项目名称: 基于事件触发机制的多智能体系统分布式协调控制研究
项目编号: No.61203027
项目类型: 青年科学基金项目
立项/批准年度: 2013
项目学科: 自动化学科
项目作者: 樊渊
作者单位: 安徽大学
项目金额: 26万元
中文摘要: 在很多应用领域中,多智能体系统均需要由大规模、低成本的智能体个体组成。将事件触发机制用于该类系统可以有效突破智能体的通信消耗、能源消耗、计算能力和驱动能力等多种限制。本项目的研究内容为:(1)针对通信非对称和拓扑时变的多智能体系统建立数学模型,并利用事件向量空间划分和辅助函数轨迹分析方法研究系统的触发事件设计、触发行为分析和稳定性分析;(2)在聚集任务背景下,利用邻居状态反馈和目标点设定法研究分布式事件触发协调控制律设计;(3)通过触发数量分析建立通信和能源消耗性能指标,并利用极小值原理和线性矩阵不等式求解法研究性能指标下的事件触发优化控制律设计。通过上述研究,本项目将提出一套有效的适用于通信非对称和拓扑时变的多智能体系统的触发事件设计和分布式协调控制律设计方法,并在此基础上提出若干种通信和能源消耗性能指标下的优化控制律设计方法,为资源受限的大规模、低成本多智能体系统的应用提供理论支持。
中文关键词: 多智能体系统;事件触发;分布式协调控制;优化控制;
英文摘要: For multi-agent systems consisting of a multitude of low-cost individual agents, to employ the event-triggering mechanism is an effective way of overcoming many emergent constraints, like those in communication and energy consumption, computational capacity and actuation capability. Taking into account all these factors, this project will focus on the following aspects. Firstly, we will investigate the modeling for multi-agent systems with asymmetric communication links and time-varying network topologies. To study the triggering event design, the triggering behavior analysis and the stability analysis, both the event-vector-space partition approach and the auxiliary function trajectory analysis approach will be applied. Secondly, in the context of rendezvous tasks, we will use the neighboring agent state feedback and the target point determination approach to study the distributed coordination controller design. Thirdly, we will formulate the communication and energy consumption performance index based on analyzing the number of triggering times. Furthermore, we will use the minimum principle and linear matrix inequalities to study the event-triggered optimal controller design under the proposed performance index. The project aims to propose a set of effective approaches to event design and event-triggered cont
英文关键词: Multi-Agent Systems;Event-Triggering;Distributed Coordination Control;Optimized Control;