不确定海洋环境中的稳健信号检测及性能预测方法

项目名称： 不确定海洋环境中的稳健信号检测及性能预测方法

项目编号： No.11274252

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 数理科学和化学

项目作者： 孙超

作者单位： 西北工业大学

项目金额： 95万元

中文摘要： 现代声呐系统面临的重大挑战之一是如何在复杂多变且不确定的海洋环境中检测弱目标信号，具体困难在于：1）随着减振降噪技术的发展，待探测目标的辐射噪声级越来越低，检测难度日益加剧；2）工作环境由深海转向浅海，水声传播特性更加复杂多变，由海洋环境决定的不确定性因素增多、影响严重。然而，目前大多数的信号检测算法没有考虑将环境的先验信息应用到检测器中，从而造成检测性能损失，同时也缺乏对检测器性能的有效预测。 本申请项目拟以统计海洋模型为基础，提取不确定海洋环境中的参数统计特性，分析海洋环境参数变化引起的声传播起伏和信号检测器性能损失，通过使用主成分量、模态加权等方法有效提取声基阵接收数据中的稳健成分，或是综合利用贝叶斯原理和广义似然比方法，构建存在海洋环境参数不确定时的快速稳健信号检测器，同时研究检测器性能的快速预测方法，为提高不确定海洋环境中声呐系统对远程弱目标的探测能力奠定必要的理论和方法基础。

中文关键词： 不确定海洋声环境；信号检测；性能检测；稳健处理；

英文摘要： One of the main challenges modern sonar systems are facing with is the detection of weak target signals in an uncertain ocean environment. Specific difficutlies stem from 1) the target to be detected is gettting weaker gradually, and 2) the ocean environment is becoming worse when sonar sytems working in shallow waters. On the other hand, most signal detection algorithms do not take the a prior information of the environment into account when designing the detectors, therefore, performance degradation is unavoidable when applied in practical situations. Meanwhile, the performance of the detectors is predicted by the Monte Carlo method, which generally takes quite a long time. Based on the realization of this situation and understanding of the unrealities of ocean acoustic environments, the project is aiming on studies of robust detection algorithms and the corresponding performance prediction method. Based on the a prior information of the ocean area under consideration and also the extracted features from data measurements, the statistical model is used to describe the uncertain ocean acoustic environment. The fluctuations of acoustic signals will be analyzed and so is the performance degradation induced by such factors. The will be designed follow two separate thoughts: 1) based on the principal component

英文关键词： uncertain ocean acoustic environment；signal detection；performance prediction；robust processing；