摘要
文章以机翼载荷为例,构建了载荷-飞参的BP神经网络模型,通过对比预测载荷与实测载荷的误差,对神经网络模型的泛化能力进行了分析。结果表明:神经网络模型预测载荷的最大平均误差为7.89%,工程上可用于其他飞行起落的载荷预测,对开展单机寿命监控、单机载荷谱的编制具有重要意义。
In this paper,the wing load is taken as an example,the BP neural network model of the load-fly parameter is constructed,and the generalization ability of the neural network model is analyzed and discussed by comparing the errors between the predicted load and the measured load.The results show that the maximum average error of the load predicted by the neural network model is 7.89%,which can be used to predict the load of other take-off and landing in engineering,and has great significance to the individual aircraft life monitoring and the compiling of individual aircraft load spectrum.
作者
陈奇
安彦
CHEN Qi;AN Yan(Avic Aircraft Co.,Ltd.,Xi'an 710089,China)
关键词
神经网络
飞行参数
泛化能力
neural network
flight parameter
generalization ability
