A Dynamic Multi-modal deep Reinforcement Learning framework for 3D Bin Packing Problem

Anhao Zhao, Tianrui Li, Liangcai Lin

Abstract: Highlights•Introducing DMRL-BPP, significantly reducing gap ratios in 3D-BPP.•Novel dynamic multi-modal encoder for effective multi-modal data handling.•New reward function to efficiently tackle sparse rewards in 3D bin packing.•Achieved 7.8%, 2.2%, 1.4% improvements in gap ratios for 16, 20, 30 boxes over current 3D-BPP methods.