A Fril knowledge base for the management of uncertainty in performance assessment of hazardous waste repositories

摘要

Performance assessment of a repository for hazardous waste involving uncertainties of many different kinds is considered. The authors describe a system written in Fril which can be used to manage these uncertainties and act as an intelligent front end both to the technical documentation describing the performance assessment and to software used in the performance assessment process. The fuzzy and probabilistic uncertainty features of Fril are an essential component in the knowledge engineering process. The automated system is used for storing, retrieving, and manipulating information within a performance assessment project. The demonstration system covers the far-field section of the SKI-Project 90 report but could be applied to any technical documentation.