基于数据驱动的Al-Cu合金多目标性能模型预测

Data-driven multi-objective property model prediction in Al-Cu alloys

摘要

铸造铝合金因其优异的力学性能广泛应用于航空航天、汽车等领域，但传统合金设计面临成分空间庞大、试错实验成本高和成分与性能之间非线性关系难以预测的问题。本工作提出一种反向传播神经网络、主成分分析和遗传算法相结合的机器学习模型，用于铸造铝合金的多目标性能预测。该模型通过反向传播神经网络非线性映射建立合金成分与性能的关系、主成分分析降维、遗传算法优化网络参数，从而提升预测精度和训练效率。结果表明，优化后的模型均方误差、决定系数和平均绝对误差分别为 36.28、0.91 和 2.44，在极限抗拉强度、屈服强度和断后伸长率的实验验证中，预测值与实验值控制在±5% 误差范围内，具有较高预测精度，证明该模型具有高效性与可靠性。

Abstract

Cast aluminum alloys are widely used in aerospace, automotive and other industries due to their excellent mechanical properties. However, traditional alloy design faces challenges such as vast composition space, high costs of trial-and-error experiments and difficulty in predicting the nonlinear relationship between composition and properties. This paper proposes a machine learning model that combines backpropagation neural networks, principal component analysis, and genetic algorithms for multi-objective property prediction of cast aluminum alloys. The model establishes the relationship between alloy composition and properties through the nonlinear mapping of backpropagation neural networks, reduces dimensionality via principal component analysis, and optimizes network parameters using genetic algorithms-thereby improving prediction accuracy and training efficiency. The results show that the optimized model has mean squared error of 36.28, correlation coefficient of 0.91, and mean absolute error of 2.44. In the experimental verification of ultimate strength, yield strength, and elongation after fracture, the error between experimental values and predicted values is controlled within the range of ±5%. This high prediction accuracy demonstrates the efficiency and reliability of the proposed model.

关键词

铸造铝合金 / 主成分分析 / 反向传播神经网络 / 遗传算法 / 力学性能

Key words

cast aluminum alloy / principal component analysis / backpropagation neural network / genetic algorithm / mechanical property

参考文献

[1] AKHTAR N, AKHTAR W, WU S J . Melting and casting of lithium containing aluminium alloys[J].International Journal of Cast Metals Research, 2015, 28(1): 1-8.

[2] 王一唱, 袁灵洋, 杨磊, 等 . Al-Zn 系高强铸造铝合金强韧化研究现状与展望[J].稀有金属材料与工程, 2023, 52(11): 3954-3970.

[3] WANG Y C, YUAN L Y, YANG L, et al. Strengthening-toughening of high-strength Al-Zn cast aluminum alloys: research progress and prospects[J].Rare Metal Materials and Engineering, 2023, 52(11): 3954-3970.

[4] NIU G D, WANG Y, ZHU L J, et al. Fluidity of casting Al-Si series alloys for automotive light-weighting: a systematic review[J].Materials Science and Technology, 2022, 38(13): 902-911.

[5] 何玄, 宋鹏, 孔德昊, 等 . 机器学习技术在航空材料领域的应用[J].航空材料学报, 2025, 45(3): 85-104.

[6] HE X, SONG P, KONG D H, et al. Application of machine learning technology in field of aviation materials[J].Journal of Aeronautical Materials, 2025, 45(3): 85-104.

[7] 邢清源, 臧金鑫, 陈军洲, 等 . 超高强铝合金研究进展与发展趋势[J].航空材料学报, 2024, 44(2): 60-71.

[8] XING Q Y, ZANG J X, CHEN J Z, et al. Research progress and development tendency of ultra-high strength aluminum alloys[J].Journal of Aeronautical Materials, 2024, 44(2): 60-71.

[9] ABOULKHAIR N T, MASKERY I, TUCK C, et al. The microstructure and mechanical properties of selectively laser melted AlSi10Mg: the effect of a conventional T6-like heat treatment[J].Materials Science and Engineering: A, 2016, 667: 139-146.

[10] MA K K, WEN H M, HU T, et al. Mechanical behavior and strengthening mechanisms in ultrafine grain precipitation-strengthened aluminum alloy[J].Acta Materialia, 2014, 62: 141-155.

[11] 张思平, 曹海龙, 吴锡伟, 等 . 航空装备用 2014 铝合金高温变形行为研究[J].铝加工, 2025(1): 42-46.

[12] ZHANG S P, CAO H L, WU X W, et al. Research on high temperature deformation behavior of 2014 aluminium alloy for aviation equipment[J].Aluminium Fabrication, 2025(1): 42-46.

[13] YANG Z W, CHEN C, LI D, et al. An additively manufactured heat-resistant Al-Ce-Sc-Zr alloy: microstructure, mechanical properties and thermal stability[J].Materials Science and Engineering: A, 2023, 872: 144965.

[14] TUAN N Q . Effects of substituting ytterbium for scandium on the microstructure and properties of Al-Sc and Al-Mg-Sc alloys[D]. Braga:Universidade do Minho (Portugal), 2014.

[15] 李晓强, 杜长林, 王俊, 等 . 锻造 7050 铝合金力学性能的 RBF 神经网络预测模型[J].航天制造技术, 2021(3): 15-18.

[16] LI X Q, DU C L, WANG J, et al. Prediction of mechanical properties of forged 7050 aluminum alloy based on RBF neural network[J].Aerospace Manufacturing Technology, 2021(3): 15-18.

[17] 魏萌, 肖凌. 基于 BP 神经网络的铝合金锻造性能预测[J].热加工工艺, 2021, 50(21): 105-107.

[18] WEI M, XIAO L . Prediction of forging performance of aluminum alloy based on BP neural network[J].Hot Working Technology, 2021, 50(21): 105-107.

[19] 崔鑫, 张建平, 张能辉. 基于 PSO-BP 模型的 5083 铝合金力学性能预测[J].锻压技术, 2019, 44(6): 183-187.

[20] CUI X, ZHANG J P, ZHANG N H . Prediction of mechanical properties for 5083 aluminum alloy based on PSO-BP model[J].Forging & Stamping Technology, 2019, 44(6): 183-187.

[21] 王硕, 王俊升, 梁婷婷, 等 . 高强、高导铝合金研发的机器学习策略[J].材料热处理学报, 2023, 44(11): 27-34.

[22] WANG S, WANG J S, LIANG T T, et al. Development of high-strength, high-conductivity aluminum alloys: a machine learning strategy[J].Transactions of Materials and Heat Treatment, 2023, 44(11): 27-34.

[23] 李婷. 基于不同机器学习算法的铝合金性能预测[J].有色设备, 2023, 37(4): 66-71.

[24] LI T . Prediction of aluminum alloy properties based on different machine learning algorithms[J].Nonferrous Metallurgical Equipment, 2023, 37(4): 66-71.

[25] 张普. 基于机器学习的 2xxx 系铝合金成分优化及组织与性能研究[D]. 成都: 西南交通大学,2023.

[26] ZHANG P . Study on machine learning-based composition optimization and microstructures and properties of 2xxx series aluminum alloys[D]. Chengdu:Southwest Jiaotong University, 2023.

[27] 定巍, 苗以升, 张坛, 等 . 基于半监督模型的铝合金性能预测与成分设计[J].邢台职业技术学院学报, 2023, 40(5): 46-53.

[28] DING W, MIAO Y S, ZHANG T, et al. The performance prediction and composition design of aluminum alloy based on semi-supervised model[J].Journal of Xingtai Polytechnic College, 2023, 40(5): 46-53.

[29] HART G L, MUELLER T, TOHER C, et al. Machine learning for alloys[J].Nature Reviews Materials, 2021, 6(8): 730-755.

[30] LIANG T T, WANG J S, XUE C P, et al. Design of high strength and electrically conductive aluminium alloys by machine learning[J].Materials Science and Technology, 2022, 38(2): 116-129.

[31] BUNDELA A S, RAHUL M R . Application of machine learning algorithms with and without principal component analysis for the design of new multiphase high entropy alloys[J].Metallurgical and Materials Transactions A, 2022, 53(10): 3512-3519.

[32] 魏雨晨, 李旭东, 刘治国, 等 . 基于 GA-BP 神经网络的航空铝合金预腐蚀疲劳寿命预测[J].环境技术, 2022, 40(5): 178-182.

[33] WEI Y C, LI X D, LIU Z G, et al. Prediction of pre-corrosion fatigue life of aviation aluminum alloy based on GA-BP neural network[J].Environmental Technology, 2022, 40(5): 178-182.

[34] 丁凤娟, 贾向东, 洪腾蛟, 等 . 基于 GA-BP 和 PSO-BP 神经网络的 6061 铝合金板材流变应力预测模型[J].稀有金属材料与工程, 2020, 49(6): 1840-1853.

[35] DING F J, JIA X D, HONG T J, et al. Flow stress prediction model of 6061 aluminum alloy sheet based on GA-BP and PSO-BP neural networks[J].Rare Metal Materials and Engineering, 2020, 49(6): 1840-1853.

[36] 王荣桥, 蒋康河, 胡殿印, 等 . 基于主成分分析的镍基单晶高温疲劳寿命模型[J].航空动力学报, 2016, 31(6): 1359-1367.

[37] WANG R Q, JIANG K H, HU D Y, et al. High temperature fatigue life model for single crystal nickel superalloy based on principal component analysis[J].Journal of Aerospace Power, 2016, 31(6): 1359-1367.

[38] 陈奎, 张天云, 郑小平, 等 . 主成分分析法的改进及其在工程材料综合评价中的应用[J].机械工程材料, 2013, 37(7): 90-93.

[39] CHEN K, ZHANG T Y, ZHENG X P, et al. Improvement of principal component analysis and its application in engineering materials comprehensive evaluation[J].Materials for Mechanical Engineering, 2013, 37(7): 90-93.

[40] 胡倩倩, 韩啸, 刘吉辉, 等 . 基于 GA-BP 神经网络的 LF 精炼过程合金加入量预测模型[J].冶金自动化, 2025, 49(1): 32-41.

[41] HU Q Q, HAN X, LIU J H, et al. Prediction model of alloy addition amount in LF refining process based on GA-BP neural network[J].Metallurgical Industry Automation, 2025, 49(1): 32-41.

[42] 秦一梅. 软计算技术在 TiAl 合金显微硬度的预测研究[D]. 西安: 陕西科技大学,2018.

[43] QIN Y M . Prediction on micro-hardness of TiAl alloy based on soft computing[D]. Xi’an: Shaanxi University of Science and Technology,2018.

[44] 张浩, 王国文, 曾凡宜, 等 . 基于 BP 神经网络的 6082 铝合金固溶时效处理后的晶粒尺寸预测[J].锻压技术, 2023, 48(3): 227-235.

[45] ZHANG H, WANG G W, ZENG F Y, et al. Grain size prediction of 6082 aluminum alloy after solution and aging treatment based on BP neural network[J].Forging & Stamping Technology, 2023, 48(3): 227-235.

[46] MA Z, JI H C, HUANG X M, et al. Research on high temperature stamping forming performance and process parameters optimization of 7075 aluminum alloy[J].Materials, 2021, 14(19): 5485.

[47] 梅益, 孙全龙, 喻丽华, 等 . 基于 GA-ELM 的铝合金压铸件晶粒尺寸预测[J].金属学报, 2017, 53(9): 1125-1132.

[48] MEI Y, SUN Q L, YU L H, et al. Grain size prediction of aluminum alloy dies castings based on GA-ELM[J].Acta Metallurgica Sinica, 2017, 53(9): 1125-1132.

[49] LI B, DU Y, ZHENG Z S, et al. Manipulation of mechanical properties of 7xxx aluminum alloy via a hybrid approach of machine learning and key experiments[J].Journal of Materials Research and Technology, 2022, 19: 2483-2496.

[50] 侯丽娜. 微合金化对 Al-Cu-Mg 合金组织及性能的影响[D]. 沈阳: 沈阳工业大学,2010.

[51] HOU L N . The effects of minor alloying on the microstructure and mechanical properties of Al-Cu-Mg alloy[D]. Shenyang:Shenyang University of Technology, 2010.

[52] 潘利文, 陈永斌, 周勇, 等 . Cr 对 Al-Cu-Mn 合金显微组织及力学性能的影响[J].广西大学学报（自然科学版）, 2024, 49(4): 835-846.

[53] PAN L W, CHEN Y B, ZHOU Y, et al. Effects of Cr on microstructure and mechanical properties of Al-Cu-Mn alloy[J].Journal of Guangxi University (Natural Science Edition), 2024, 49(4): 835-846.

基金资助

国家科技部创新方法专项(2020IM020700)