多智能体系统的分布式预测控制器设计  

Distributed Model Predictive Controller Design for Multi-Agent Systems

作　　者：王鹏[1] 丁宝苍[1] 韩崇昭[1] 

机构地区：[1]西安交通大学电信学院,西安710049

出　　处：《复杂系统与复杂性科学》2012年第1期82-87,共6页Complex Systems and Complexity Science

基　　金：国家自然科学基金(60934007;60921003;41074176);国家重点基础研究973项目(2007CB311006)

摘　　要：针对多智能体系统的跟踪及编队控制问题,研究其分布式预测控制器的设计方法。考虑时变通讯拓扑情况,在优化问题中附加一项辅助约束,保证了优化问题的递归可行性及多智能体系统的闭环指数稳定性。在每个采样时刻,各智能体根据通讯拓扑实时更新其代价函数及辅助约束,通过求解带有状态、输入及辅助约束的优化问题来计算其控制输入。仿真结果验证了结论的有效性。The distributed model predictive controller design for tracking and formation of multi-agent system is studied in this paper. Considering the time-varying communication topology, we present an auxiliary constraint in the optimization problem, which guarantees the recursive feasibility of the optimization and exponential stability of the overall closed-loop system. At each sam pling time, each agent solves a constrained optimization problem, respecting input, state and auxiliary constraints to obtain its input. The effectiveness of our result is verified through an il lustrative example.

关 键 词：分布式控制 预测控制 时变通讯拓扑 递归可行性 闭环稳定性 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置] N94[自动化与计算机技术—控制科学与工程]