Optimization of some properties of hydraulic asphalt concrete mix using RSM and NSGA-II

The traditional mix ratio optimization research is affected by the operation conditions, and most of them only consider a single target, which cannot fully consider the correlation between different performance indicators. In this study, considering the multi-objective demand, to meet the basic mechanical properties of asphalt concrete requirements, to seek both economic and low carbon emissions of the theoretical asphalt concrete mix ratio. The response surface method (RSM) and non-dominated sorting genetic algorithm-II (NSGA-II) were combined to balance the economy and low carbon of hydraulic asphalt concrete. The Box-Behnken design (BBD) was used to arrange a series of experimental studies. The filler asphalt ratio X 1 , mortar aggregate ratio X 2 , and fine aggregate ratio X 3 were used as design variables, and porosity, stability, and flow value were used as response values. According to the technical indicators, the spatial constraints of the mix ratio were constructed. Through NSGA-II, the Pareto optimal solution under the goal of economy and low carbon is sought, and the linear weighted sum method is used to obtain the optimal mix ratio scheme with different biases. The research shows that:(1) The fusion of RSM-NSGA-II can effectively optimize the multi-objective optimization of the mixture ratio; (2) Regression fitting to establish the prediction model of performance with respect to mix ratio parameters is highly significant, and the established spatial constraints can greatly reduce the screening range of mix ratio parameters. (3) The optimized mix ratio of environmental protection bias is X 1 = 1.2, X 2 = 0.21, X 3 = 0.6, the optimized mix ratio of economic bias is X 1 = 1.2, X 2 = 0.22, X 3 = 0.59, and the optimized mix ratio of balanced type is X 1 = 1.2, X 2 = 0.218, X 3 = 0.6. This study has theoretical guiding significance for mix ratio design in engineering practice.