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
-
摘要: 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。Abstract: 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.
-
Key words:
- neural network /
- wave-height/tide-level /
- resolution /
- real time
-
郑正奇,蒋传纪,阮志伟,等.Ku波段海表波高/潮位探测系统的研究[J].微波学报,2000,16(3):214-219. 高小榕,杨福生.采用BP算法进行多层前向神经网络的训练[J].计算机学报,1996,19:687-694. MOODY J,ANTSAKLIS P J.The dependence identification neural network construction algorithm[J].IEEE Transaction on Neural Networks,1996,7(1):3-15. -
计量
- 文章访问数: 784
- HTML全文浏览量: 18
- PDF下载量: 780
- 被引次数: 0
下载:
