Decentralized frontier occupancy assignment by using multiple exploration robots

U radu je implementiran algoritam decentralizirane raspodjele toˇcaka fronte u svrhu istraživanja nepoznatog prostora višerobotskim sustavom, temeljen na market modelu, uz korištenje robota Pioneer P3-DX. Istovremenu lokalizaciju i izgradnju karte obavlja Google Cartographer. Algoritam raspodjele toˇcaka ukljuˇcuje filtriranje (grupiranje) toˇcaka fronte, koje detektira Google Cartographer, pomo´cu aglomerativnog hijerarhijskog algoritma za grupiranje. Raspodjela zadataka vrši se rješavanjem problema trgovacˇkog putnika i mad¯arskim algoritmom. Algoritam je odabran tako da minimizira vrijeme potrebno za istraživanje prostora. Eksperimenti su provedeni u virtualnom okruženju gdje se uspored¯uju dva algoritma raspodjele toˇcaka fronte.This thesis considers decentralized frontier points occupancy assignment by using multiple exploration robots. The robots negotiate amongst themselves to minimize the exploration time. The exploration algorithm of an unknown environment includes clustering (filtering) of frontier points using hierarchical clustering and assigning missions to robots using an algorithm that solves the traveling salesman problem and that uses the Hungarian algorithm. In order to compare algorithms, experiments were conducted in a simulation environment