Multi‐Objective Path Planning for Unmanned Data Collection Vehicles Based on Deep Reinforcement Learning
Concurrency and Computation: Practice and Experience · 2026-11 · 38 (22)
DOI: 10.1002/cpe.71039 ↗Kurz gefasst
Fachartikel in der Zeitschrift Concurrency and Computation: Practice and Experience, online erschienen am 09.11.2026. Verfasst von Can Tang und 2 weiteren Autor:innen. Die Arbeit nennt 25 Literaturangaben.
Abstract (Originalsprache)
In rugged and complex field environments, unmanned ground vehicles (UGVs) are widely used to collect network data from distributed sensor nodes. However, due to the large number and wide distribution of data sensing sites, planning a path for a UGV to travel from the starting point through multiple sub‐goal points to the final destination requires substantial time and energy. To address the challenges of network convergence difficulty, local optima, and low planning efficiency in multi‐objective path planning for UGVs, this paper proposes an improved DDQN algorithm—DEDT DDQN. The algorithm enhances planning performance in complex multi‐objective environments by introducing a high‐quality experience differentiation selection mechanism and dynamically integrating prior knowledge from DDQN and Averaged DDQN. Simulation modeling in real‐world field scenarios verifies the effectiveness of the proposed algorithm. Experimental results show that DEDT‐DDQN achieves path lengths comparable to those of classical and reinforcement learning baselines while reducing the number of turning points and increasing the frequency of generating high‐quality feasible paths during training. Validation in a DEM‐derived real‐terrain environment further demonstrates its applicability to complex terrain scenarios.
Zitieren
APA
Tang, C., Peng, T., & Chen, X. (2026). Multi‐Objective Path Planning for Unmanned Data Collection Vehicles Based on Deep Reinforcement Learning. Concurrency and Computation: Practice and Experience, 38(22). https://doi.org/10.1002/cpe.71039
Vancouver
Tang C, Peng T, Chen X. Multi‐Objective Path Planning for Unmanned Data Collection Vehicles Based on Deep Reinforcement Learning. Concurrency and Computation: Practice and Experience. 2026;38(22). doi:10.1002/cpe.71039
BibTeX
@article{Tang2026,
author = {Tang, Can and Peng, Tao and Chen, Xuewei},
title = {Multi‐Objective Path Planning for Unmanned Data Collection Vehicles Based on Deep Reinforcement Learning},
journal = {Concurrency and Computation: Practice and Experience},
year = {2026},
volume = {38},
number = {22},
doi = {10.1002/cpe.71039}
}
EUVIMED