Probabilistic risk assessment and intelligent decision support systems

The purpose of this paper is to review the progress made in recent years in both the area of PRA as a support to AI applications and the area of AI applications in PRA. The emphasis is on the areas that have made some progress in the past few years, with a brief description of the methods and a discussion of the potential uses and weaknesses. Also included is a brief review of recent developments in the theory of uncertainty in the AI community that may impact uncertainty modeling in PRA. AI techniques could be applied to the related field of PRA in several ways. In this discussion, however, the scope is limited to emergency diagnosis and accident management, because these are the areas that have attracted most of the attention in recent years. The potential use of PRA as a support to these applications is discussed in detail, and this is followed by a survey of recent developments in these areas. (orig./GL)