The Hydraulic Power Plant Accident Management Expert System Based On CBR And SVM

With the development of information technology, mass data has been accumulated in Database of enterprise during the process of management. How to find the valuable knowledge concealed in these data and use them to serve the enterprise have become a problem to be settled urgently. The tradition rule based reasoning meets some difficulties in obtaining the knowledge and rule.But the case based reasoning can solve the problem successfully by borrowing the way that human deal with the matter, this method features in using the previously knowledge and experience and is coincident with the thought in actual work. On the basis of large sum of reading about relative domestic and overseas literatures and analyzing the relational technology of the CBR, the expression of case, structure, storage, revise, self-study and other critical problems were studied in this paper in aiming to build the universal CBR model. The methods of altering weights of case searching KNN is used to search the best case scheme, and the SVM theory is used to automatism classify. Lastly, according to the four processes of the CBR technology and the actual accidental mode, the design and implement of the hydraulic power plant accident management expert system based on CBR is given using the techniques of the Object Oriented language Java and AAF framework.