浅海波导中基于模糊逻辑的恒虚警检测方法研究

项目名称： 浅海波导中基于模糊逻辑的恒虚警检测方法研究

项目编号： No.60802072

项目类型： 青年科学基金项目

立项/批准年度： 2009

项目学科： 电工技术

项目作者： 郝程鹏

作者单位： 中国科学院声学研究所

项目金额： 25万元

中文摘要： 浅海波导中目标的检测有别于传统的声纳信号检测，其一混响的统计分布特性远离瑞利分布，导致虚警概率增加；其二目标由常规的点目标变为具有多个散射中心的距离扩展目标，给目标信号的检测带来困难。模糊逻辑是一种解决不确定、不精确信息的方法，而声纳信号所具有的不确定性往往是由模糊性所引起的。本课题开展距离扩展目标的模糊恒虚警检测方法研究，将目标的多散射中心特征及混响的非高斯特性创新地加以利用。设计模糊隶属函数平滑混响背景，降低虚警概率，突出信号分量，研究基于距离扩展目标检测特征的模糊检测方法，有效而稳健地检测目标。搭载某鱼雷型号湖海试验，进行该方法的实验研究。该方法能显著提高高精度自导系统对弱信号的检测能力，提高其对假目标和干扰的对抗能力，为新体制鱼雷的检测手段提供新思路和理论依据。

中文关键词： 波导；检测；恒虚警；模糊逻辑

英文摘要： The detection of target during waveguide in shallow water is different from the traditional sonar signal detection. The first difference is that the statistic property of the reverberation deviates from Rayleigh distribution, and the other difference is that the target becomes a range-extended target with multiple scattering centers instead of point-like target. These differences bring many difficulties for the detection of target. Fuzzy logic is a method to resolve uncertain and non-accurate information. The uncertainty of a sonar signal is often caused by fuzzy property. In this project, we deal with fuzzy constant false alarm rate (CFAR) detection of range-extended target, which will make full use of both multiple scattering centers property of the target and non-Gaussian property of the reverberation. More precisely, we devise a fuzzy membership function to smooth reverberation background, to decrease false alarm rate, and to highlight signal component. Moreover, we propose a fuzzy detection method based on the detection properties of range-extended target in order to detect target more effectively and robustly. By participating in lake and sea trial of some torpedo project, we can verify this method. This method not only can significantly improve the detection performance of high resolution sonar system for weak signal, but also can enhance its anti-interference capability. Therefore, a new idea and the theory basis for detection means of modern torpedo are introduced.

英文关键词： waveguide; detection; CFAR; fuzzy logic