The intelligent BIT design of aviation integrated computer system based on improved BP neural network

In this paper, aiming at some difficulties such as in-flight fault detection and fault isolation of aviation integrated computer system for its complex structure and function, an intelligent BIT design method of aviation integrated computer system, which combining integrated BIT and information-enhanced BIT, has been given after analyzing the research content and development of intelligent BIT. The improved BP neural network has been introduced and the algorithm is applied to the intelligent BIT design. The power process module of aviation computer system is taken as an example for MATLAB simulation. Availability of the intelligent BIT method based on improved BP neural network is partly verified by the simulation result. The insufficient of traditional BIT´s fault diagnostics ability is overcome, which leads to high false alarm rate. The intelligent BIT design improves the testability and the capability of fault diagnostics and has strong practical value.