Distributed Robust Consensus Tracking Control of Higher-order
Nonlinear Systems
Gang Wang
1
, Chaoli Wang
2
, Qinghui Du
1
1. Business School, University of Shanghai for Science and Technology, Shanghai, 200093
E-mail: 2010wanggang@gmail.com
2. Department of Control Science and Engineering, University of Shanghai for Science and Technology, Shanghai, 200093
E-mail: clclwang@126.com
Abstract: In this paper, a multi-agent consensus problem has been considered with a dynamic leader and fixed commu-
nication topology. Each follower node is modeled by a higher-order nonlinear systems with unknown nonlinear dynamics
and an unknown disturbance. The leader node is also a higher-order nonlinear system. Only a part of the networked group
has access to the information of the leader. And a distributed robust consensus controller is designed to ensure all the
follower nodes synchronize to the trajectory of the leader node exponentially. Meanwhile, a smooth and time-varying
sliding mode controller is presented to eliminate chattering phenomenon. The stability of the proposed methods is proved
rigorously. Simulation results confirm the effectiveness of the proposed methods.
Key Words: Multi-agent, Consensus, Nonlinear, Robust, Synchronization
1 INTRODUCTION
In recent years, there has been an increasing research inter-
est in the distributed synchronization control of multi-agent
systems due to its potential applications in sensor network-
s, monitoring and surveillance, unmanned-air-vehicle for-
mations, etc. A lot of effective control methods have been
presented to deal with the synchronization problem in ear-
ly work, and interested readers are referred to the survey
papers [1].
Consensus problems have been studied from several dif-
ferent perspectives in the existing literature. For exam-
ple, consensus for systems on communication graphs with
fixed or time-varying has been researched [2]. And some
scholars have focused on leaderless consensus and leader-
following consensus. For the leaderless consensus, all the
agents are driven to a common value [1]. As for the leader-
following consensus, which is also called the cooperative
tracking problem, there is a leader node which acts as time-
varying signals. And only a part of the follower nodes has
access to the information of the leader, all the followers can
synchronize to the trajectory of the leader finally. There
are also many works that investigated for different types
of agent dynamics including first-order integrator systems
[3, 4], second-order integrator systems [5, 6] and higher-
order integrator systems [7, 8, 9, 10]. Hou et al. [3] pre-
sented a robust adaptive control method to achieve consen-
sus for the first-order multi-agent systems with the uncer-
tainties and external disturbances. In [4, 6], Das and Lewis
have solved the cooperative tracking problem for first-order
This paper was partially supported by Scientific Innovation pro-
gram(13ZZ115), The National Natural Science Foundation (61374040),
Hujiang Foundation of China (C14002), Graduate Innovation program of
Shanghai(54-13-302-102).
nonlinear systems and second-order nonlinear systems with
non-identical unknown nonlinear dynamics and unknown
disturbances. A distributed adaptive controller is designed
to guarantee the uniform ultimate boundedness of the track-
ing errors. And the results are generalized to higher-order
nonlinear systems in the Brunovsky form by Zhang and
Lewis in [7]. In [8], Wang, Zhang, and Guo have stud-
ied the adaptive output consensus tracking of a class of
higher-order systems with mismatched unknown parame-
ters. And the estimator is used for every agent to estimate
the consensus reference and exchanged among the neigh-
bors. Thus, the communication burden may be increased.
In addition, whether the sufficient condition to ensure the
output consensus tracking has been satisfied is not easy to
check. Motivated by those observations, Wang et al. pro-
posed a backstepping based distributed adaptive controller
to achieve the output consensus tracking with the multi-
agent systems as similar to [8] in [9]. However, a clear
drawback of the backstepping technique is that the recur-
sive design procedure is more complicated with increasing
the order of the systems.
In this paper, we study the consensus tracking control of
higher-order nonlinear systems as similar to Zhang et al.
in [7]. Different with [7], our methods can let the tracking
errors converge to zero exponentially or asymptotically by
adding an extra condition that every agent must know its
neighbor’s control input. Also due to the difference of the
Lyapunov function, Lemma 1 in [7] which plays a key role
is not needed here. In addition, the use of the Frobenius
norm in the Lyapunov function which is instrumental in [7]
is not effective any more. The main contributions of this
paper are twofold. Firstly, the distributed robust consensus
control approach is designed to make the tracking errors
go to zero exponentially. Secondly, to eliminate drastically
2096
978-1-4799-7016-2/15/$31.00
c
2015 IEEE