Multi-objective intelligent optimization design method for mix proportions of hydraulic asphalt concrete facings

Abstract

Hydraulic asphalt concrete facing serves as a critical impervious structure, ensuring the long-term and stable operation of the upper reservoir in pumped-storage power station. However, conventional mix design theories still rely heavily on empirical criteria focused on optimizing single material properties, which fall short of meeting the increasingly multi-dimensional demands of modern engineering design. To facilitate the intelligent and efficient construction of impervious facings, this research proposes a multi-objective mix design approach that integrates statistical experimental optimization with metaheuristic search techniques, considering material performance, economic cost, and environmental sustainability. First, response surface models were established between mix parameters and material properties using the central composite design (CCD) method. The significance and predictive accuracy of the models were validated through analysis of variance (ANOVA) and residual analysis. Subsequently, a set of material performance constraints was formulated based on engineering requirements, and a multi-objective optimization model was constructed with economic cost and carbon emissions as dual objectives. The Non-dominated sorting whale optimization algorithm (NSWOA) was then employed to derive the Pareto optimal solutions iteratively. Finally, the technique for order preference by similarity to ideal solution (TOPSIS) was applied to perform decision analysis on the Pareto optimal front, enabling the selection of optimal compromise mix proportions under different preferences. Specifically, the optimal trade-off solution between resource-saving and environmentally-friendly corresponds to an asphalt content of 6.705%, a filler content of 12.680%, and a gradation index of 0.264. The results provide a valuable theoretical reference for the efficient and sustainable design of asphalt concrete facings.