基于海杂波混沌特性的弱目标检测

发布时间：2018-05-26 06:41

本文选题：海杂波 + 混沌　；参考：《大连海事大学》2017年硕士论文

【摘要】：随着雷达高新技术的发展,海杂波背景下的目标检测成为海面雷达信号处理的重要组成部分。海杂波通常指海洋表面的雷达后向散射回波,导致雷达对海面目标的检测性能产生严重的干扰。海杂波中弱目标的检测,传统方法是基于海杂波统计特性,但是统计特性并不能很好地反映海杂波的内在动力学特性,因此检测效果很不理想。随着海杂波混沌特性的出现为海面弱目标检测提供了新方法。本文通过分析海杂波的混沌特性研究了基于自适应RBF模糊神经网络的弱目标检测方法。主要研究内容分为以下四部分:第一部分包含本文的第1至2章内容:这部分首先叙述了海杂波研究的意义,从统计理论、分形理论、混沌理论阐述了海杂波的研究现状。然后对混沌理论、几种常见的混沌吸引子的模型进行描述,最后阐述了 IPIX雷达海杂波数据的来源与采集情况。第二部分包含本文的第3章内容:该部分对实测雷达海杂波数据的混沌动态特性进行了验证。阐述了著名的Takens定理并对海杂波时间序列进行相空间重构,对相空间重构的关键参数延迟时间和嵌入维数进行了计算。本文采用自相关函数法和互信息量法计算延迟时间,采用虚假近邻法和Cao方法计算嵌入维数。采用GP方法和小数据量方法分别计算关联维数和最大Lyapunov指数,并根据有限的关联维数和最大Lyapunov指数大于零,验证了海杂波的混沌特性。第三部分包含本文的第4章内容:本部分利用海杂波的混沌特性,研究了基于自适应RBF模糊神经网络的混沌背景下弱目标检测的方法,建立了单步预测模型,从预测误差中检测出淹没在混沌背景中的微弱目标。分别以Lorenz系统和实测的IPIX雷达数据作为混沌背景进行实验研究。仿真实验说明此方法能够有效的将混沌背景中的信号检测出来。最后通过与RBF神经网络预测误差进行比较,可得出自适应RBF模糊神经网络具有更好的预测效果,因此能够对目标实现更好的检测。第四部分包含本文的第五章内容：本部分根据软件模块化思想,将上述相关算法在Matlab软件上进行集成,设计了良好的人机交互性能的海杂波混沌特性目标检测平台。该平台包括相空间重构、混沌识别、目标检测三大功能模块,具有较好的实用性。
[Abstract]:With the development of radar technology, target detection in sea clutter background becomes an important part of sea surface radar signal processing. Sea clutter usually refers to the radar backscattering echo on the ocean surface, which results in serious jamming to the detection performance of the sea surface target. The traditional method of weak target detection in sea clutter is based on the statistical characteristics of sea clutter, but the statistical characteristics can not well reflect the inherent dynamic characteristics of sea clutter, so the detection effect is not satisfactory. With the emergence of chaotic characteristics of sea clutter, a new method is provided for weak target detection. In this paper, the weak target detection method based on adaptive RBF fuzzy neural network is studied by analyzing the chaotic characteristics of sea clutter. The main research contents are as follows: the first part includes chapters 1 to 2 of this paper: this part first describes the significance of the study of sea clutter, from the statistical theory, fractal theory, chaos theory to explain the research status of sea clutter. Then the chaos theory and several common models of chaotic attractor are described. Finally, the source and acquisition of sea clutter data of IPIX radar are described. The second part includes the third chapter of this paper: in this part, the chaotic dynamic characteristics of radar sea clutter data are verified. The famous Takens theorem is introduced and the phase space reconstruction of sea clutter time series is carried out. The key parameter delay time and embedding dimension of phase space reconstruction are calculated. In this paper, autocorrelation function method and mutual information method are used to calculate the delay time, and false nearest neighbor method and Cao method are used to calculate the embedding dimension. The GP method and the small data method are used to calculate the correlation dimension and the maximum Lyapunov exponent respectively, and the chaos characteristics of sea clutter are verified according to the finite correlation dimension and the maximum Lyapunov exponent greater than zero. The third part includes the content of chapter 4 of this paper: in this part, the method of weak target detection based on adaptive RBF fuzzy neural network is studied, and a one-step prediction model is established, which is based on the chaotic characteristics of sea clutter. Weak targets submerged in chaotic background are detected from prediction errors. The chaotic background of Lorenz system and the measured IPIX radar data are studied experimentally. The simulation results show that this method can effectively detect the signals in chaotic background. Finally, by comparing with the prediction error of RBF neural network, it can be concluded that adaptive RBF fuzzy neural network has better prediction effect, so it can detect the target better. The fourth part includes the fifth chapter of this paper: according to the idea of software modularization, the above algorithms are integrated on Matlab software, and a good human-computer interaction platform is designed to detect the chaotic characteristics of sea clutter. The platform consists of three functional modules: phase space reconstruction, chaos recognition and target detection.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN957.51

【相似文献】