【作者】 许小可；

【导师】 柳晓鸣；

【作者基本信息】 大连海事大学 ， 通信与信息系统， 2008， 博士

【摘要】 海杂波通常是指海洋表面的雷达后向散射回波,严重干扰了雷达对海面目标的检测性能,因此海杂波研究对于雷达系统设计、雷达信号处理和海面目标检测具有非常重要的意义。混沌和分形都是非线性科学的重要分支,它们在诸多领域获得了广泛应用。本文关注的是如何使用混沌、分形等非线性理论来处理海杂波和检测弱小目标这一前沿课题。在研究中注重将理论研究和实证研究相结合,一方面采用最新的非线性理论对海杂波进行分析和研究,另一方面结合IPIX雷达实测数据来检验海杂波处理和弱小目标检测方法的有效性。本文分析了海杂波在统计模型建模后运用最大似然比检测准则下,很难检测出弱小目标的缺陷,改进了非广延分布模型在海杂波建模和弱小目标检测领域的应用。使用替代数据法分析了海杂波非线性性质,根据海杂波和弱小目标非线性性质差异,提出了基于替代数据的弱小目标检测方法。在确定海杂波具有非线性性质后,使用混沌和分形两种非线性方法对海杂波进行了研究。首先,使用Cao方法弥补了伪临近点法进行海杂波相空间重构不准确的缺陷,并使用该方法定性分析出海杂波是由随机成分和确定成分共同组成的。其次,通过仿真实验和实际数据分析,发现噪声严重影响了关联维数和最大Lyapunov指数的计算,指出使用这两种不变量判断时间序列是否具有混沌特性的局限性。最后,使用分形理论分析了海杂波的分形特性,提出基于空间分形特征差异的目标检测算法,提高了海杂波中弱小目标的检测性能。本文使用非线性理论分析了海杂波性质,拓展了非线性理论的实际应用领域,深化了海杂波研究领域对海杂波性质和物理机制的理解。研究了关联维数和最大Lyapunov指数等混沌不变量在处理海杂波时间序列的局限性,为其他领域实际时间序列的非线性分析提供了借鉴。根据海杂波和弱小目标数据的非线性特征差异,改进了非广延模型的目标检测算法,提出了基于替代数据和空间分形特征差异的目标检测方法。这些非线性目标检测方法,能在不增加硬件成本的前提下,提高海杂波中弱小目标的检测能力,因此具有一定的理论价值和实际应用的指导意义。更多还原

【Abstract】 Sea clutter refers to radar backscatter wave from sea surface, and it seriously interferes in detection performance of targets within sea clutter. So it is great significance to study sea clutter for radar system design, radar signal processing and targets detection within sea clutter. Chaos and Fractal are two branches of nonlinear science, which have comprehensive applications on a great many of research areas. It has been researched how to use the Chaos, Fractal and other nonlinear theories on sea clutter processing and small targets detection in this paper. Theoretical research and experiment research are combined in this dissertation. On the one hand, the lastest nonlinear theories are used to analyze and process sea clutter. On the other hand, IPIX real-life data are used to verify the effectiveness of the theories and methods.The limitation that small and weak targets within sea clutter are difficult to be detected by statistical model and likelihood ratio test is analyzed in this paper. The application of nonextensive distribution model for sea clutter modeling and small targets detecting is improved. Surrogate data method is used to analyze nonlinear character of sea clutter. According to nonlinear characters difference between sea clutter and small targets, small targets detecting method based on surrogate data method is proposed. After nonlinear characters of sea clutter are confirmed, sea clutter is researched by two nonlinear methods of Chaos and Fractal. Firstly, False Nearest Neighbors method is improved by Cao method, and accurate phase reconstruction parameters of sea clutter and targets data are obtained. It is analyzed that sea clutter is a composite of stochastic component and determinate component quantitatively by Cao method. Secondly, through the simulation experiment and real-life data analysis, it is found that the calculation of correlation dimension and largest lyapunov exponent is seriously influenced by noise. And the limitations of judging whether time series has chaotic character or not by the two chaotic invariants are pointed out. At last, fractal theory is used to analyze fractal characters of sea clutter. Targets detection method based on spatial fractal character difference is proposed, which improved the detection performance of small targets within sea clutter.Nonlinear theories are used to analyze the characters of sea clutter, which widens the practical applications of nonlinear theories and deepens the understandings of characters and physical mechanism of sea clutter. The limitations of correlation dimension and largest lyapunov exponent on processing sea clutter time series are researched, which can provide help for other real-life time series nonlinear analysis. According to nonlinear characters difference between sea clutter and small targets, targets detection method of nonextensive model is improved, targets detection methods based on surrogate data and spatial fractal character difference are proposed. All these nonlinear detection methods can improve detection performance of small targets within sea clutter, which no need to add any hardware facility. So this research has important theoretical value and referenced significance in real engineering application.更多还原