扰动线性多智能体系统的分布式资源分配算法
Distributed resource allocation algorithms for linear multi-agent systems with disturbances
时侠圣 1孙长银 1穆朝絮2
作者信息
- 1. 安徽大学自主无人系统技术教育部工程研究中心,合肥 230601;安徽大学安徽省无人系统与智能技术工程研究中心,合肥 230601;安徽大学人工智能学院,合肥 230601
- 2. 安徽大学自主无人系统技术教育部工程研究中心,合肥 230601;天津大学电气与自动化信息工程学院,天津 300192
- 折叠
摘要
针对一般异构线性多智能体系统的分布式资源分配问题,结合智能体动力学特性设计一类分布式优化算法,旨在实现扰动抑制的同时完成资源的最优分配.面对扰动模型已知干扰信号,本文利用智能体状态信息设计一种自适应分布式优化算法,其中拉格朗日(Lagrange)乘子对偶变量控制参数自动增加,直至拉格朗日乘子对偶变量达到一致.当智能体状态信息不可测时,本文将上述算法修改为基于输出反馈的分布式优化算法.当全局成本函数强凸时,借助拉萨尔(LaSalle)不变性引理证明上述所提算法渐近收敛到全局最优解.面对未知扰动信号,本文利用扩张状态观测器技术设计一种分布式优化算法.当未知扰动信号为常数或其微分趋于零时,利用李雅普诺夫(Lyapunov)理论证明上述算法渐近收敛至全局最优解.最后利用数值仿真验证上述算法有效.
Abstract
For the distributed resource allocation problem over general heterogeneous linear multi-agent systems,this study designs a class of distributed optimization algorithms based on the dynamics of agents to suppress disturbances and simultaneously complete the optimal allocation of resources.This study designs an adaptive distributed optimization employing the state information of each agent for structure-known disturbances.In addition,the control parameter of the dual-variable Lagrange multiplier adaptively increases until consensus is achieved.Furthermore,the proposed algorithm is modified to be an output-based distributed optimization algorithm when the state information is not accessible.Based on LaSalle's invariance principle,we show that the designed algorithms asymptotically converge to the optimal point when the global cost function is strongly convex.For unknown disturbances,this study proposes a distributed optimization algorithm based on an extended state observer.The proposed algorithm asymptotically converges to the theoretical optimal solution based on Lyapunov stability theory when the disturbance is constant,or its derivative tends to zero.Finally,the effectiveness of our proposed algorithms is illustrated by the provided simulation examples.
关键词
异构多智能体系统/资源分配/自适应通信权重/状态反馈/输出反馈/未知扰动Key words
heterogeneous multi-agent systems/resource allocation/adaptive communication weight/state feedback/output feedback/unknown disturbance引用本文复制引用
基金项目
国家自然科学基金创新研究群体项目(61921004)
国家自然科学基金重点项目(62136008)
国家自然科学基金重点项目(62236002)
出版年
2024
NETL
NSTL
万方数据