.png){kind=icon}
Return
Distributed model predictive control with asynchronous optimization for multi-agent systems over directed graphs
.png){kind=icon}
PRE
AI
DOI：10.1016/j.automatica.2026.113296.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This paper investigates the distributed model predictive control (DMPC) problem for linear multi-agent systems under both decoupled and coupled constraints over directed graphs. By utilizing the Lagrangian method and the Fenchel conjugate function, we reformulate the DMPC optimization problem into its dual problem to address coupled constraints. To solve this dual problem, we propose a novel asynchronous distributed push-sum constrained optimization (ADPSCO) algorithm, which eliminates the need for clock synchronization, significantly improving computational efficiency. Under mild conditions on the fixed step-size, we prove that the proposed ADPSCO algorithm converges to the optimal solution at an R -linear rate of O ( δ k ) with 0 < δ < 1 . We also develop a distributed stopping criterion to terminate the ADPSCO algorithm when the solutions reach the tolerance accuracy while satisfying both decoupled and coupled constraints, which avoids infinite iterations. Furthermore, we propose an ADPSCO-based iterative DMPC approach, and prove its recursive feasibility and the stability of the resulting closed-loop system. Finally, a simulation example demonstrates that our approach outperforms state-of-the-art methods.
Keywords:
Linear multi-agent systems
Distributed model predictive control
Asynchronous distributed push-sum constrained optimization algorithm
R-linear convergence rate
Distributed stopping criterion
Journal
IF:
5.9
Papers:
1.2W
Citations:
5.2W
Organization
Cited Papers
No cited papers available