Design And Analysis Of Consensus Protocols For Multi-agent Systems By Using Learning Control

Coordination control and cooperation of multi-agent systems can improve the intelligent level of individual behavior, and can complete a lot of tasks which cannot be complete by an individual. Therefore, the coordination and cooperative control for multi-agent systems have attracted a lot of attention from various research communities. Most of the available literatures have been focused on the consensus of multi-agent systems in the one-dimensional system framework evolving along the time axis. In order to achieve the objective of precise consensus for multi-agent systems, iterative learning control has been employed for dealing with the consensus problems. At present, the consensus of multi-agent systems by using iterative learning control is still in the preliminary study stage. This paper mainly makes full use of the characteristics of distributed multi-agent systems and learning control, and studies the consensus problem of multi-agent systems in the repeatable control environment. The distributed learning protocols are presented to ensure that all follower agents can achieve perfect tracking consensus. The proposed scheme in this paper can realize the effective combination of learning control theory and distributed multi-agent system. The main works of this paper are as follows:1. Leader-following consensus problems for linear multi-agent systems are studied by using iterative learning control approach. Assuming that the leader node is globally reachable in the multi-agent systems, the iterative learning control protocols are designed for single-integrator and the general linear multi-agent dynamics, respectively. Then, some sufficient conditions to guarantee the multi-agent consensus are derived for the directed communication topologies. Consequently, all follower agents can track the leader perfectly in the finite time interval [0,T].2. The consensus problem for the first-order parametric nonlinear leader-following multi-agent systems was studied by utilizing adaptive iterative learning control. Combined with the characteristics of adaptive iterative learning control and the distributed system, a new adaptive iterative learning control protocol is designed. The proposed protocol guarantees that each follower agent can track the leader on [0, T] perfectly. Then, an efficient framework is proposed for the consensus and formation control of leader-following multi-agent systems with second-order dynamics and unknown time-varying parameters. By means of an adaptive iterative learning control approach, the distributed adaptive iterative learning protocols are designed such that all follower agents track the leader uniformly on [0,T]. In addition, the proposed protocol can solve the formation control of multi-agent systems effectively.3. We propose a distributed adaptive fuzzy iterative learning control algorithm to deal with coordination control problems for first-order leader-following multi-agent systems. A fuzzy logical system is used to approximate the nonlinearity of each follower agent. In order to avoid the use of initial reset condition, distributed initial-state learning protocols are designed and each follower agent can take arbitrary initial state. Combined time-domain and iteration-domain adaptive laws are used to tune the controller parameters. The protocol guarantees that all follower agents track the leader for the consensus problem and keep at a desired distance from the leader for the formation problem on [0, T] when the communication topology is connected. In addition, the adaptive fuzzy iterative learning control scheme is proposed for coordination problems of M-th order (M≥2) distributed multi-agent systems. With distributed initial state learning, the unified distributed protocols combined time-domain and iteration-domain adaptive laws guarantee that the follower agents track the leader uniformly on [0, T]. Then, the proposed algorithm extends to achieve the formation control.4. The coordination problems are studied for a kind of leader-follower heterogeneous multi-agent systems on [0, T] by applying iterative learning control scheme. The heteroge-neous multi-agent systems are composed of first-order and second-order dynamics in two aspects. The leader is assumed to have second-order dynamics and the trajectories of the leader are only accessible to a subset of the followers. To overcome the strict identical initial condition commonly used in ILC, the distributed initial state learning controller for each follower is designed, thus each follower agent can take arbitrary initial state. Distributed itera-tive learning protocols guarantee all follower agents to achieve perfect tracking consensus for both fixed and switching connected communication topologies, respectively. In addition, the proposed scheme is also extended to achieve formation control for heterogeneous multi-agent system.5. We propose adaptive repetitive control framework for uncertain nonlinear multi-agent systems. Based on the framework, by learning periodic uncertainties, consensus-based learning control protocols are designed for nonlinear multi-agent systems with time-varying parametric uncertainty. The learning-based updating law is utilized to compensate for periodic time-vary ing parametric uncertainties. With the dynamic of the leader unknown to any follower agents, a new auxiliary control is designed for each follower agent to deal with the leader’s dynamic. Then, the proposed learning control protocol guarantees that all follower agents can track the leader. Furthermore, as an extension of the consensus problem, the formation problem is studied. Finally, simulation examples are given to illustrate the effectiveness of the proposed method in this article.