.png){kind=icon}
Return
Trajectory Optimization for Drone Logistics Delivery via Attention-Based Pointer Network
.png){kind=icon}
PRE
AI
DOI：10.1109/TITS.2022.3168987.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Drone logistics delivery is a potential booster to redefine the logistics system efficiency, which has been a new special hot research field. Among that, how to optimize drone trajectory data and find optimal delivery path is a crucial problem. However, most existing studies fail to derive the feasible trajectory data due to underestimating drone energy consumption and delivery multi-variant constraints for air transportation. In this paper, we propose a novel self-driven learning procedure, named attention-based pointer network (A-Ptr-Net) model, which can solve drone delivery trajectory optimization problem. The generated A-Ptr-Net model coupled with attention mechanism is effective on adapting to the new drone trajectory data automatically, regardless of explicit distance matrix. We go into developing the convex function constraints related to drone nonlinear energy consumption, customer demand and service time windows, which is applied on A-Ptr-Net model for optimizing the drone logistics delivery. Numerical experiments indicate that the proposed method performs significantly better than classical heuristics for drone trajectory analysis and optimization.
Keywords:
Drones
Logistics
Heuristic algorithms
Energy consumption
Traveling salesman problems
Trajectory optimization
Mathematical models
Drone logistics delivery
trajectory optimization
pointer network model
attention mechanism
Journal
IF:
8.4
Papers:
9.7K
Citations:
6.3W
Organization
Cited Papers
.png){kind=icon}
ShareSaveDesign and simulation of a DC electric vehicle charging station connected to a MVDC infrastructure
PREAI
ShareSave
ShareSaveShareSaveShareSave
View more