【作者】 唐艳红；

【导师】 郝燕玲；

【作者基本信息】 哈尔滨工程大学 ， 导航、制导与控制， 2010， 博士

【摘要】 海浪是最为常见且最为重要的海洋波动现象,它与海洋探索、海洋开发利用等许多重要问题,特别是海洋军事技术和海上航行安全等有着十分密切的关系。雷达技术的发展,特别是航海雷达技术的进步为观测海浪提供了重要的观测条件。航海雷达观测海浪主要是通过分析来自海洋表面、雷达接收的海面回波随时空的演变,经过波谱分析计算得到实时的海表面流和海况参数(波峰周期、波峰波向和有效波高)。由于航海雷达观测海浪时,无需直接接触海面进行观测,具有其它海浪观测方法无法比拟的优越性和广泛应用价值,因此,利用航海雷达观测海浪越来越引起国内外学者的广泛关注。本文在系统分析了海浪遥测技术理论研究成果及其发展概况的基础上,对海浪图像数据采集,海浪图像预处理,海流估计以及海浪参数反演等关键技术进行了深入研究。针对课题的具体需求,本文在研究应用雷达单元构成原理和各路输出信号功能特点的基础上,提出了一种以间歇采样的模式对雷达多路信号同步采集的方案。通过深入分析PXI-9820高速数据采集卡的功能特性,对数据采集硬件平台进行了方案设计,同时,研究并实现了基于Windows/PXI系统的雷达回波信号数据采集软件。通过实地岸基试验验证了数据采集系统设计方案的可行性以及系统运行的可靠性。考虑到海浪遥感测量时,获取的雷达原始回波信号中具有雨雪杂波和同频干扰噪声的情况,本文在系统分析大量降雨图像特征的基础上,提出利用图像的统计特性—均值和变异系数作为图像是否受到降雨影响的判断依据。同时,提出一种基于小波变换的海浪图像同频干扰噪声抑制算法,在对海浪图像进行多分辨率分解提取出含有细节信息的高频分量后,利用图像直方图的阈值来区分图像的同频干扰噪声和海浪信号,然后对检测出的噪声点进行恢复。实验结果表明,本文算法能够取得较好的降噪效果。考虑到海流估计的主要技术难点,为了减少因海流的误差带来的海浪方向谱的误差,本文提出一种基于误差序列方差最小的迭代估流算法,以方差最小为估计海流的目标函数,利用属于某个色散关系的集合的整体属性,对海流进行估计,同时,利用高斯正态分布原理,提出了确定属于某个色散关系的能量集合的方法。实验结果表明,利用本文提出的算法进行海流估计,能够提高海流估计的稳定性和精度。考虑到海浪参数反演的具体要求,本文提出一种改进的带通滤波算法,通过对色散关系方程的单调性、连续性以及收敛性的深入研究,确定发生色散关系的主要区域,以此为基础,设计了一个更加合理、准确的带通滤波器,用以改善海浪参数的反演精度。实验结果表明,利用本文提出算法进行海浪参数反演,能够更加准确的提取海浪的频谱信息。更多还原

【Abstract】 Ocean wave is the most common and important ocean fluctuation phenomenon, which has close relations with many important issues such as ocean exploration, ocean development and utilization, especially marine military technology and navigation safety. An important condition of ocean wave observation is provided by the development of radar technology, especially in marine radar technology.The marine radar observing waves is realized through the analysis of sea echo varied with time and space, which come from the ocean surface, the real-time near-surface current and sea state parameters (peak wave period, peak wave direction and significant wave height.) obtained by wave spectral analysis.The method of wave observed by marine radar has incomparable superiority and greatly potential application value compared with other method, for there is no direct contact between sea and marine radar. Therefore, the use of marine radar observing waves is increasingly attracted wide attention from scholars at home and abroad.In this dissertation, based on the theory and development of ocean remote sensing technology are systematically analyzed, the key technologies of ocean wave observation, wave image data acquisition, wave image pre-processing, current estimation and ocean wave parameters inversion are deeply researched.In view of specific demands of the subject, based on the radar units and the function feature of radar output signal are researched, a multi-channel signal acquisition method based on interval sampling mode is presented in this dissertation. Through profound analysis of function feature of PXI-9820 high-speed data acquisition card, the hardware platform of data acquisition is designed. At the same time, the radar echo signal data acquisition software is designed and realized based on Windows/PXI system. At last, the feasibility and reliability of the platform is proved by the subsequent shore-based experimentation.In this dissertation, the rain clutter and the same frequency interference noise are taken into account.Firstly, on the basis of systematic analysis of a large number of rainfall image features, an analyzing and processing method is proposed for rainfall observation data in wave monitoring system by marine radar which uses the statistic characteristic-mean value and variation coefficient of radar echo image in order to distinguish rainfall and non-rainfall image.Secondly, in order to inhibit the same frequency interference noise effectively, a novel algorithm for co-channel interference noise suppression is proposed based on the wavelet transform in this dissertation. The high-frequency details part of image are extracted by the wavelet transform.According to the histogram distribution of wave image, a threshold value is used to determine sea echo signal points and noise points.Then the property values of determined noise points are substituted with the adjacent signal points.The result of Experiments shows that the algorithm is effective for the image noise suppression.Considering the main problem of current estimates, a novel algorithm for retrieving ocean surface currents is presented in order to reduce errors in wave directional spectrum, which based on the minimum variance and iterative algorithm. The objective function of current estimate is based on minimum variance.The ocean current is determined by using dispersion relations. At the same time, the energy set of the dispersion relation is determined using the Gaussian normal distribution principle.Experimental results show that the stability and accuracy of the currents estimation could be improved by the proposed algorithm in this dissertation.Taking into account the specific requirements of ocean wave parameter inversion, an improved band-pass filtering algorithm is proposed in this dissertation. Through a deep study of the monotonicity, continuity and convergence of the dispersion relation equation, the main areas of dispersion relations could be determined. Based on this algorithm, a more reasonable and accurate band-pass filter is designed to improve the accuracy of wave parameters inversion. Experimental results show that more accurate information of the wave spectrum for wave parameters inversion could be extracted by the improved band-pass filtering algorithm.更多还原