基于事件驱动的多智能体系统一致性研究
发布时间：2019-04-02 14:01
【摘要】：近年来,关于多智能体系统的协调控制问题越来越受到工程技术领域和科学研究领域的关注。目前,已在无人机编队、交通控制等领域得到应用。多智能体系统控制问题的基本研究包括一致性问题、编队控制、会和控制和聚结控制,后三个为一致性问题的推广。在实际应用中,内嵌微处理器的多智能体系统能源有限,所以要求系统尽可能地减少能源消耗,保障系统正常运转。多智能体系统传统的控制方式多为时间驱动或是周期采样,存在通信带宽和计算资源浪费的弊端,有较大的能源消耗。采用事件驱动的控制方式,可以有效地减少控制器的更新次数,缓解每个个体能量、资源受限的问题。本文重点研究了基于事件驱动的多智能体系统一致性问题,包括分组一致性和跟踪一致性两方面内容。基于事件驱动的分组一致性问题,针对无向拓扑下一阶和二阶多智能体系统,采用时间相关的分散式事件驱动函数,设计了事件驱动分组一致性协议,依靠矩阵论和线性系统理论知识,得出了系统实现分组一致性需要满足的代数条件;然后又将结果拓展到了带时延的情况,通过构造李雅谱诺夫-克拉索夫斯基(Lyapunov-Krasovskii)函数,以线性矩阵不等式(Linear Matrix Inequality)的形式展现了保证系统实现分组一致性的充分条件。同时,还证明了系统不会发生奇诺现象。基于事件驱动的二阶多智能体系统的跟踪一致性问题,从离散系统的角度,针对固定有向拓扑下的二阶领导-跟随多智能体系统,采用时间相关的事件驱动函数,设计了相应的事件驱动跟踪控制协议和领航者速度观测器,通过代数图论和矩阵论等工具,给出了控制参数和采样周期的选择办法,得到了使系统实现跟踪一致性的代数条件;然后又对切换拓扑的情形做了详细的分析,通过构造李雅谱诺夫(Lyapunov)函数,以线性矩阵不等式的形式给出了保证系统实现跟踪一致性的充分条件。通过数值仿真,证实了控制协议的合理性和驱动函数的有效性。
[Abstract]:In recent years, more and more attention has been paid to the coordinated control of multi-agent systems in the field of engineering technology and scientific research. At present, it has been applied in UAV formation, traffic control and other fields. The basic research of multi-agent system control includes consistency problem, formation control, union control and coalescence control. The last three are the generalization of the consistency problem. In practical application, the energy of multi-agent system with embedded microprocessor is limited, so the system is required to reduce the energy consumption as far as possible and ensure the normal operation of the system. The traditional control mode of multi-agent system is time-driven or periodic sampling, which has the disadvantages of communication bandwidth and the waste of computing resources, and has a large energy consumption. Event-driven control can effectively reduce the number of updates of the controller and alleviate the energy and resource constraints of each individual. This paper focuses on the unifying problem of event-driven multi-intelligence architecture, including packet consistency and tracking consistency. Based on event-driven packet consistency problem, an event-driven packet consistency protocol is designed for undirected topology-based first-order and second-order multi-agent systems using time-dependent decentralized event-driven functions. Based on the knowledge of matrix theory and linear system theory, the algebraic conditions that need to be satisfied to realize the group consistency of the system are obtained. Then, the result is extended to the case with delay. By constructing the Lyapunov-Krasovskii function, the sufficient conditions to ensure the packet consistency of the system are presented in the form of linear matrix inequality (Linear Matrix Inequality). At the same time, it is proved that the Chino phenomenon will not occur in the system. Based on the problem of tracking consistency of event-driven second-order multi-agent systems, the time-dependent event-driven function is used for the second-order leader-follower multi-agent systems in fixed directed topology from the point of view of discrete system. The corresponding event-driven tracking control protocol and navigator velocity observer are designed. By means of algebraic graph theory and matrix theory, the selection method of control parameters and sampling period is given, and the algebraic conditions to make the system realize tracking consistency are obtained. Then the switching topology is analyzed in detail. By constructing Lyapunov (Lyapunov) function, the sufficient conditions to ensure the tracking consistency of the system are given in the form of linear matrix inequality (LMI). The rationality of the control protocol and the validity of the driving function are verified by numerical simulation.
【学位授予单位】：华北电力大学(北京)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18
本文编号：2452620
[Abstract]:In recent years, more and more attention has been paid to the coordinated control of multi-agent systems in the field of engineering technology and scientific research. At present, it has been applied in UAV formation, traffic control and other fields. The basic research of multi-agent system control includes consistency problem, formation control, union control and coalescence control. The last three are the generalization of the consistency problem. In practical application, the energy of multi-agent system with embedded microprocessor is limited, so the system is required to reduce the energy consumption as far as possible and ensure the normal operation of the system. The traditional control mode of multi-agent system is time-driven or periodic sampling, which has the disadvantages of communication bandwidth and the waste of computing resources, and has a large energy consumption. Event-driven control can effectively reduce the number of updates of the controller and alleviate the energy and resource constraints of each individual. This paper focuses on the unifying problem of event-driven multi-intelligence architecture, including packet consistency and tracking consistency. Based on event-driven packet consistency problem, an event-driven packet consistency protocol is designed for undirected topology-based first-order and second-order multi-agent systems using time-dependent decentralized event-driven functions. Based on the knowledge of matrix theory and linear system theory, the algebraic conditions that need to be satisfied to realize the group consistency of the system are obtained. Then, the result is extended to the case with delay. By constructing the Lyapunov-Krasovskii function, the sufficient conditions to ensure the packet consistency of the system are presented in the form of linear matrix inequality (Linear Matrix Inequality). At the same time, it is proved that the Chino phenomenon will not occur in the system. Based on the problem of tracking consistency of event-driven second-order multi-agent systems, the time-dependent event-driven function is used for the second-order leader-follower multi-agent systems in fixed directed topology from the point of view of discrete system. The corresponding event-driven tracking control protocol and navigator velocity observer are designed. By means of algebraic graph theory and matrix theory, the selection method of control parameters and sampling period is given, and the algebraic conditions to make the system realize tracking consistency are obtained. Then the switching topology is analyzed in detail. By constructing Lyapunov (Lyapunov) function, the sufficient conditions to ensure the tracking consistency of the system are given in the form of linear matrix inequality (LMI). The rationality of the control protocol and the validity of the driving function are verified by numerical simulation.
【学位授予单位】：华北电力大学(北京)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18
【参考文献】
相关期刊论文 前4条
1 关永强;纪志坚;张霖;王龙;;多智能体系统能控性研究进展[J];控制理论与应用;2015年04期
2 朱亚锟;关新平;罗小元;;线性和非线性动态异构多自主体系统的有限时间一致性（英文）[J];自动化学报;2014年11期
3 闵海波;刘源;王仕成;孙富春;;多个体协调控制问题综述[J];自动化学报;2012年10期
4 王震宇,高更君,黄卫;基于多智能主体的高速公路控制系统结构[J];交通运输工程学报;2003年01期
相关博士学位论文 前2条
1 卫佳;网络化多智能体系统的一致性研究[D];华中科技大学;2013年
2 胡松林;基于事件触发机制的网络化控制系统的分析与综合[D];华中科技大学;2012年
相关硕士学位论文 前2条
1 高澜;分布式事件触发技术在多智能体一致性控制中的应用研究[D];重庆大学;2015年
2 傅梓瑛;多智能体理论及其在厚度控制系统中的应用研究[D];东北大学;2012年
，本文编号：2452620
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2452620.html
