线性多智能体系统的自适应事件触发协同控制

Multi-agent systems (MASs) have a strong engineering background and can find a broad scope of applications in many fields such as multi-robot cooperation, smart grids and logistic scheduling. Although the cooperative control theory of MASs has undergone remarkable developments in the past decade, the existing results do not simultaneously satisfy the two basic requirements of scalability and high efficiency yet, thus imposing great limitations in practical applications. In view of this disadvantage, the problem of adaptive event-triggered cooperative control of linear MASs will be investigated in this project, aiming at proposing adaptive event-triggered control protocols which simultaneously satisfy the requirements of scalability and high efficiency. Firstly, we study the event-triggered consensus problem of linear multi-agent systems, in order to obtain event-triggered control protocols with less conservatism. Based on this development, we will investigate the problem of adaptive event-triggered consensus control of linear MASs and propose state-feedback based and output-feedback based adaptive event-triggered control protocols, respectively, as well as the corresponding design method. In addition, we will further solve the problem of adaptive event-triggered cooperative output regulation of heterogeneous linear MASs. Finally, the proposed results on cooperative control theory will be verified on the multi-robot cooperation platform. The goal of this paper is to develop systematic methods for designing adaptive event-triggered cooperative control protocols for linear MASs, improving and completing the existing cooperative control theory of MASs.

多智能体系统具有很强的工程背景，在多机器人协作、智能电网、物流调度等领域具有广泛的应用。尽管多智能体系统的协同控制理论在最近十年里得到了长足的发展，但是现有结果无法同时满足可扩展性和高效用的设计需求，因而在应用中具有很大的局限性。为此，本项目对线性多智能体系统的自适应事件触发协同控制问题展开研究，提出同时满足可扩展性和高效用要求的自适应事件触发控制协议。首先研究线性多智能体系统的事件触发一致性控制问题，获得具有更低保守性的事件触发控制协议；在此基础上，研究自适应事件触发一致性控制问题，分别提出基于状态反馈和基于输出反馈的控制协议及其设计方法；此外，进一步解决异质线性多智能体系统的自适应事件触发协同输出调节问题；最后，在多机器人协作系统上验证取得的协同控制理论研究成果。本项目的宗旨是提出一套较为完整的线性多智能体系统的自适应事件触发协同控制设计方法，完善现有多智能体系统协同控制理论。

多智能体系统具有很强的工程背景，在多机器人协作、智能电网、物流调度等领域具有广泛的应用。尽管多智能体系统的协同控制理论在最近十年里得到了长足的发展，但是现有结果无法同时满足可扩展性和高效用的设计需求，因而在应用中具有很大的局限性。本项目系统深入研究了线性多智能体系统的自适应事件触发协同控制问题，提出了同时满足可扩展性和高效用要求的自适应事件触发控制协议。首先研究了线性多智能体系统的自适应一致性控制问题，给出了新颖的降阶自适应一致性控制协议。较之已有结果，新协议同时具有更低的阶次且满足完全分布式的要求。在此基础上，进一步提出了基于状态反馈和基于输出反馈的自适应事件触发一致性控制协议，构建了无需智能体绝对信息的自适应事件触发条件。由于仅依赖于智能体间的相对信息，提出的控制协议具有更广泛的适用性。此外，设计了智能体相对信息的迭代解算方法，解决了为获取相对信息而需要对智能体进行持续监测的实现难题。最后，研究了多欧拉-拉格朗日系统的自适应事件触发一致性控制方法，为将来进一步开展多机器人系统的自适应事件触发协同的研究和验证打下了基础。