基于BP神经网络的飞行动态实时预测方法

Method of Flying Dynamic Real-Time Forecast Based on BP Neural Network

针对传统Back Propagation网络 (简称BP网络 )收敛速度慢、网络灵敏度过高和隐含层数难以确定等缺陷 ,提出一个改进型BP网络 ,提高了网络预测的实时性和精确性 ;然后将之应用到飞行动态预测问题上 ,充分发挥网络模型的学习、记忆和动态自适应性的优势 ,力图解决飞行器轨迹的描述和预测问题 .

The disadvantages of the traditional BP neural network include low speed of convergence, high sensitivity and indefinite layers, etc. This paper develops a new BP network model, which can improve the quality of the real time and precision, and applies the developed BP network model to the dynamic forecast of aircraft track. Owing to the advantages of BP network in the aspects of acquisition, remembrance and self-adaptation, it can make the results of the dynamic forecast more efficient and more precise.