.png){kind=icon}
.png){kind=icon}
Return
Optimization of direct extrusion process for Nd-Fe-B magnets using active learning assisted by machine learning and Bayesian optimization
.png){kind=icon}
OA
AI
DOI：10.1016/j.scriptamat.2021.114341.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Process optimization of permanent magnet is time-consuming as the microstructure that depends on alloy compositions and process parameters must be optimized to achieve high coercivity. Given a raw material of fixed composition, the optimization of process involves the refinement of grains size, the alignment of crystallographic orientation and the formation of intergranular phase. In this paper, we implemented an active learning pipeline assisted by machine learning and Bayesian optimization (ALMLBO) for predicting magnetic properties from process parameters and propose optimum conditions leading to high coercivity and remanence in Nd-Fe-B anisotropic magnets fabricated by direct hot extrusion. ALMLBO allowed us to optimize the process to exhibit high coercivity, mu 0 H c -1.7 T, and remanence, mu 0 B r -1.4 T, simultaneously, resulting in an excellent maximum energy product, (BH) max -380 kJ/m 3 . We show that an ALMLBO pipeline is an effective tool for optimizing process for Nd-Fe-B anisotropic magnets. (c) 2021 The Authors. Published by Elsevier Ltd on behalf of Acta Materialia Inc. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )
Keywords:
Active learning
Machine learning
Bayesian optimization
Nd-Fe-B magnet
Hot extrusion
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
5.6
Papers:
1.6W
Citations:
5.1W