文档介绍：南京邮电大学
硕士学位论文
多普勒气象雷达原始回波数据混合压缩算法研究
姓名:李啸
申请学位级别:硕士
专业:信号与信息处理
指导教师:曹雪虹
2011-03
南京邮电大学硕士研究生学位论文摘要
摘要
多部多普勒气象雷达组成分布式网络雷达系统,雷达探测区域互相重合,相互通信,进
一步提高了对灾害性天气的监测预警能力。但是,多部雷达产生的原始回波数据量十分巨大,
不适合当前气象网络传输。因此,本文对分布式全固态网络气象雷达系统中的数据传输与压
缩算法进行研究,目的在于压缩原始回波数据以降低数据传输率。研究工作主要从以下几个
方面展开:
(1) 在研究了气象雷达原始回波数据格式的基础上,从时域、频域、相关性和熵值特性四个
方面对数据进行仿真分析,挖掘出原始回波数据的特征和相关性。
(2) 结合气象雷达原始回波数据的特点,重点研究基于上下文模型的无损压缩算法,对实测
数据进行对比压缩实验,提出无损压缩算法改进方案。
(3) 重点分析了使用小波变换模极大值的方法对气象雷达原始回波数据去噪声的思想,结合
对原始回波数据特征的挖掘以及无损压缩改进算法,提出一种基于小波变换模极大值去噪声
的混合数据压缩系统,给出算法实现和仿真分析结果。
(4) 从气象基数据参数(强度、速度、谱宽)的角度对去噪声后的数据进行损失评估,计算损失
信噪比和数据压缩比。实验结果表明:本文提出的算法能够在一定失真率的前提下有效的对
气象雷达原始回波数据进行压缩。

关键词:多普勒气象雷达、数据压缩、小波变换模极大值、去噪
I
南京邮电大学硕士研究生学位论文 Abstract
Abstract
work of Doppler Weather Radar System is constituted by multiple radar sites.
The detection area is coincident municating with each other to further improve the warning
capabilities of disaster weather. However, several radars generate enormous amount of radar raw
data which is not suitable work transmission. Therefore, pression algorithm in this
network of distributed solid state weather radar system is studied for the purpose of reducing the
data transfer rate. Research is carried out as follows:
(1) Based on studying the weather radar raw data format, simulation and analysis experiments are
carried out from time domain, frequency domain, correlation property and entropy property to the raw
data for mining data features.
(2) Combined with the characteristics of raw data, pression algorithm based on context
model is studied. The improved algorithm is proposed paring the result pression
experiments.
(3) The idea of reducing noise using wavelet transform modulus maxima(WTMM) method for
weather radar echo data is studied, combining with features of the original echo and the improved
pression algorithm, a mixed pression system based on WTMM denoising is
given and simulation results is shown.
(4) pression ratio and SNR r