西南石油大学学报(自然科学版) ›› 2005, Vol. 27 ›› Issue (4): 56-59.DOI: 10.3863/j.issn.1000-2634.2005.04.17

• 石油机械工程及其它 • Previous Articles     Next Articles

A THREE-VALUED LOGIC NEURON MODEL AND ITS APPLICATIONS IN UNCERTAIN REASONING

CHEN Xing,DIAO Yong-feng.   

  1. lnstitute of Coniputer Application,China West Normal University, Nanchong Sichuan, 637002
  • Received:2004-05-18 Revised:1900-01-01 Online:2005-08-20 Published:2005-08-20
  • Contact: CHEN Xing

Abstract:

As to the weakness of the binary logic neuron in uncertain knowledge representation and reasoning, a new neuron model (three valued logic neuron, TLN) was presented in this paper.It can express three valued logic knowledge.There are two different TLNs which are TLN-AND(three valued “logic and” neuron) and TLN-OR(three valued “logic or” neuron).This kind of neuron can represent uncertainty of rules and truth value.Neural chains made up of this kind of neurons can be used as three valued logic inference system, and can present and inference ternary, even the preconditions are not sufficient, conclusions can get.


Key words: neuron, ternary adjacent rule, uncertain knowledge representetion and reasoning

CLC Number: