基于事件触发机制的多智能体系统脉冲一致性研究

How to achieve consensus with limited on-board resources, limited computing, communicating and actuating capabilities is one of the key problems for the applications of consensus theory of multi-agent systems. In this project, the advantage of impulsive control and event-based control theories will be effectively applied to study the consensus of multi-agent systems. The asynchronous event-triggering time sequences will be determined by some reasonable distributed triggering function and triggering condition and the event-based impulsive controller will be designed. The event-based networked impulsive control method will be presented by improving the stability theory of impulsive large scale systems with asynchronous impulsive sampling instants. The consensus conditions of multi-agent systems will be obtained under complex environment such as time delays and time-varying topologies. Furthermore, the Zeno-behavior of the event-triggering time sequences will be excluded. Moreover, the analytical relationship between the convergence rate and the impulsive sampling instants will be established to optimize convergence rate and controller update frequency. The implementation of this project will improve the method of impulsive control and event-based control. Theories and technologies of multi-agent systems based on digital platform will be enriched and developed, which will present new ideas and methods on theory analysis, control and design of complex systems.

如何在通信网络带宽、在线资源、智能体自身计算处理能力等受限的条件下，实现多智能体系统一致性，是多智能体系统一致性理论实际应用的关键问题之一。本项目充分运用脉冲控制与基于事件触发控制的优势，通过设计合理的分布式事件触发函数与激励条件，确定异步事件触发时间序列，设计基于事件触发机制的脉冲控制器，发展脉冲采样时刻点异步的脉冲大系统稳定性理论，建立一套基于事件触发机制的网络化脉冲控制方法。获得多智能体系统在复杂网络环境下达到一致的条件，弄清楚时滞、动态拓扑结构等因素对多智能体系统一致性的影响。排除事件触发时间序列可能存在的Zeno现象，建立一致性收敛速度与脉冲采样时刻点之间的解析关系，优化多智能体系统一致性收敛速度与控制器更新频率。项目的实施将丰富与发展多智能体系统数字化控制的新理论、新技术，有效克服现有脉冲控制方法与事件触发控制方法的不足，为复杂系统的理论分析、控制与设计提供新的思想和方法。

如何在通信网络带宽、在线资源、智能体自身计算处理能力等受限的条件下，实现多智能体系统一致性，是多智能体系统一致性理论实际应用的关键问题之一。本项目基于脉冲（时滞）系统稳定性理论、事件触发控制理论，开展了分数阶多智能体系统一致性、具有时滞的多智能体系统一致性、动态拓扑与量化通信下的多智能体系统一致性、基于事件触发机制的多智能体系统一致性、基于自适应事件触发机制的多智能体系统一致性、多智能体系统异步脉冲一致性、基于事件触发机制的连续动态系统脉冲控制理论、基于事件触发机制的多智能体系统脉冲一致性、基于事件触发机制的多智能体系统编队等系列问题研究，建立了基于事件触发机制的网络化脉冲控制方法，取得一批具有较高学术水平的多智能体系统一致性新理论、新成果，在Automatica、IEEE Transactions on Neural Networks and Learning Systems、Science China: Information Sciences等期刊上发表SCI检索学术论文21篇，申请专利4项，其中已授权实用新型发明专利1项。本项目的研究为复杂系统的分析、设计提供有力的思想方法和工具。