基于目标与混响散射特性的特征空间联合建模与分离

项目名称： 基于目标与混响散射特性的特征空间联合建模与分离

项目编号： No.51279033

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 水利工程

项目作者： 李秀坤

作者单位： 哈尔滨工程大学

项目金额： 62万元

中文摘要： 水下目标探测识别技术是水声设备和水中武器系统智能化的关键技术之一。由于目标回波信号微弱、回波产生机理复杂、海底混响干扰强以及海底沉积层吸收等原因，使得安静型沉底与掩埋目标探测识别至今仍是一个世界级难题。本项目针对远距离掩埋水雷目标探测的需求，提出了可将混响视为与其散射特性相关的具有稳定信号性质和特征的一类信号的概念；研究基于目标与混响散射特性的信号特征；建立目标与混响的联合特征空间以及研究在特征空间下对目标与混响进行分离的稳定性和普适性。为掩埋目标与混响在特征空间建模与识别构建理论框架，为主动声纳识别技术提供理论基础。

中文关键词： 主动声呐；混响；特征空间；盲分离；目标分类

英文摘要： Underwater target detection and recognition is one of the key technologies of intelligent acoustic equipments and weapon systems in the water. Since the target echo signal is weak, echo mechanism is very complex, the strong bottom reverberation interference, seabed sediment layer absorption and other reasons make the detection and recognition of quiet type targets either sink or buried is still a world-class problem. For the needs of long distance buried mines detection, the concept of reverberation will be regarded as a class of signal with stable scattering properties; the signal characteristics of target and reverberation; the establishment of joint feature space and the researches of separation stability and universality based on the feature space are proposed in this project. The work of this project will build a framework for modeling and identification of buried targets and reverberation in the feature space and provide a theoretical basis for active sonar recognition technology.

英文关键词： active sonar；reverberation；feature space；blind source separation；target classification