Application of genetic algorithms for sequencing of AS/RS with a triple-shuttle module in class-based storage

Materials handling systems, as a main support to the dynamism of logistic systems, must be highly productive, well utilised, and very efficiently controlled. In the case of high volume, large capacity warehouse systems, an AS/RS (automatic storage/retrieval system) is a typical solution to these requirements. The performance of an AS/RS is closely related to the implementation of different control policies. The sequencing policy, used to create minimum overall handling time tours, has a great impact on system performances. The selection of an appropriate sequencing method is even more significant in a more complex case of multiple shuttle S/R (storage/retrieval) devices because of the mutual paring between more than one storing and retrieving tasks. This paper is focused on the sequencing problem of a triple-shuttle AS/RS in a class-based storage system under a modified sextuple command cycle policy with a planning horizon that comprises the realisation of several successive cycles of S/R device. To solve the problem, three commonly used greedy heuristics (nearest neighbour, reversed nearest neighbour, and shortest leg) are adapted, and a genetic algorithm is proposed.