This paper investigates the applicability of artificial neural network (ANN) for the prediction ofthe flow number of dense asphalt-aggregate mixtures. Percentages of coarse aggregate, filler,bitumen, air voids, voids in mineral aggregate, and Marshall Quotient were employed as thepredictor variables. A comprehensive experimental database was used for the development of themodel. The statistical measures of coefficient of determination, coefficient of efficiency, rootmean squared error, and mean absolute error were used to evaluate the performance of themodel. Sensitivity and parametric analyses were conducted and discussed. The ANN modelaccurately characterizes the flow number of asphalt mixtures resulting in a very good predictionperformance. The proposed model remarkably outperforms several existing prediction modelsfor the flow number of asphalt mixtures.
展开▼