Lennard-jones potential field-based swarm systems for aggregation and obstacle avoidance

Abstract

Main concern of this paper is constructing a swarm model using Lennard-Jones potential for N agents. The Lennard-Jones potential [1] is one of basic models to interpret motion of atoms or molecules in molecular dynamics. The swarm model generates interaction forces composed of attractive force and repulsive force as a function of initial positions of every agent on the basis of Lennard-Jones potential model with velocity matching. In that case, motion of the agents in the swarm shows aggregation behavior. Also, one of main issues of swarm system is how to avoid obstacles. In our case, we apply additional repulsive potential function into our swarm system for obstacle avoidance. After conducting a simulation, we confirm that a group of agents appears swarm behavior. We expect that this behavior can be applied to control movement of multiple robots or UAVs.