首页|具有复杂动力学的多智能体系统分布式优化综述

具有复杂动力学的多智能体系统分布式优化综述

扫码查看
多智能体系统分布式优化由于其高效性、灵活性和可靠性等特点吸引了大量学者的关注,在多机器人协同控制、无线传感器网络、能源系统等领域具有广泛的应用前景.分布式优化的基本目标是利用智能体的个体目标函数梯度、自身及其邻居状态信息设计分布式控制协议,驱动所有智能体的状态或输出到全局目标函数的最优解,系统动力学是影响智能体状态演化的重要因素.鉴于此,在回顾现有连续时间分布式优化算法的基础上,根据系统动力学分类,尽可能全面地评述具有复杂动力学的多智能体系统分布式优化问题的最新研究进展,并对未来发展方向进行展望.
A survey on distributed optimization for multiagent systems with complex dynamics
Distributed optimization for multiagent systems has attracted much attention on account of its high-efficiency,flexibility and reliability with extensive application prospects in cooperative control of multiple robots,wireless sensor networks,energy systems,etc.The basic goal of distributed optimization is designing a distributed control protocol by utilizing the individual objective function gradient and the state information of the agent and its neighbors to drive the states or outputs of all the agents towards the optimal solution of the global objective function.System dynamics is an important factor to affect the state evolution.Based on reviewing the research results on continuous-time distributed optimization algorithms,a systematic survey on the recent development of distributed optimization for multiagent systems with complex dynamics is conducted according to the categories of system dynamics.The future development directions for the research are also discussed.

distributed optimizationmultiagent systemscyber-physical systemscontinuous systemslinear systemsnonlinear systems

郭戈、康健

展开 >

东北大学流程工业综合自动化国家重点实验室,沈阳 110819

东北大学秦皇岛分校控制工程学院,河北秦皇岛 066004

东北大学信息科学与工程学院,沈阳 110819

分布式优化 多智能体系统 信息物理系统 连续系统 线性系统 非线性系统

国家自然科学基金国家自然科学基金

62173079U1808205

2024

控制与决策
东北大学

控制与决策

CSTPCD北大核心
影响因子:1.227
ISSN:1001-0920
年,卷(期):2024.39(7)