TY - JOUR
T1 - Cooperative Optimization of Tasks and Trajectories for Multiple Unmanned Ground Vehicles Based on Path Perception
AU - Hu, Mingjing
AU - Li, Xueyuan
AU - Su, Yunquan
AU - Ye, Junhan
AU - Yang, Fan
N1 - Publisher Copyright:
© 2026, China Ordnance Industry Corporation. All rights reserved.
PY - 2026
Y1 - 2026
N2 - The task planning for cooperative multiple unmanned ground vehicles (UGVs) needs to take into account the actual trajectory constraints of the vehicles. However, integrating both task allocation and trajectory planning into a unified model often leads to low computational efficiency, which limits its practical applicability. This paper proposes a path-aware-based loosely-coupled cooperative optimization method for task and trajectory. A 2.5D off-road map is utilized to enhance the A* algorithm for acquiring feasible path information. A Hungarian algorithm is employed to generate an allocation scheme coupled with the path information. The cooperative trajectory planning is formulated as an optimal control problem, which is solved iteratively using the Runge-Kutta discretization method and an interior-point method, yielding a cooperative task-trajectory strategy for the system. The results demonstrate that the path-aware-based loosely-coupled cooperative optimization method outperforms the unified modeling approach in terms of reducing the computational complexity and improving the strategy generation efficiency, and has excellent adaptability and robustness in large-scale scenarios. Real-world vehicle validation further confirms its reliability in practical applications.
AB - The task planning for cooperative multiple unmanned ground vehicles (UGVs) needs to take into account the actual trajectory constraints of the vehicles. However, integrating both task allocation and trajectory planning into a unified model often leads to low computational efficiency, which limits its practical applicability. This paper proposes a path-aware-based loosely-coupled cooperative optimization method for task and trajectory. A 2.5D off-road map is utilized to enhance the A* algorithm for acquiring feasible path information. A Hungarian algorithm is employed to generate an allocation scheme coupled with the path information. The cooperative trajectory planning is formulated as an optimal control problem, which is solved iteratively using the Runge-Kutta discretization method and an interior-point method, yielding a cooperative task-trajectory strategy for the system. The results demonstrate that the path-aware-based loosely-coupled cooperative optimization method outperforms the unified modeling approach in terms of reducing the computational complexity and improving the strategy generation efficiency, and has excellent adaptability and robustness in large-scale scenarios. Real-world vehicle validation further confirms its reliability in practical applications.
KW - multi-vehicle cooperative engagement
KW - off-road path planning
KW - task allocation
KW - trajectory optimization
KW - unmanned ground vehicle
UR - https://www.scopus.com/pages/publications/105054172160
U2 - 10.12382/bgxb.2026.0193
DO - 10.12382/bgxb.2026.0193
M3 - Article
AN - SCOPUS:105054172160
SN - 1000-1093
VL - 47
JO - Binggong Xuebao/Acta Armamentarii
JF - Binggong Xuebao/Acta Armamentarii
M1 - 260193
ER -