AS/R system design via dual-objective nonlinear mixed integer programming

In designing an AS/RS, the major objective is to provide a system with maximal operation efficiency at a minimal cost under certain system constraints. The paper presents a design strategy to meet such an objective using a dual-objective nonlinear mixed integer programming technique. To simplify the design model without losing generality, key design parameters of a unit load steady state AS/RS are defined and incorporated into the system. It is then solved by a branch-and-bound algorithm in a compact computer program that converges to optimal solutions in few iterations. A shoe-making AS/RS design example is presented to verify the effectiveness of the method. The design process results in all optimal design curve comprising solutions of various weights among different storage assignment types. The design curve thus obtained is used to perform trade-off study in tayloring a customized AS/RS. The result of simulations indicates that the proposed method is effective and reliable for AS/RS design and evaluation.