Ni Sun and DanDan Wang
DFG algorithm, floor storage robot, obstacle avoidance, collaborative optimisation, 5G-MEC integration
E-commerce is growing at an average annual rate of 23%, driving research on multi-level warehouse dynamic obstacle avoidance and vertical resource scheduling to become a critical industry challenge. Traditional single-level algorithms often overlook the spatiotemporal correlation between elevator scheduling and automated guided vehicle (AGV) paths, resulting in 30% order delays and 25% AGV idle rates. Current research still faces challenges such as spatiotemporal decoupling, insufficient dynamic adaptability, and lack of multi-objective coordination. To address this, this study proposes the differential evolution fuzzy greedy fusion (DFG) algorithm, establishing a three-dimensional collaborative optimisation framework. In a four-level warehouse experiment, DFG achieved a 98.7% task completion rate, a 5% improvement over traditional differential evolution (DE). By integrating elevator waiting times, cross-level task response speeds increased by 3.2 times, with elevator utilisation exceeding 91.2%. The 5G-MEC architecture reduced dynamic response latency to 28 ms, enabling real-time scheduling of 1,500 orders per hour. DFG achieved zero path conflicts in narrow channels (1.5 m), with energy consumption fluctuations of only 5.3% and a 42% extension in battery life. This study provides a three-dimensional collaborative technical framework for intelligent warehousing.
Important Links:
Go Back
