基于小波神经网络的波浪参数预报  

Wave Parameter Prediction Based on Wavelet Neural Network

作　　者：王红萍[1] 余义德[1] 张丹[1] 

机构地区：[1]91550部队,大连116023

出　　处：《舰船电子工程》2016年第12期80-84,共5页Ship Electronic Engineering

摘　　要：预报未来时刻水下航行体的出水姿态,关键在于海浪信息的预报。在获得实测海浪谱之后,还需对其进行预报,才能得到未来某时刻的海浪谱信息。使用Matlab软件完成小波神经网络的实际构建,利用测波雷达实测得到的海浪谱数据完成频谱预报,预报目标为6~8小时后的海浪频谱数据。最后经过仿真计算,小波神经网络方法可以对海浪谱进行较为精确的预报,海浪谱预报的有效时长可达6小时。The key to predict wave lies to predict future exceeding water gesture of an underwater vehicle. On the basis of the measured spectrum, the information of wave spectrum is gained in the future. In this paper, firstly theory and method are studied to construct wavelet neural network by using Matlab. Secondly future spectrum after 6？8 hours can be forecast by analysis of data of wave observation radar . The simulation shows that this method is a high precision method to predict wave spectrum, thus possibly available for 6 hours.

关 键 词：小波神经网络 波浪参数 预报 

分 类 号：TB566[交通运输工程—水声工程]