Title :
Application of sliding mode control to swarms under conflict
Author :
Cepeda-Gomez, R. ; Olgac, Nejat ; Sierra, D.A.
Author_Institution :
Dept. of Mech. Eng., Univ. of Connecticut, Storrs, CT, USA
fDate :
12/30/1899 12:00:00 AM
Abstract :
A robustising sliding mode control strategy is implemented on two competing multi-agent swarms, called pursuers and evaders, with equal membership count. Newtonian dynamic models are considered, which include drag forces as well as the inter-agent attraction/repulsion forces. The proposed sliding mode control law achieves the stability and the capture of the evaders by the pursuers despite the uncertainties in the evader behaviour. The group pursuit is conceived in two phases: the approach phase during which the two swarms act like two individuals and the assigned pursuit phase when each pursuer is assigned to an evader. Furthermore, the authors take into account a turning action for the evaders, which adds to their agility. This property is considered as a part of the uncertainty in the dynamics. The control parameters are separately studied to assess their influences on the pursuit.
Keywords :
Newton method; multi-agent systems; multi-robot systems; robust control; stability; variable structure systems; Newtonian dynamic model; inter agent attraction repulsion forces; multi-agent swarms; robustising sliding mode control strategy; stability;
Journal_Title :
Control Theory & Applications, IET
DOI :
10.1049/iet-cta.2010.0201