The reliability theory based expert system research for Subway construction safety risk identification

According to the existence of uncertainty in the subway risk identification, introduced rule-based expert system technology, integrated expertise into the knowledge base, automatically identified construction parameters based on CAD drawings, discussed and designed automatic subway security risk identification expert system based on confidence factor theory. And taken underflow sand pit in subway station construction as an example, elaborated the key issues in the knowledge base construction in details, such as risk mechanism analysis, fact base structuring and rule base structuring. And proposed to combine the contribution (C (eij)) with fact credibility(F (eij))to confirm the evidence credibility CF(eij) =C(eij) X F(eij), solving the credibility assignment conflict when the fact of certainty and uncertainty existed simultaneously; Selected the threshold of risk conclusions credibility in representative conditions as the sub-boundary for risk potential classification, avoiding the risk omission problem. In this way, the application of expert intelligent decision technology was broadened in subway construction.