Experimental Implementation of Distributed Average Tracking for Heterogeneous Physical Agents Using Neighbors’ Positions

The focus of this thesis is on average tracking algorithm implementation for a group of heterogeneous physical agents consisting of single-integrator, double-integrator and Euler-Lagrange dynamics. In the algorithms, each agent is able to track the average of the time-varying reference inputs, where each agent has access to only its own position, its own input signals and the relative positions between itself and its neighbors. The algorithms are experimentally implemented on a multi-robot platform under an undirected communication topology. Simulation results and the experimental results based on the multi-robot platform are shown to validate the proposed algorithms. Finally, the error in simulation and experiment is analyzed based on experimental environment and robot characteristics.