Application of BPNN and CBR on Fault Diagnosis for Missile Electronic Command System

Based on the complexity of the mobile missile electronic command system (MMECS), applying the single method in system fault diagnosis can hardly achieve satisfactory results. The fault diagnosis system combining the BP neural network (BPNN) method and the case-based reasoning (CBR) method was presented. The framework of the mixed neural network and the case presentation was put forward. The question of redundancy reasoning was solved, moreover, it can interpret the diagnoses by providing the successful case. Finally, with the example of voice interrupt, the system´s correctness and validity was proved. It is shown that the system is suitable for both the operators training and online decision making for the army