A palletize-planning system for multiple kinds of loads using GA search and traditional search

Abstract

In logistics, it was impossible to automatically palletize different kinds of loads on pallets without human thinking. We have developed a planning system for multiple kinds of loads by hybridizing a GA (to make the combinations of loads and pallets) and the traditional beam search algorithm that allows us to use heuristic knowledge in loading process (to stack loads on a pallet effectively in space). The system has shown the performance to make a plan in reasonable time, and has enabled a picking machine to automate the work of palletizing multiple kinds of loads. The developed system is successfully applied to multiload automatic storage.