用BP网络算法获取电子浓度真高剖面

OBTATNING THE PROTILE OF IONOSPHERIC ELECTION DENSITY VS REAL HEIGHT WITH BACKPROPAGATION NETWORK

提出了一种从垂测电离图获取电子浓度真高剖面的新方法——网络逼近法．该方法通过ＢＰ网络算法求解群路径积分方程，得到了电子浓度真高剖面．并对给定的抛物模式作了验证计算。

A new method approximation with BP network to obtain the profile of ionospheric electron density versus real height is described, how to resolve the group path integral equation with BP neural network is demonstrated, and the computation result with this method in resolving the ionogram correspondent with a parabolic model is givn in this paper.