TY - GEN
T1 - BN+BN
T2 - 16th Australian Conference on Artificial Intelligence, AI 2003
AU - Kim, Kyung Joong
AU - Cho, Sung Bae
N1 - Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 2003.
Copyright:
Copyright 2020 Elsevier B.V., All rights reserved.
PY - 2003
Y1 - 2003
N2 - In the philosophy of behavior-based robotics, design of complex behavior needs the interaction of basic behaviors that are easily implemented. Action selection mechanism selects the most appropriate behavior among them to achieve goals of robot. Usually, robot might have one or more goals that conflict each other and needs a mechanism to coordinate them. Bayesian network represents the dependencies among variables with directed acyclic graph and infers posterior probability using prior knowledge. This paper proposes a method to improve behavior network, action selection mechanism that uses the graph of behaviors, goals and sensors with activation spreading, using goal inference mechanism of Bayesian network learned automatically. Experimental results on Khepera mobile robot show that the proposed method can generate more appropriate behaviors.
AB - In the philosophy of behavior-based robotics, design of complex behavior needs the interaction of basic behaviors that are easily implemented. Action selection mechanism selects the most appropriate behavior among them to achieve goals of robot. Usually, robot might have one or more goals that conflict each other and needs a mechanism to coordinate them. Bayesian network represents the dependencies among variables with directed acyclic graph and infers posterior probability using prior knowledge. This paper proposes a method to improve behavior network, action selection mechanism that uses the graph of behaviors, goals and sensors with activation spreading, using goal inference mechanism of Bayesian network learned automatically. Experimental results on Khepera mobile robot show that the proposed method can generate more appropriate behaviors.
UR - http://www.scopus.com/inward/record.url?scp=7444239161&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=7444239161&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-24581-0_84
DO - 10.1007/978-3-540-24581-0_84
M3 - Conference contribution
AN - SCOPUS:7444239161
SN - 9783540206460
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 979
EP - 991
BT - AI 2003
A2 - Gedeon, Tamas D.
A2 - Fung, Lance Chun Che
A2 - Gedeon, Tamas D.
PB - Springer Verlag
Y2 - 3 December 2003 through 5 December 2003
ER -