非线性多自主体协同输出调节问题及应用研究

The distributed output regulation approach has been recognized as a powerful control scheme for cooperative control of multi-agent systems. The salient feature of this approach lies in that it can handle heterogeneous dynamics, system uncertainties and external disturbances at the same time. The general problem formulation of cooperative output regulation includes many cooperative control problems of uncertain multi-agent systems such as consensus, synchronization, containment control, and formation as its special cases. However, in most of existing literature, the solvability of cooperative output regulation problem relies on the assumptions that the exosystem is certain, no communication delay exists, and network topologies are time invariant. In real applications, the controlled plants and communication conditions are much more complicated. For instance, the exosystem may contain uncertain parameter, the communication delay is indispensable and the communication topologies may be time-varying. All these settings, in spite of making the problem more practical, raise great technical challenges. Therefore, this project aims at developing new distributed control schemes to solve several critical and challenging issues in cooperative output regulation. ..First, novel distributed control schemes based on internal model approach will be developed to handle the cooperative robust output regulation problem of several classes of nonlinear uncertain multi-agent systems subject to unknown exosystem. ..Second, by integrating distributed feedforward approach and distributed internal model approach, a hierarchical control scheme will be adopted to construct distributed control laws to solve the cooperative robust output regulation problem of several classes of nonlinear uncertain multi-agent systems under arbitrarily bounded time-varying communication delays. ..Third, a general framework for cooperative robust output regulation of nonlinear uncertain multi-agent systems will be established, which unifies the average dwell time approach, the distributed feedforward approach, and the distributed internal model approach. Based on this framework, a general distributed switched control law can be developed such that the cooperative output regulation problem under unknown exosystem, communication delays and jointly connected switching network topologies can be handled at the same time...Finally, the developed distributed control schemes will be applied to the coordination control of multiple mobile robots. The success of this research project will enable us to answer a number of fundamental and critical problems in cooperative control of multi-agent systems and its implementation. The expected results from this project will provide more powerful tools with fewer assumptions that are applicable to more practical situations, and thus considerably narrow the gaps between theories and applications.

非线性协同输出调节问题包括许多协同控制问题,近年来受到广泛关注。然而，现有文献中此问题的可解性大都基于一些共同的假设：外部系统已知、不存在通讯时延并且网络拓扑是时不变的，往往难以满足实际工程应用的要求。为此，本项目首先基于分布式内模方法提出新颖的控制策略，求解在外部系统含未知参数条件下的鲁棒协同输出调节问题。第二，结合分布式前馈方法和分布式内模方法提出分层式控制策略，求解在系统含不确定参数并且受到任意有界时变网络通讯时延影响条件下的鲁棒协同输出调节问题。第三，建立一个解决非线性多自主体鲁棒协同输出调节问题的广义理论框架，使平均驻留时间方法、分布式前馈方法和内模方法得到有机统一，进而提出一般性的分布式切换控制律。最后，将以上提出的理论应用于机器人多自主体的协同运动控制。本项目将解决协同控制理论及应用中的一些基本问题，突破多自主体理论研究中的一些关键技术瓶颈，拓广该理论的工程应用范围。

非线性协同输出调节问题包括许多协同控制问题，近年来受到广泛关注。然而，现有文献中此问题的可解性大都基于一些共同的假设：外部系统已知、不存在通讯时延且网络拓扑是时不变的。而实际工程中上述假设常常无法满足。随着上述假设的逐一放宽，多自主体协同控制方法的应用场景将被极大地拓展。同时，这也带来了诸多新的控制方法设计方面的技术挑战。为此，本项目首先基于分布式内模方法提出新颖的控制策略，求解在外部系统含未知参数条件下的鲁棒协同输出调节问题。第二，建立了一个解决多自主体鲁棒协同输出调节问题的分层式控制理论框架，将复杂条件下的非线性异质多自主体协同输出调节问题解耦成外部系统含未知参数的网络一致性控制问题和各子系统的输出调节问题。采用该理论方法，成功地解决了线性异质多自主体系统，欧拉-拉格朗日多自主体系统以及二阶非线性多自主体系统在外部系统含未知参数、存在通讯时延、网络拓扑时变等各种复杂条件下的鲁棒输出调节问题。最后，以上提出的理论被应用于几类常见多自主体系统的协同控制中，包含多机器人包围目标编队控制问题，多航天器姿态协同控制问题，无人机编队控制问题等。本项目解决了复杂条件下协同控制理论及其应用中的一些基本问题，突破了多自主体理论研究中的一些关键技术瓶颈，极大地拓展了该理论的工程应用范围。