Machine Learning-Enhanced Feature Engineering for High-Fidelity Quasi-Isentropic Waveform Prediction under Data Sparsity Constraints
Source abstract
Abstract
The quasi-isentropic loading method based on multidimensional gradient structures plays a critical role in acquiring dynamic physical parameters and significantly enhances the service safety of materials. This study proposes a physics-guided machine learning (PGML) framework for high-fidelity prediction of quasi-isentropic loading waveforms under data-scarce conditions. The framework systematically integrates physical principles into the learning process through mathematically regulated inductive biases and deep feature engineering. By integrating shock propagation physics and an attention mechanism, this approach transforms sparse data into interpretable visual representations, achieving an R 2 > 0.96 and 18.5 m/s MAE with only 528 samples while reducing shape alignment error by over 35% compared to baselines. The 4 × 4 augmentation strategy delivers a breakthrough performance in small-data prediction, achieving radically enhanced accuracy and training efficiency across impact velocities. A PGML framework, constrained by shock wave propagation behavior, coupled with SHAP analysis, reveals that the Hill coefficient ( h ) and curvature modulation parameter ( K ) dominate the loading path, providing interpretable insights into waveform modulation. The presented framework establishes a transformative, data-efficient paradigm for graded structure material design, substantially diminishing the dependency on resource-intensive simulations and experiments.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Model Reduction and Neural Networks · Machine Learning in Materials Science · Cellular and Composite Structures
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
State Key Lab of Advanced Technology for Materials Synthesis and Processing
Wuhan University of Technology
School of Materials Science and Engineering
Hubei Longzhong Laboratory
School of Mathematics and Statistics
Funding
Department of Science and Technology of Hubei Province · Basic and Applied Basic Research Foundation of Guangdong Province · China Postdoctoral Science Foundation · Natural Science Foundation of Hubei Province · National Natural Science Foundation of China · Xiangyang City
References (49)
Show source-provided references
1. 10.1126/science.abg148710.1126/science.abg1487 ↗
2. 10.1016/j.ijimpeng.2025.10544010.1016/j.ijimpeng.2025.105440 ↗
3. 10.1021/am507725k10.1021/am507725k ↗
4. 10.1021/acsami.9b1288010.1021/acsami.9b12880 ↗
5. 10.1088/0965-0393/22/2/02502110.1088/0965-0393/22/2/025021 ↗
6. 10.1016/j.ijimpeng.2015.02.01010.1016/j.ijimpeng.2015.02.010 ↗
7. 10.1007/s00193-015-0558-310.1007/s00193-015-0558-3 ↗
8. 10.1063/12.000104210.1063/12.0001042 ↗
9. 10.1016/j.ejor.2021.04.03210.1016/j.ejor.2021.04.032 ↗
10. 10.1038/s41598-024-73643-x10.1038/s41598-024-73643-x ↗
11. 10.1016/j.ijimpeng.2023.10455810.1016/j.ijimpeng.2023.104558 ↗
12. 10.1016/j.cma.2025.11804610.1016/j.cma.2025.118046 ↗
13. 10.1007/s11831-024-10063-010.1007/s11831-024-10063-0 ↗
14. 10.1002/idm2.1218010.1002/idm2.12180 ↗
15. 10.1016/j.jmrt.2024.03.10610.1016/j.jmrt.2024.03.106 ↗
16. 10.1016/j.msea.2024.14774010.1016/j.msea.2024.147740 ↗
17. 10.1007/s10853-025-10913-710.1007/s10853-025-10913-7 ↗
18. 10.1016/j.ijplas.2021.10307210.1016/j.ijplas.2021.103072 ↗
19. 10.1016/j.engfailanal.2024.10870010.1016/j.engfailanal.2024.108700 ↗
20. 10.1016/j.compositesb.2014.05.02410.1016/j.compositesb.2014.05.024 ↗
21. 10.1155/2020/883783810.1155/2020/8837838 ↗
22. 10.1088/1742-6596/500/14/14203410.1088/1742-6596/500/14/142034 ↗
23. 10.1051/epjconf/2015940106510.1051/epjconf/20159401065 ↗
24. 10.1016/j.compstruct.2025.11940810.1016/j.compstruct.2025.119408 ↗
25. 10.1038/s41586-018-0337-210.1038/s41586-018-0337-2 ↗
26. 10.1021/acsami.2c0889110.1021/acsami.2c08891 ↗
27. 10.1021/acsami.2c1531110.1021/acsami.2c15311 ↗
28. 10.1016/j.infrared.2025.10582210.1016/j.infrared.2025.105822 ↗
29. 10.1016/j.actamat.2017.11.05310.1016/j.actamat.2017.11.053 ↗
30. 10.1016/j.compstruct.2019.11126410.1016/j.compstruct.2019.111264 ↗
31. 10.1016/j.addma.2025.10470110.1016/j.addma.2025.104701 ↗
32. 10.1016/j.compositesb.2024.11175110.1016/j.compositesb.2024.111751 ↗
33. 10.1016/j.matdes.2020.10850910.1016/j.matdes.2020.108509 ↗
34. 10.1016/j.matdes.2024.11293210.1016/j.matdes.2024.112932 ↗
35. 10.1016/j.ijmecsci.2025.11023010.1016/j.ijmecsci.2025.110230 ↗
36. 10.1016/j.dt.2024.07.00710.1016/j.dt.2024.07.007 ↗
37. 10.1109/TASE.2025.354993210.1109/TASE.2025.3549932 ↗
38. 10.1109/TII.2024.345220310.1109/TII.2024.3452203 ↗
39. 10.1016/j.jclepro.2021.12907210.1016/j.jclepro.2021.129072 ↗
40. 10.1002/idm2.1223310.1002/idm2.12233 ↗
41. 10.1016/j.cmpb.2021.10658410.1016/j.cmpb.2021.106584 ↗
42. 10.1021/acsami.4c2227210.1021/acsami.4c22272 ↗
43. 10.1021/acsami.4c0877810.1021/acsami.4c08778 ↗
44. 10.3390/sym1611142610.3390/sym16111426 ↗
45. 10.1002/adma.20220090810.1002/adma.202200908 ↗
46. 10.1016/j.eswa.2021.11589210.1016/j.eswa.2021.115892 ↗
47. 10.48550/arXiv.1807.0652110.48550/arXiv.1807.06521 ↗
48. 10.48550/arXiv.1703.0154110.48550/arXiv.1703.01541 ↗
49. 10.48550/arXiv.1106.181310.48550/arXiv.1106.1813 ↗