基于事件触发数字通信的多智能体系统分布式协同控制与优化

In this project the features like bandwidth constraints of digital channel and on-board resource constraints exist in multi-agent systems of actual complex networks are researched in mines. The project aims at solving the problem for distributed coordination control and optimization of multi-agent systems in these networks and carries out research in distributed coordination control and optimization of multi-agent systems with event-triggered digital communications. The advantages of digital communication and event-triggered communication control method will be combined to design reliable dynamic communication strategies, high-performance triggering functions and triggering conditions, and the algorithms of distributed coordination control and optimization of multi-agent systems with good robustness, easy expansibility and no steady-state error will be obtained and the Zeno-behavior of the event-triggering time sequences will be excluded. Meanwhile, for different networks, more pervasive algorithms of distributed coordination control and optimization will be present in the project. The analytical relation between the algorithm performance and network parameters will be further explored to optimize algorithmic design and enrich relevant theory, and provide a referential theoretical framework of distributed coordination control and optimization of multi-agent systems. The fulfillment of the project will extend and expand the design of the algorithm of distributed coordination control and optimization of multi-agent systems in resource-constrained networks, overcome the defects of the prior methods effectively, enrich and develop theories and technologies of event-triggered digital communications.

本项目针对现实复杂网络中多智能体系统存在的数字信道带宽受限和在线资源受限等问题，以解决此类网络下多智能体系统分布式协同控制与优化问题为目标，开展对基于事件触发数字通信的多智能体系统分布式协同控制与优化问题的研究。通过结合数字通信技术与事件触发机制的优势，设计可靠的动态通信策略和高性能的事件触发函数与激励条件，获得具有较强鲁棒性、可扩展性且无稳态误差的多智能体系统分布式协同控制与优化算法，并排除事件触发时间序列可能存在的芝诺现象。同时，针对不同的网络拓扑结构，提出更为普适的协同控制与优化算法，进一步探寻算法性能与网络参数之间的解析关系，进而优化算法设计和丰富相关理论，为多智能体系统分布式协同控制与优化提供可参考的理论框架。本项目的完成能够延伸和扩展资源受限网络下的多智能体系统分布式协同控制与优化算法的设计，有效克服现有方法的不足,丰富和发展基于事件触发数字通信技术的理论与应用。

本项目主要研究内容侧重于考虑现实复杂网络中多智能体系统存在的通信信道带宽受限和在线资源受限等问题，研究基于事件触发机制的数字通信技术，解决上述网络环境下的多智能体系统分布式协同控制与优化问题。通过项目组成员之间的团结合作，本项目充分挖掘了有限带宽数字通信技术与事件触发控制技术的优势，设计了可靠动态通信策略和高性能事件触发函数与激励条件。针对信息传输时延、数据丢包等通信不确定性情形，提出了具有强鲁棒性且无稳态误差的多智能体系统分布式协同控制与优化算法，排除了事件触发时间序列可能存在的芝诺行为。针对复杂的网络拓扑结构，提出了更具一般性的协同控制与优化算法，解析了算法性能与网络参数之间的关系，发展了多智能体系统分布式协同控制与优化的理论框架。本项目扩展了多智能体系统分布式协同控制与优化算法在网络带宽、在线资源、智能体自身计算能力等受限情形下的设计方法，丰富了基于事件触发的数字通信控制理论。