Design And Study On The Reliability Data Analysis System

Reliability data analysis (RDA), which is the essential part of reliability engineering, is the foundation of Probabilistic Safety Assessment (PSA). Therefore the research on analysis methods and the development of RDA software have become an important branch in the worldwide reliability analysis field.Although much attention has been paid to reliability data analysis by abroad research community, little work has been done at home. Developing a software system with the independent intellectual property rights has become inevitable and urgent. In this dissertation, a user-friendly software system for RDA (named RiskWeibull) is designed based on the comprehensive investigation on the recent progress. The object-oriented programming technology and database design technology, which ensure reliability, expandability and maintainability of the system, are fully utilized in the design.The median rank function, which plays an important role in the Weibull analysis, is optimized to attain the more precise result using genetic algorithms. The feasibility of improved function has been proved in the dissertation.Division of Reactor Technology of Institute of Plasma Physics, Chinese Academy of Sciences is now engaged in PSA research. RiskA, developed by FDS team, is a large-scale reliability/PSA software system with the independent intellectual property rights. RiskWeibull can be used to evaluate the failure data from nuclear power plants, the processed result is offered to reliability database management system (named RiskBase), another self-developed software also by FDS team, and then to the quantitative analysis of RiskA.