Multi-robot, multi-target Particle Swarm Optimization search in noisy wireless environments

Abstract

Multiple small robots (swarms) can work together using Particle Swarm Optimization (PSO) to perform tasks that are difficult or impossible for a single robot to accomplish. The problem considered in this paper is exploration of an unknown environment with the goal of finding a target(s) at an unknown location(s) using multiple small mobile robots. This work demonstrates the use of a distributed PSO algorithm with a novel adaptive RSS weighting factor to guide robots for locating target(s) in high risk environments. The approach was developed and analyzed on multiple robot single and multiple target search. The approach was further enhanced by the multi-robot-multi-target search in noisy environments. The experimental results demonstrated how the availability of radio frequency signal can significantly affect robot search time to reach a target.

Distributed target identification in robotic swarms

“…In multi-robot systems, many authors ( [5], [7]) have dealt with the problem of jointly estimating the location of a common target by fusing the individual readings of the robots. However, each robot still enjoys a wide point of view over the scene.…”

Properties of classes, subclasses and objects in an abstraction model

“…Another direction of future research will be focused on the integration of the results. Several other models, applications and structures will be used [20]- [32].…”

On-line task allocation for the robotic interception of multiple targets in dynamic settings

“…Several off-line methodologies for the MTSP in multitarget interception by multiple pursuers have been reported in the literature [11][12][13][14][15][16][17][18][19]. These methods perform pursuertarget assignments only once, at the start of travel, or when a pursuer has detected a nearby target.…”

“…These methods perform pursuertarget assignments only once, at the start of travel, or when a pursuer has detected a nearby target. While this one-time determination of pursuer-target pairings can be applicable to scenarios in which targets are static [13,16,19] or where the targets' trajectories can be defined a priori [18], they would not be suitable for targets with unknown trajectories and high maneuverability as the optimality of the initial pairings would not hold. In [20], an algorithm was presented for online re-assignment of unmanned aerial vehicles containing and intercepting evading targets using a differential games theory framework.…”