problem diagnosis
2.The authors present a new fast training algorithm multilayer feedforward neural network based on error back propagation. The simulation study on the fault diagnosis of certain missile attitude control system prove that BP and FBP is effective algorithms to solve the problem of fault diagnosis of missile attitude control system, and FBP is much faster in training multilayer feedforward network than BP.
3.This paper discusses a new fast training algorithm for multilayer feedforward neural network based on error back propagation (i e FBP) The simulation study of the fault diagnosis of certain missile attitude control system proves that BP and FBP are effective algorithms to solve the problem of fault diagnosis of missile attitude control system,and FBP is much faster in training multilayer feedforward neural network than BP
5.To solve the problem visualization technology is introduced into fault diagnosis field,which makes the diagnosis course visualize and more importantly,inference made by experts is simulated with plots so that it is easy for knowledge engineers to understand their thinking with imagery and extract their deep-level knowledge.

