Adaptive neural network consensus tracking control for uncertain multi-agent systems with predefined accuracy
- PDF / 1,149,391 Bytes
- 14 Pages / 547.087 x 737.008 pts Page_size
- 52 Downloads / 210 Views
ORIGINAL PAPER
Adaptive neural network consensus tracking control for uncertain multi-agent systems with predefined accuracy Dajie Yao · Chunxia Dou · Dong Yue · Nan Zhao · Tingjun Zhang
Received: 24 April 2020 / Accepted: 4 August 2020 © Springer Nature B.V. 2020
Abstract This paper proposes the consensus tracking control problem for a class of uncertain nonlinear multi-agent systems. By using a group of nonnegative functions, an adaptive neural network controller is addressed based on the technique of backstepping. Compared with existing results about uncertain nonlinear multi-agent systems, the advantage of the proposed scheme is that it can ensure the consensus of multiagent systems within a given accuracy by using two nth-order continuous differentiable functions. Finally, simulation results confirm the correctness of the proposed scheme. Keywords Nonlinear multi-agent systems · Adaptive neural network control · Consensus tracking control · Predefined accuracy
D. Yao (B)· D. Yue · N. Zhao · T. Zhang College of Automation and College of Artificial Intelligence, Nanjing University of Posts and Telecommunications, Nanjing 210023, China e-mail: [email protected] D. Yao School of Mechanical and Electrical Engineering, Chizhou University, Chizhou 247000, China C. Dou · D. Yue · N. Zhao · T. Zhang Institute of Advanced Technology, Nanjing University of Posts and Telecommunications, Nanjing 210023, China
1 Introduction In recent decades, the research of multi-agent systems (MASs) has received great attention because of widely applied in a variety of areas such as sensor networks, satellites, mobile robots, unmanned vehicles and autonomous underwater vehicles [1–6]. Among the many properties of multi-agents, it is worth emphasizing that the consensus problem is an important and fundamental feature for MASs, it requires that all agents converge to a same value. Especially, if all agents converge to the state of a certain agent, that is, the leader-following problem, see, e.g., [7–10]. With that in mind, many scholars have studied this problem from different perspectives based on the knowledge of graph theory and a number of significant results have emerged, for instance the consensus of second-order MASs [11–15], the finite-time consensus for MASs [16–19], the fixed-time consensus for MASs [9,20–22], even in terms of saving resources for event-triggered MASs [10,23–26]. In the existing research results, the research objects have been linear dynamic models and nonlinear dynamic models, and there have been significant differences in the methods of studying two class of models. Generally speaking, nonlinear properties are closer to the essence of things, and uncertainty exists widely in engineering applications. Hence, in theoretical modeling, some unknown nonlinear functions are introduced to indicate that the nonlinear dynamic mod-
123
D. Yao et al.
els are unknown. To name a few, unknown functions Fi (x i ) (i = 1, 2, . . . , n) for uncertain multiinput and multi-output (MIMO) nonlinear systems σ (t) were add
Data Loading...
