【作者】 杨坤德；

【导师】 马远良；

【作者基本信息】 西北工业大学 ， 武器系统与运用工程， 2003， 博士

【摘要】 本文以浅海环境底质声学参数反演和匹配场被动定位为应用背景，对水声信号匹配场处理技术进行了探索性研究。论文从浅海环境参数反演、复杂水声条件下的被动定位、以及匹配场用于空域离散噪声干扰抵消等三方面展开，对匹配场的理论、方法及实验验证进行了深入地探讨。主要研究内容和取得的成果有： (1)从目标函数的建立、拷贝场计算、全局寻优算法、参数反演的不确定性分析四个方面建立了浅海环境参数反演的完整体系；提出了采用快速精确的差异进化算法进行全局寻优，可以取得比模拟退火算法和遗传算法更好的全局收敛性能；并采用后验概率分布进行反演结果的唯一性分析，改善了全局寻优反演方法的完整性，提高了反演结果的可信度。 (2)根据不同的前提条件，采用似然比的方法推导了宽带匹配场反演的最大似然目标函数；深入地研究了宽带匹配场处理中的相干与非相干问题；在分析参数反演的敏感性之后，提出了浅海环境参数宽带匹配场反演的多步优化策略，并与全参数反演方法进行了性能上的仿真比较。 (3)对东中国海实验(ASIAEX)的实际海洋环境参数反演进行了实验研究。利用实测爆炸声传播数据，对实验海域半径投弹方向进行了水平不变的参数反演，对实验海域西北和东南局部区域进行了水平变化的参数反演，得到了不同实验区域的沉积层和海底地声参数，并与Hamilton地声模型经验值、ASIAEX2000和ASIAEX2001海底采样值进行了比较，验证了差异进化算法、多步优化策略的正确性与可行性。 (4)研究了浅海复杂水声环境下的目标被动定位问题。为了降低被动定位对快拍数的要求，建立了降阶自适应匹配场处理的通用结构，比较分析了多种降阶算法；对环境参数失配条件下的两种稳健AMFP算法进行了性能对比；采用运动补偿的方法提高被动定位的精度；并采用实测数据对各种方法进行了评估。针对存在水面强干扰、非平稳噪声场和海洋环境参数失配的一般浅海复杂水声环境，提出了基于子空间分解的运动补偿算法和自适应匹配场定位的系统解决方案，并通过仿真数据和实验数据进行了验证。 (5)将平面波波束形成后干扰抵消与匹配场处理技术进行有机结合，探索性地提出了拖线阵声纳本舰干扰的匹配场抵消方法和匹配场处理后的干扰抵消方法，为本舰多途干扰抵消、水面强干扰抑制和水下弱目标检测提供了新的技术途径。同时，提出将匹配场干扰抵消应用于同方向强干扰背景下的弱目标检测，为解决平面波假设条件下无法克服的干扰抵消难题提供了新的技术途径。更多还原

【Abstract】 This dissertation studies mainly the inversion of ocean environmental parameters by MFP (Matched Field Processing). Meanwhile, adaptive matched field passive localization in complex environment and discrete noise rejection with MFP are discussed. The main contributions are as follows:(1) DE (Differential Evolution) algorithm is proposed to invert the ocean acoustic parameters in shallow water in order to get faster and more accurate results than GA (Genetic algorithm) and SA (Simulated Annealing algorithm). Also a posteriori probability analysis method is applied to evaluate the uncertainty of inversion results.(2) Maximum likelihood objective functions for broadband MFI are derived according to different conditions. Meanwhile, coherent and non-coherent broadband MFP is studied. After analyzing the sensitivity of different inversion parameters, a new approach named MS-MFI (Multi-step MFI) is proposed for inversion in order to improve the efficiency.(3) Applying MS-MFI, ASIAEX2001 (Asian Seas International Acoustic Experiments) data are used to invert environmental parameters in the East China Sea. According to the different ocean environment, range dependent and range independent inversion are studied and the inversion results are analyzed.(4) In order to reduce the snapshots needed by passive localization in complex ocean environment, reduced rank AMFP frame is founded and many algorithms are compared. In mismatch environment, two robust AMFP algorithms are compared. Meanwhile, motion compensation method is used for AMFP to improve the localization effect. With strong interferes at surface, non-stationary noise field and mismatch environment, subspace motion compensation method and efficient AMFP system frame are proposed and confirmed by simulation and real data.(5) MFPIC (Matched Field Postbeamformer Interference Canceller) is proposed to cancel tow ship’s noise. The advantage of MFPIC is that it can be used to cancel the tow ship’s noise for all arrival directions. Furthermore, this new approach can be used to detect weak target, which has the same bearing angle of a strong interference. Numerical results show that the MFPIC outperforms the conventional PIC. MFP-PIC (Matched Field Processing with Postbeamformer Interference Canceller) is also proposed to cancel strong interference and detect the underwater weak target.更多还原