Chen Guo
Software Technology Institute, Dalian Jiaotong University, Dalian, China
Ming Huang
Software Technology Institute, Dalian Jiaotong University, Dalian, China
Xu Liang
Software Technology Institute, Dalian Jiaotong University, Dalian, China
ABSTRACT
Uncertainty problem is the focus of the Expert System Research. This study utilizes the thought of Expert Systems confidence as guidance and the mechanical products design as example to introduce the process of the uncertainty reasoning research of the expert system detailedly. In the process of uncertainty research, we firstly propose a new matching algorithm of uncertainty reasoning for matching the method of the first component of rules. Compare with the existing method, this algorithm taking fully into account of the experts judgement for known first component of rules and realistic conditions influence for the first component of rules. It makes the algorithms result fit the actual rules choosing better. Furthermore, in the process of uncertainty dissemination, this study uses a new algorithm to solve the supporting conclusions evidence confidence. Compare with the existing method, the confidence of conclusions which use this algorithm are all over 0.8. As the result, it shows the uncertainty problem in inference engine has been solved very well. And it also improves the choosing decision-making function of the expert system.
PDF References Citation
Received: August 02, 2013;
Accepted: October 06, 2013;
Published: November 12, 2013
How to cite this article
Chen Guo, Ming Huang and Xu Liang, 2013. Uncertainy Reasoning Research Based on the Expert System of
Mechanical Product Design. Journal of Applied Sciences, 13: 4902-4906.
DOI: 10.3923/jas.2013.4902.4906
URL: https://scialert.net/abstract/?doi=jas.2013.4902.4906
DOI: 10.3923/jas.2013.4902.4906
URL: https://scialert.net/abstract/?doi=jas.2013.4902.4906