Product reliability design knowledge reasoning method based on rough sets and certainty factors theory

The empirical data of product reliability design are often vague, comprehensive and incomplete. Information retrieval for reliability design depends on mining the knowledge applied to specific designing from mass empirical data. At the same time, designers are often lack of the judgments to accept or reject the reasoning results of reliability design knowledge. For these two issues, this paper presents a method based on rough sets and certainty factors theory. The method first calculates the belief of rules, which extracts from original information using rough sets theory. Then find the belief of results via certainty factors theory. Unlike the traditional way, the method considers the experiment data more comprehensively, and achieving the quantitative expression of rules´ belief.