Research and Application of Knowledge Reduction Method for Case Base

The method of case-based reasoning is introduced into the adjustment of rail transit metro diagram, which can solve the problem that it takes a long time for the simulation system to compile operation diagram and meet the needs of solving emergencies. However, with the incremental learning of knowledge in the case base of rail transit metro diagram, the problem of knowledge redundancy will become more and more serious, which greatly affects the efficiency of case retrieval. Therefore, on the premise of keeping the decision ability of case-based reasoning system unchanged, the research on knowledge reduction of case base is a focus of current case base maintenance and case-based reasoning research and application. According to the characteristics of urban rail transit case base knowledge, this paper studies from two aspects of case attribute reduction and case value reduction, and proposes two knowledge reduction methods for the case base system, attribute reduction method based on fuzzy rough set and heuristic value reduction algorithm. The proposed fuzzy rough set attribute reduction method based on the standard European distance and F-measure method is accurate to 93.33% in case identification of rail transit case base, which is accurate to 93.33% in case identification of rail transit case base, and its effectiveness is evaluated from the precision rate of retrieval results. The heuristic value reduction algorithm is introduced to reduce the value of the rail transit decision information table, and an integration model based on rules and case-based reasoning (CBR) is constructed. The case database system of rail transit energy consumption is designed, and the validity of the knowledge reduction method is verified by the retrieval time of actual cases. According to the characteristics of the case database of the metro diagram, the fuzzy rough set attribute reduction method is introduced, and the heuristic value reduction method is proposed, which improves the efficiency of the case attribute value reduction and the accuracy of the reduced result. A new way of thinking for the retrieval of large scale data sets is provided. It further meets the energy saving requirements and operating efficiency of the case base of rail transit.