Distributed adaptive optimization-based formation tracking with double parameter projections for multi-agent systems

doi:10.1016/j.jfranklin.2022.05.041

Distributed adaptive optimization-based formation tracking with double parameter projections for multi-agent systems

2022-07-01

7

PRE

AI

DOI：10.1016/j.jfranklin.2022.05.041

Original

Original request for help

Save

Abstract

En 中文

In this paper, a distributed adaptive optimization-based formation tracking strategy with double parameter projections for multi-agent systems is addressed under a centralized task allocation and distributed task execution (CTA-DTE) framework. Since a pre-described formation strategy is unable to adapt to a complex and dynamic environment, an optimization-based approach is proposed to transfer the formation tracking problem into a time-varying optimization one, subject to some constraints with several time-varying parameters which determine the rule of formation configuration change adaptively. These parameters are computed by a centralized unit and allocated to each agent as a global mission. Furthermore, each agent cooperates with others to execute this mission under a distributed optimization-based strategy, which combines a geometric center observer technology and a novel double parameter projections technology. The former ensures accurate tracking of a continuous reference trajectory. The latter guarantees that all agents enter into a time-varying security region and never escape from it, and meanwhile, all agents are attracted towards the best time-varying formation configuration via a gradient descent with a compensation. Finally, some simulation results are illustrated to verify the effectiveness of the strategy. (c) 2022 The Franklin Institute. Published by Elsevier Ltd. All rights reserved.

Keywords:

VARYING FORMATION TRACKING

PREDICTIVE CONTROL

AVOIDANCE

Distributed adaptive optimization-based formation tracking with double parameter projections for multi-agent systems

Abstract

Journal

Organization