Composites have been increasingly applied in aerospace and engineering construction due to their multi-physical field coupling and anisotropic characteristics. However, their complex nonlinear structures posed significant challenges to the traditional research and development paradigm. The conventional trial-and-error approach, which relied heavily on physical experiments, was limited by long cycles, high costs, and difficulties in exploring the vast design space. The introduction of machine learning technologies brought novel research insights to materials science, enabling the exploration of the deep structure–property relationships of materials through a data-driven approach. This study aims to provide a systematic review of the current application status and transformative potential of machine learning in composite materials research, structured across three primary dimensions: structural design, performance characterization, and process optimization. The article first elucidated how high-fidelity surrogate models developed through machine learning significantly enhanced the efficiency of optimizing structural design parameters. Furthermore, it introduced breakthroughs in multi-scale performance characterization, delving into the research frontiers by exploring data-driven performance prediction and the utilization of deep learning architectures for the precise formulation of constitutive relations. Finally, the paper discussed intelligent manufacturing technologies of composites based on intelligent regulation of process parameters and intelligent scheduling in the production flow. It points out the challenges and future development directions of further promoting composites research using machine learning in view of current data quality, model interpretability, and industrial chain integration in the machine learning field.
BaiX P, ZhangX C. Artificial intelligence-powered materials science[J]. Nano-Micro Letters, 2025, 17(1): 135.
[4]
AlamM A, SohelA, HasanK M, et al. Machine learning and artificial intelligence in diabetes prediction and management: a comprehensive review of models[J]. Innovatech Engineering Journal, 2024, 1(1): 107-124.
YangZhongwei, WangZhuangzhuang, DuanHongji, et al. Construction of interlayer interlocking structure in carbon fiber reinforced epoxy electromagnetic shielding material and its property regulation[J]. Journal of North University of China (Natural Science Edition), 2025, 46(3): 278-284. (in Chinese)
DaiMengyan, LiuZhe, PengWenlian, et al. Material discovery and generation accelerated by artificial intelligence[J]. CBRN Defense, 2022, 1(1): 53-58. (in Chinese)
[11]
KauweS K, RhoneT D, SparksT D. Data-driven studies of Li-ion-battery materials[J]. Crystals, 2019, 9(1): 54.
[12]
LiuY L, NiuC, WangZ, et al. Machine learning in materials genome initiative: a review[J]. Journal of Materials Science & Technology, 2020, 57: 113-122.
CuiGuoxiang, LiRui, YuanChangchi, et al. AI-driven design of next-generation interconnect materialsfor integrated circuits: progress and challenges[J]. Materials China, 2025, 44(5): 436-450. (in Chinese)
[15]
HuangJ S, LiewJ X, AdemiloyeA S, et al. Artificial intelligence in materials modeling and design[J]. Archives of Computational Methods in Engineering, 2021, 28(5): 3399-3413.
[16]
WangR Q, LiuY, LiuX, et al. Assessment of mechanical properties for three-dimensional needled composites: a geometric partitioning strategy dealing with mesoscopic needling damage[J]. Materials, 2022, 15(16): 5659.
JiaYuxiang, WangJiafu, ChenWei, et al. Research progress on rapid optimization design methods of metamaterials based on intelligent algorithms[J]. Journal of Radars, 2021(2): 220-239. (in Chinese)
XieBingbing, ZhaoFeng, GuoXinxing, et al. A probabilistic fatigue life prediction method for wind turbine towers based on a physics-informed neural network[J]. China Mechanical Engineering, 2026, 37(5): 1017-1025. (in Chinese)
[21]
BéjiH, KanitT, MessagerT. Prediction of effective elastic and thermal properties of heterogeneous materials using convolutional neural networks[J]. Applied Mechanics, 2023, 4(1): 287-303.
[22]
MäättäJ, BazaliyV, KimariJ, et al. Gradient-based training and pruning of radial basis function networks with an application in materials physics[J]. Neural Networks, 2021, 133: 123-131.
DuChuanzhong, LiYuwei. How the deep integration of AI and big data drives paradigms of scientific and technological innovation: mechanisms and pathways[J]. Social Science Research, 2026(1): 30-40. (in Chinese)
YinWanjian, LiuMiao, GongXingao. Paradigm shift in materials science driven by artificial intelligence: methods, platforms, and frontiers of applications[J]. Chinese Science Bulletin, 2025, 70(24): 4012-4014. (in Chinese)
[27]
WardC. Materials genome initiative for global competitiveness: [R/OL].2012-06-20.
[28]
KrizhevskyA, SutskeverI, HintonG E. ImageNet classification with deep convolutional neural networks[C]//Advances in Neural Information Processing Systems, 2012: 1097-1105.
[29]
GravesA, MohamedA R, HintonG. Speech recognition with deep recurrent neural networks[C]//2013 IEEE International Conference on Acoustics, Speech and Signal Processing, 2013: 6645-6649.
WangYalong, WangYiqiu. Research on technological innovation of performance optimization of polymer materials driven by artificial intelligence[J]. China Invention & Patent, 2025, 22(S1): 153-160. (in Chinese)
MauriziM, GaoC, BertoF. Predicting stress, strain and deformation fields in materials and structures with graph neural networks[J]. Scientific Reports, 2022, 12: 21834.
[38]
ChenC T, GuG X. Generative deep neural networks for inverse materials design using backpropagation and active learning[J]. Advanced Science, 2020, 7(5): 1902607.
ZhangHengrui, XuXijun, LiXunlu, et al. Applications of generative artificial intelligence in battery research: current status and prospects[J]. Acta Physico-Chimica Sinica, 2025, 41(10): 1-10. (in Chinese)
ShengYuqiang, AnShaohang, MaMengyao, et al. Application of artificial intelligence in novel molecular sieve gas adsorption materials[J]. Modern Applie Physics, 2025, 16(1): 199-205. (in Chinese)
[43]
MillerB, ZiemiańskiL. Accelerating multi-objective optimization of composite structures using multi-fidelity surrogate models and curriculum learning[J]. Materials, 2025, 18(7): 1469.
[44]
JiaQ T, XuK, LiC H, et al. Transfer learning-assisted multi-objective optimization of mechanical properties for particle reinforced aluminum matrix composites[J]. Materials Futures, 2026, 5(1): 015602.
YuanJie, WangJunbo, ChenXin, et al. Research progress on application of artificial intelligence in ultra-high performance concrete[J]. Journal of Jilin University (Engineering and Technology Edition), 2025, 55(3): 771-789. (in Chinese)
[47]
FangW, DuncanM, DuaM, et al. Machine learning-assisted design of multilayer thermoplastic composites: robust neural network prediction and feature importance analysis[J]. Macromolecular Materials and Engineering, 2025, 310(9): 70005.
HeLin, HeChangqing, SuiHong. Prospects for the creation of novel interfacial separation materials driven by artificial intelligence[J]. Chemical Industry and Engineering Progress, 2024, 43(4): 1649-1654. (in Chinese)
[50]
LiZ, DuJ, YiH, et al. Machine learning framework for evaluating microscale interface cohesive zone models of polymer composites[J]. ACS Applied Materials & Interfaces, 2025, 17(45): 62590-62605.
YaoLixiang, HuangXianjun, ChenHongting, et al. Advances and perspectives in artificial intelligence-empowered electromagnetic protection materials research[J]. High Power Laser and Particle Beams, 2025, 37(8): 089001. (in Chinese)
[55]
SunX H, YueL, YuL X, et al. Machine learning-evolutionary algorithm enabled design for 4D-printed active composite structures[J]. Advanced Functional Materials, 2022, 32(10): 2109805.
[56]
DuschlbauerD, BöhmH J, PettermannH E. Computational simulation of composites reinforced by planar random fibers: homogenization and localization by unit cell and mean field approaches[J]. Journal of Composite Materials, 2006, 40(24): 2217-2234.
[57]
PadovecZ, RužičkaM, SedláčekR, et al. Comparison of the thermoelastic properties of a randomly reinforced composite computed by the classical lamination theory and by the Monte Carlo simulation[J]. Mechanics of Composite Materials, 2017, 53(2): 149-158.
ZhangZhen, ZhuMingyu, FuXukai, et al. Hot deformation behavior and constitutive relationship of monel K-500 alloy based on machine learning[J]. Journal of Netshape Forming Engineering, 2025, 17(11): 242-251. (in Chinese)
QuTongming, FengYuntian, WangMengqi, et al. Constitutive relations of granular materials by integrating micromechanical knowledge with deep learning[J]. Chinese Journal of Theoretical and Applied Mechanics, 2021, 53(9): 2404-2415. (in Chinese)
[65]
MujtabaA, IslamF, KaedingP, et al. Machine-learning based process monitoring for automated composites manufacturing[J]. Journal of Intelligent Manufacturing, 2025, 36(2): 1095-1110.
[66]
XieJ B, ChenX M, ZhangY F, et al. Experimental and numerical investigation of the needling process for quartz fibers[J]. Composites Science and Technology, 2018, 165: 115-123.
QiYunchao, FangGuodong, LiangJun, et al. Optimization of process parameters of three-dimensional needled preforms for C/C-SiC composites[J]. Journal of Materials Engineering, 2020, 48(1): 27-33. (in Chinese)
[69]
WuC C, ZhangR M, ZhaoP Y, et al. Curing simulation and data-driven curing curve prediction of thermoset composites[J]. Scientific Reports, 2024, 14: 31860.
ZhangRunbo, YuBaifeng, YuJiayun. Research on the application of artificial intelligence in the manufacturing of fiber reinforced composite materials[J]. Hi-Tech Fiber and Application, 2024, 49(6): 37-47. (in Chinese)
CaiZhiliang, HuangJieming, YaoXuexuan, et al. The application of artificial intelligence in the development of energy storage battery materials and chemical engineering[J]. Leather and Chemicals, 2025, 42(5): 4-6. (in Chinese)