CBR-Based Grey Relational Identification Method for Technical Risks in Equipment Tests

Considering that poor/limited data about technical risks is obtained in equipment tests, it's not a good idea to use existing methods to identify and deal with new risks. In this paper, a four-step method which combining case based reasoning and grey relational analysis is proposed. By utilizing prior information, a case library, in which every case is described in details representing for certain kind of technical risk, is built. With a new kind of risk symptom rising, the previous library is retrieved to find the case with largest conditional degree, and then a recommendation solution for the new risk symptom is given according to the solution which is taken to solve the risk with largest conditional degree mentioned above. The knowledge comes into being in the process of dealing with the new risk symptom, is combined into a new case, by creating or modifying the existing case. At last, the new case is stored in the case library. A practical example is provided. The results show that the CBR-based grey relational technical risk identification method is sensible and feasible.