File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Conference Paper: Adaptive fuzzy approach to obstacle avoidance

TitleAdaptive fuzzy approach to obstacle avoidance
Authors
Issue Date1998
Citation
Proceedings Of The Ieee International Conference On Systems, Man And Cybernetics, 1998, v. 4, p. 3418-3423 How to Cite?
AbstractReinforcement learning based on a new training method previously reported guarantees convergence and an almost complete set of rules. However, there are two shortcomings remained: first, the membership functions of the input sensor readings are determined manually and take the same form; and second, there are still a small number of blank rules needed to be manually inserted. To address these two issues, this paper proposes an adaptive fuzzy approach using a supervised learning method based on back propagation to determine the parameters for the membership functions for each sensor reading. By having different input fuzzy sets, each sensor reading contributes differently in avoiding obstacles. Our simulations show that the proposed system converges rapidly to a complete set of rules, and if there are no conflicting input-output data pairs in the training sets, the proposed system performs collision-free obstacle avoidance.
Persistent Identifierhttp://hdl.handle.net/10722/158260
ISSN

 

DC FieldValueLanguage
dc.contributor.authorYung, NHCen_US
dc.contributor.authorYe, Cen_US
dc.date.accessioned2012-08-08T08:58:47Z-
dc.date.available2012-08-08T08:58:47Z-
dc.date.issued1998en_US
dc.identifier.citationProceedings Of The Ieee International Conference On Systems, Man And Cybernetics, 1998, v. 4, p. 3418-3423en_US
dc.identifier.issn0884-3627en_US
dc.identifier.urihttp://hdl.handle.net/10722/158260-
dc.description.abstractReinforcement learning based on a new training method previously reported guarantees convergence and an almost complete set of rules. However, there are two shortcomings remained: first, the membership functions of the input sensor readings are determined manually and take the same form; and second, there are still a small number of blank rules needed to be manually inserted. To address these two issues, this paper proposes an adaptive fuzzy approach using a supervised learning method based on back propagation to determine the parameters for the membership functions for each sensor reading. By having different input fuzzy sets, each sensor reading contributes differently in avoiding obstacles. Our simulations show that the proposed system converges rapidly to a complete set of rules, and if there are no conflicting input-output data pairs in the training sets, the proposed system performs collision-free obstacle avoidance.en_US
dc.languageengen_US
dc.relation.ispartofProceedings of the IEEE International Conference on Systems, Man and Cyberneticsen_US
dc.titleAdaptive fuzzy approach to obstacle avoidanceen_US
dc.typeConference_Paperen_US
dc.identifier.emailYung, NHC:nyung@eee.hku.hken_US
dc.identifier.authorityYung, NHC=rp00226en_US
dc.description.naturelink_to_subscribed_fulltexten_US
dc.identifier.scopuseid_2-s2.0-0032316543en_US
dc.identifier.volume4en_US
dc.identifier.spage3418en_US
dc.identifier.epage3423en_US
dc.identifier.scopusauthoridYung, NHC=7003473369en_US
dc.identifier.scopusauthoridYe, C=7202201245en_US

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats