J Intell Robot Syst (2010) 58:165–189
DOI 10.1007/s10846-009-9361-7
Multi-Agent Formation Control Based
on Bell-Shaped Potential Functions
Kristian Hengster-Movri´ c · Stjepan Bogdan ·
Ivica Draganjac
Received: 17 July 2009 / Accepted: 17 August 2009 / Published online: 29 August 2009
© Springer Science + Business Media B.V. 2009
Abstract In this paper we analyze stability properties of multi-agent control system
with an artificial potential based on bell-shaped functions. In our approach attractive
and repulsive forces created by potential gradient have the same form. This particular
property allows definition of target formation that is parameter invariant. Due to
the fact that agents are identical, the proposed structure of formation potential is
invariant to the interchange of agents configurations, hence, target in which particu-
lar agent would eventually end up, depends only on formation initial condition.
It has been demonstrated that stability analysis, given for stationary targets, applies
to moving targets formation as well. We show that position of unwanted stable equi-
libria can be controlled by a single parameter that defines an elementary potential
function. This fact has been used for synthesis of an adaptation algorithm, such that
arrival of agents at required formation is guarantied. Simulation results, presented at
the end of the paper, confirm correctness of the proposed control scheme.
Keywords Multi-agent systems · Formation control · Potential navigation function
1 Introduction
In the last decade multi-agent systems have attracted a lot of attention among
researchers. As unmanned vehicles are becoming more capable and more afford-
able, their potential for employment in a wide range of applications is increasing.
Effective deployment of autonomous mobile teams made up from numerous units
calls for sophisticated communication, cooperation and control mechanisms. Use of
cooperative formation of agents enables achieving goals that would not be possible
K. Hengster-Movri´ c · S. Bogdan (B ) · I. Draganjac
LARICS—Laboratory for Robotics and Intelligent Control Systems,
Department of Control and Computer Engineering, Faculty of EE&C,
University of Zagreb, Unska 3, 10000 Zagreb, Croatia
e-mail: stjepan.bogdan@fer.hr