人工神经网络在潮位探测系统中的应用研究

Research on applying artificial neural network to detecting system of wave height/tide level

叙述了人工神经网络算法在“Ku波段微波海洋波高/潮位探测系统”中的应用研究成果,介绍了检测微波回波信号中的中频信号和9/256点插值神经网络的设计方法及训练模型,并把这两个网络应用于探测系统,使系统波高/潮位的探测精度提高了1个数量级,还保证了监测的实时性,提出了某些减少训练时间的方法。

The study results of artificial neural network algorithm applied to Ku-band microwave ocean wave-height and tide-level detecting system are specified. The way to detect the microwave returned intermediate frequency (fif) and the neural network (NN) design method of the 9/256 insertion and also training model are introduced. The precision of the measured wave-height and tide-level is risen to one order of magnitude as before but still kept low time delay while using NN. Some steps are also taken to reduce the training time of NN.