Jim Perkins

A Decentralized Searching Control Algorithm for a Scalable Robotic Swarm based on Mesh-free Particle Hydrodynamics

M.S. Thesis Defense
Date: Tuesday August 16, 2005

Department of Electrical and Computer Engineering

Abstract
This thesis seeks to develop an innovative positional organization and movement algorithm for large-scale distributed robotic systems. This control system is based on the algorithm, “Smoothed Particle Hydrodynamics” which is commonly used for simulating fluids, solids, and astrophysical phenomena. This algorithm can be adapted for a variety of different uses and in this thesis will illustrate the changing of the swarms “pose” with regards to object detection in an unknown area. This particular use can help mitigate “false alarms” in a mobile sensor net. This algorithm must be computationally simple, fault tolerant, and rely solely on local interactions. Feasibility has been verified by simulating this control algorithm using a group of mobile agents in 2 dimensional space.

These results show that the algorithm works well for adapting the swarms pose when encountering detectable objects and obstacles.

Thesis Committee:
Bahram Shafai (Thesis Advisor)
Stefano Basagni
Mehmet Dokmeci