基于BP神经网络的智能BIT故障诊断系统研究

Research of Fault Diagnosis System of Intelligent BIT Based on BP Neural Network

论述了智能BIT的设计、检测、诊断、决策四个方面的主要研究内容,分析了BP神经网络的网络模型及工作原理,构建了基于BP神经网络的智能BIT故障诊断系统,并用某雷达录取终端的故障实际数据进行了验证。结果表明将神经网络与智能BIT结合是一种有效的诊断方法,解决了传统BIT故障诊断能力不足,导致系统虚警率过高、自适应性能差等问题,使被测系统具有更高的故障诊断能力。

The chief research content of intelligent BIT design, test, diagnosis and decision is discussed. The BP net- work model and its working principle is analysed. Fault diagnosis system of intelligent BIT based on BP neural network is constructed. Availability of the method of intelligent BIT based on BP neural network is verified by fault data of a radar extraction terminal. The experimental results indicate the method of intelligent BIT based on BP neural network is availability, insufficient of traditional BIT＇s fault diagnosis function is overcomed, which leads to high false alarm rate, and reduces the testability and the capability of fault diagnosis.