事件触发机制下随机多智能体系统的有限时间一致性研究

项目名称： 事件触发机制下随机多智能体系统的有限时间一致性研究

项目编号： No.61503053

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 自动化技术、计算机技术

项目作者： 孙凤兰

作者单位： 重庆邮电大学

项目金额： 20万元

中文摘要： 多智能体系统的一致性是协调控制中的基本问题，也是当前控制领域的研究热点和前沿。由于实际中的系统经常处于复杂的环境中，网络结构可能是不确定的，因此研究随机多智能体系统的有限时间一致性是很有意义的。本项目针对随机多智能体系统，分别设计理想情形下和有传输时延或噪声的一致性协议，给出相应的事件触发函数和控制方法，研究事件触发机制下随机多智能体系统的有限时间一致性问题，得到系统达到有限时间一致的充分或必要条件，并从理论和实际两方面分析一致性协议的合理性和适用性。初步分析影响系统一致性和一致性性能的因素，得出事件触发控制系统的性能指标与表征系统环境的参数（时延、噪声等）之间的关系，采用合适的事件触发控制机制，设计鲁棒性强且易于实现的一致性协议。该研究结果将更加深刻地揭示复杂系统涌现性产生的内在规律，这对于发展网络化控制系统的理论研究和实际工程的应用都具有重要的意义和价值。

中文关键词： 随机网络；多智能体系统；有限时间一致性；事件触发控制

英文摘要： Consensus of multi-agent systems is not only the basic problem in the cooperative control, but also the research hotspot and the forefront in control field. Because the actual system is usually in complex environment, the network structure may be uncertain. So it’s very significance to study the finite-time consensus problem of stochastic multi-agent system. Based on the event-triggered control method, the consensus protocols both with and without the transmission delay or noise are designed in this project, and the corresponding event-triggered function and control methods are given. The sufficient or necessary condition for the finite-time consensus of stochastic multi-agent systems is given. Furthermore, the rationality and applicability of the consensus protocol is analyzed from both the theory and practical application aspects. By analyzing the influence factors of the consensus and consensus performance preliminarily, getting the relationship between the performance index of event-triggered control systems and the parameters of the system environment characterization (delay, noise, etc.). The consensus protocol with strong robustness and ease implementation is designed by adopting the appropriate event-triggered control mechanism. The result of this project will more deeply reveal the inherent law of emergent properties in the complex system, which is very important and significance both for the theoretical study of the networks control and the application of the actual engineering.

英文关键词： stochastic networks;multi-agent systems;finite-time consensus;event-triggered control