TY - GEN
T1 - Game theoretic optimal position computation of collaborating agents for visual area coverage∗
AU - Papatheodorou, Sotiris
AU - Hamidou, Tembine
AU - Smyrnakis, Michalis
AU - Tzes, Anthony
N1 - Funding Information:
∗This research work was partially supported by the U.S. Air Force Office of Scientific Research under grant number FA9550-17-1-0259 and by the European Union Horizon 2020 Research and Innovation Programme under the Grant Agreement No. 644128, AEROWORKS.
Publisher Copyright:
© 2018 Copyright held by the owner/author(s).
PY - 2018/7/9
Y1 - 2018/7/9
N2 - The computation problem of the optimal position of collaborating mobile robots equipped with omnidirectional cameras for visual area coverage is the subject of this paper. The planar area has several stationary obstacles resulting in a computationally intractable search scheme. Rather than using gradient-based search methods, the multi-agent swarm is clustered to partially deal with the dimensionality curse. Each cluster is end-to-end connected and its area of responsibility is assigned based on its collective Voronoi tessellation. This area is then coarsely sampled and a game-theoretic approach is employed relying on fictitious play amongst the cluster’s members. The search scheme is then switched into a fine-spatial sampling and initialized using the previously attained coarse optimal positions of the agents. The provided adaptive-size game-theoretic optimization search approach provides the optimal location of the agents with a tenfold faster convergence compared to the gradient-search methods. Simulation studies are offered to highlight the efficiency of the search scheme.
AB - The computation problem of the optimal position of collaborating mobile robots equipped with omnidirectional cameras for visual area coverage is the subject of this paper. The planar area has several stationary obstacles resulting in a computationally intractable search scheme. Rather than using gradient-based search methods, the multi-agent swarm is clustered to partially deal with the dimensionality curse. Each cluster is end-to-end connected and its area of responsibility is assigned based on its collective Voronoi tessellation. This area is then coarsely sampled and a game-theoretic approach is employed relying on fictitious play amongst the cluster’s members. The search scheme is then switched into a fine-spatial sampling and initialized using the previously attained coarse optimal positions of the agents. The provided adaptive-size game-theoretic optimization search approach provides the optimal location of the agents with a tenfold faster convergence compared to the gradient-search methods. Simulation studies are offered to highlight the efficiency of the search scheme.
KW - Area coverage
KW - Collaborative control
KW - Game theory
KW - Multi–agent optimization
UR - http://www.scopus.com/inward/record.url?scp=85052018456&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85052018456&partnerID=8YFLogxK
U2 - 10.1145/3200947.3201015
DO - 10.1145/3200947.3201015
M3 - Conference contribution
AN - SCOPUS:85052018456
T3 - ACM International Conference Proceeding Series
BT - Proceedings - 10th Hellenic Conference on Artificial Intelligence, SETN 2018
PB - Association for Computing Machinery
T2 - 10th Hellenic Conference on Artificial Intelligence, SETN 2018
Y2 - 9 July 2018 through 12 July 2018
ER -