Forecasting of Hydrogen Content in Hydrogen Decrepitation Process Based on Data Drive

1.
School of Information Engineering of Inner Mongolia University of Science and Engineering, Baotou 014010, China. 2.Liaoyang Branch of CPE., Liaoyang 111003, China

摘要: 针对目前钕铁硼氢粉碎工艺过程中氢含量无法在线检测、动力学模型难以建立等问题,提出利用数据驱动方法实现钕铁硼氢粉碎过程中氢含量的预报,详细介绍了基于LS-SVM的氢粉碎工艺氢含量预报模型的建模过程。预测结果表明,基于LS-SVM的氢含量预报模型对氢粉碎工艺中氢含量的预测误差较小,可满足生产工艺的要求。
- 钕铁硼材料 /
- 氢粉碎工艺 /
- 氢含量预报 /
- 在线检测 /
- 数据驱动 /
- LS-SVM模型
Abstract: In order to solve problems that hydrogen content cannot be detected on-line and kinetic model is difficult to be established in NdFeB hydrogen decrepitation process at present, the paper proposed a scheme which uses data driven method to implement forecasting of hydrogen content, and introduced modeling process of forecasting model of hydrogen content based on LS-SVM in details. The forecasting result showed that the model can forecast hydrogen content with small errors which would meet requirement of NdFeB hydrogen decrepitation process.
- NdFeB, hydrogen decrepitation technics, forcast of hydrogen content, on-line detection, data driven, LS-SVM model