Stochastic Priority Model for Aggregate Blending

Abstract

Aggregate blending models that incorporate the optimization of two objectives with their priority levels are presented. The two objectives include the minimization of the mean deviation (or mean absolute deviation) from the midpoint of specification limits and the minimization of the unit cost of the blend. The models are applicable to any number of aggregates and can be used to provide the optimum proportions corresponding to a given priority level or to establish trade‐off curves between mean deviation and cost. The stochastic elements of aggregate gradations are formulated and incorporated into the models. Both the deterministic and stochastic models are applied to a numerical aggregate blending problem, and extensions of the models to accommodate some practical cases are presented.