Aggregate Blending Using Fuzzy Optimization

Abstract

Aggregate blending is a process that blends available aggregates to create a blend that meets gradation specifications while minimizing the unit cost of the blend. Today’s aggregate blending process is a multiobjective optimization problem that involves not only minimizing cost but also satisfying other specifications of the blend, such as fineness modulus, plasticity index, and specific gravity. Gradation limits are usually expressed in terms of the range of the percentage passing each sieve. In the traditional approach, rigid values are used for the gradation limits, and as a result, gradation and other specifications are frequently met at the limits during optimization, which is not particularly desirable. This paper proposes an approach for aggregate blending using fuzzy optimization method. The approach selects the best mix of aggregates such that not only are the individual gradation specifications met within the specification limits, but also their desirability within each range is satisfied as much as possible. In the proposed model, functions that represent the decision maker’s satisfaction with respect to the blend’s unit cost and physical properties are introduced. The proposed model is compared with two traditional optimization models using the example data found in two previous models. The results show that the proposed approach is useful for real-world application as it is robust in terms of its ability to deal with many constraints and objectives, the practitioner’s uncertainty about the limits of the specification ranges, and the desire to achieve different objectives.