Optimization of Asphalt Mixture Mix Proportion Based on Bayesian Optimization Algorithm Combined with Marshall Test

Guodong Yuan; Qinyu Tang

doi:10.54097/71hg7z11

Authors

Guodong Yuan
Qinyu Tang

DOI:

https://doi.org/10.54097/71hg7z11

Keywords:

Asphalt mixture; mix proportion optimization; Bayesian optimization; Marshall test.

Abstract

This study presents an optimization framework that integrates physical testing with intelligent algorithms. With the amount of asphalt and the passing rates of key sieves aggregate gradation as design variables, and Marshall stability, flow value, void ratio, mineral aggregate void ratio, and asphalt saturation as optimization goals, a high-performance prediction model constructed. First, an initial data set is obtained through a small number of carefully designed Marshall tests (such as orthogonal tests, Latin hypercube sampling). Then, a Gaussian model is constructed to replace the complex “variables-performance” black box function. Finally, the Bayesian optimization algorithm is used, with the expected improvement as the acquisition function, to iter recommend the “most potential” mixture ratio for the next round of Marshall test verification until convergence. The results show that this method can quickly locate the Pareto optimal combination frontier that all technical specifications under the premise of significantly reducing the number of tests (expected to reduce by 40%-60%), and can quantify the contribution of each variable to the index. This study provides a new paradigm of data-driven and intelligent efficient for asphalt mixture design, which has important reference value for promoting the intelligent design of road engineering materials.

Downloads

Download data is not yet available.

References

[1] Liang Yongning. Investigation on Mix Proportion and Marshall Test of SMA-13 Asphalt Mixture fopressway Pavement [J]. China Highway, 2025, (16): 108-110.

[2] Shu Wei. Discussion oand Marshall Test of Asphalt Mixture for Thick Pavement [J]. Subgrade Engineering, 1998, (02)45.

[3] Yingcheng L, Weiguang Z, Yongli Z, et al. Mechanical Property Evaluation for Steel Slag in Asphalt Mixture with Different Skeleton Structures Using Modified Marshall Mix D Methodology [J]. Journal of Materials in Civil Engineering, 2022, 34(1): DOI:10.161/(ASCE)MT.1943-5533.0004014.

[4] Xu Jianbin. Comprehensive Report on Marshall Stability Test of Asphat Mixture [J]. Sichuan Cement, 2020, (08): 341-342.

[5] Hao Shuwei. Influence of Agte Gradation on Marshall Test Results of Asphalt Mixture [J]. Transport World (Construction & Maintenance. Machinery), 2010,(01): 200-202.

[6] Song M, Chen Z, Xia J, et al. Prediction of post-dryout heat transfer based on physics-embedded machine learning with Bayesian optimizati algorithm [J]. Progress in Nuclear Energy, 2026, 194: 106263-16263. DOI:10.1016/J.PNUCENE.2026.106263.

[7] "Technical Specifications for Construction of Highway Aspha Pavement" (JTG F40-2004)

[8] Yu Liwu. Civil Engineering Materials [M]. Nanjing Southeast University ress: 202305.

[9] Zhu Mingxuan, Li Wenkai, Shao Jinggan, et al. Study on Road Performance of Asphalt Mixtures with Different Gradationer Different Loading Conditions [J]. Henan Science, 2026, 44(01): 98-107.

[10] Nie Guihai. Theory, Design Method and Performance Evaluation of Flowing Asphalt Mixture [D]. Guangzhou University, 2025. DOI:10.27040/d.cnki.ggzdu.2025.000033.

[11] Xu Hanyong, Xu Wen, Lanhao. Study on Correlation between Stability and Dynamic Stability of Asphalt Mixture [J]. Transportation Science and Technology, 2025,06): 16-20.

[12] Feng Jun. Research on the Influence of Gradation Segregation on Asphalt Mixture Performance and Its Quality Control [D]. Chang'an Uniersity, 2024. DOI:10.26976/d.cnki.gchau.2024.001480.

[13] Zhu Zehua. Research on Noise Reduction Function and Long-term Acoustic Performance of Thermoplastic Polyurethane/Crumb Rubber Composite Modified Asphalt Pavement [D]. Yangzhou University, 2025. DOI:10.27441/d.cnki.gyzdu.2025000004.

[14] Qi Tingliang. Study on Fatigue Properties of Warm Mix Asphalt Mixture with High RAP Content [D]. Chongqing Jiaotong Unversity, 2025. DOI:10.27671/d.cnki.gcjtc.2025.001718.5] Chen Wenyi. Research on Aggregate Gradation Optimization of AC-13 Asphalt Mixture Based on Data Envelopment Analysis [D]. Jilin Jianzhu Unversity, 2024. DOI:10.27714/d.cnki.gjljs.2024.000172.

[15] Department of China Journal of Highway and Transport. Review of Academic Research on Chinese Subgrade Engineering · 2025 [J]. China Journal of Highway and Tansport, 2025, 38(12): 1-153. DOI:10.19721/j.nki.1001-7372.2025.12.001.

[16] Si, X. W. Study on Damage Characteristics of Grouting Semi-rigid Pavement Materials under High Temperature and Heavy Load Conditions [henyang Jianzhu University, 2024. DOI: 10.27809/d.cnki.gsjc.2024.000839.

[17] Jiang, B. Study on Road Performance of Composite Fiber Asphalt Mixture in High Humidity and Saline Environment [D]. Cngqing Jiaotong University, 2024. DOI: 10.27671/d.cnki.gcjc.2024.000185.

[18] Liao, Y. Q. Mix Design and Performance Research of Surface Modified Steel Slag Asphalt Mixture in Hot and Humireas [D]. Chongqing Jiaotong University, 2025. DOI: 10.27671/d.cnkigcjtc.2025.001748.

[19] Niu, H. Application of Bayesian Optimization in Parameter Combination Screening of Functional Textiles [J]. Dyeingg Technology, 2025, 47(09): 29-31.

[20] Chu, Y. L.Chen, H. K., Wang, J. Q., et al. Harmonic Source Modeling of Complex Signals Based on Multivariate Gaussian Process Regression and Spatial Clustering [J]. Transactions of China trotechnical Society, 2025, 40(23): 7621-7634. D: 10.19595/j.cnki.1000-6753.tces.242151.

[21] Lin, S. Y., Zou, D. Performancelysis of Two-tier Heterogeneous Cellular Networks Based on Generalized Matern Hard-core Point Process [J]. Application of Electronic Technique,025, 51(06): 54-58. DOI: 10.16157/j.issn.028-7998.246025.