Research of CBR Learning Strategies Based on Genetic Algorithm and BP Neural Network

In this paper we studied the insufficient of conventional case learning strategies in CBR reasoning and the research status of improved algorithm. Based on this, we further discussed the feasibility of implement case study based on genetic algorithm and the BP neural network combined, and expounds the technical details. Finally, with the example of diesel locomotive type case library in railway rescue command decision system, we validated the algorithms. The result of MATLAB 7.5 Simulation with traditional BP neural network compared indicate that BP neural network method based on genetic algorithm can effectively increase the case study ability of the CBR systems.