Diffusion Reinforcement Learning Based Online 3D Bin Packing Spatial Strategy Optimization

TL;DR

This paper introduces a diffusion reinforcement learning algorithm for online 3D bin packing, enhancing packing efficiency in logistics and manufacturing through a novel model-based approach.

Contribution

It presents a diffusion reinforcement learning framework with a Markov decision process and height map state representation, improving packing performance over existing DRL methods.

Findings

Significantly increased average number of packed items.

Effective in complex online packing scenarios.

Outperforms state-of-the-art DRL methods.

Abstract

The online 3D bin packing problem is important in logistics, warehousing and intelligent manufacturing, with solutions shifting to deep reinforcement learning (DRL) which faces challenges like low sample efficiency. This paper proposes a diffusion reinforcement learning-based algorithm, using a Markov decision chain for packing modeling, height map-based state representation and a diffusion model-based actor network. Experiments show it significantly improves the average number of packed items compared to state-of-the-art DRL methods, with excellent application potential in complex online scenarios.