论文部分内容阅读
首先对基于ＬＭ 算法的神经网络预测性能进行研究。相对快速ＢＰ网络而言，其预测精度高、收敛速度快。然后提出将该方法应用于电子设备ＢＩＴ输出及相关量的状态预测，对存在于航空电子设备中的压力、温度等环境应力的典型变化曲线进行了预测，并在环境应力影响下的ＢＩＴ状态综合预测中得到验证。结果表明，利用时空两方面信息进行状态预测和综合分析是一条提高ＢＩＴ诊断能力、降低虚警的重要思路。
Firstly, the prediction performance of neural network based on LM algorithm is studied. Relatively fast BP network, its prediction accuracy, fast convergence. Then, this method is proposed to predict the state of BIT output and related quantity of electronic equipment, predict the typical curve of environmental stress such as pressure and temperature existing in avionics, and synthesize the BIT state under the influence of environmental stress Projections have been verified. The results show that using both spatio-temporal information for state prediction and comprehensive analysis is an important idea to improve the diagnostic ability of BIT and reduce false alarms.