!3936 [华为诺亚决策推理实验室][公司贡献][MindSpore][M2N]

Merge pull request !3936 from XiangCHEN.dvi/master

research/hpc/m2n/README.md

+# Mesh Movement Networks for PDE Solvers
+
+To cater to an abrupt requirement, this repository is hastily created to for the purpose of a [MindSpore](https://www.mindspore.cn/) implementation of the work [M2N: Mesh Movement Networks for PDE Solvers](https://arxiv.org/abs/2204.11188), to the best of our knowledge the first learning-based end-to-end mesh movement framework that can greatly accelerate the mesh adaptation process by 3 to 4 orders of magnitude, whilst achieving comparable numerical error reduction to traditional sota methods.
+
+Our code is originally written in PyTorch. We plan to translate it into MindSpore and release the code before the end of 2022.
+
+## Citation
+
+```latex
+@article{song2022m2n,
+  title={M2N: Mesh Movement Networks for PDE Solvers},
+  author={Song, Wenbin and Zhang, Mingrui and Wallwork, Joseph G and Gao, Junpeng and Tian, Zheng and Sun, Fanglei and Piggott, Matthew D and Chen, Junqing and Shi, Zuoqiang and Chen, Xiang and others},
+  journal={arXiv preprint arXiv:2204.11188},
+  year={2022}
+}
+```

research/hpc/m2n/README_CN.md

+# Mesh Movement Networks for PDE Solvers
+
+该项目旨在实现[M2N: Mesh Movement Networks for PDE Solvers](https://arxiv.org/abs/2204.11188)的[MindSpore](https://www.mindspore.cn/)版本。M2N是据我们所知的第一个基于学习的端到端的网格移动算法框架，可以在达到与传统sota方法（Monge–Ampere）的PDE数值误差相当水平的情况下，实现网格优化过程3-4个数量级的加速。
+
+我们的原始代码是用PyTorch实现的。由于通知比较紧急，我们计划在2022年底之前实现并开源MindSpore版本。
+
+## 引用
+
+```latex
+@article{song2022m2n,
+  title={M2N: Mesh Movement Networks for PDE Solvers},
+  author={Song, Wenbin and Zhang, Mingrui and Wallwork, Joseph G and Gao, Junpeng and Tian, Zheng and Sun, Fanglei and Piggott, Matthew D and Chen, Junqing and Shi, Zuoqiang and Chen, Xiang and others},
+  journal={arXiv preprint arXiv:2204.11188},
+  year={2022}
+}
+```