基于膜算法的雷达辐射源信号聚类分选方法研究

项目名称： 基于膜算法的雷达辐射源信号聚类分选方法研究

项目编号： No.61461051

项目类型： 地区科学基金项目

立项/批准年度： 2015

项目学科： 无线电电子学、电信技术

项目作者： 陈韬伟

作者单位： 云南财经大学

项目金额： 43万元

中文摘要： 针对复杂电磁环境下截获的未知雷达脉冲信号，本项目应用自然计算新分支-膜计算，研究基于膜计算优化理论的复杂雷达辐射源信号聚类分选方法。本项目从实现电子对抗侦察系统自动化和智能化的要求出发，与雷达信号分选关系最密切的信号特征挖掘、分析和特征选择为对象，在膜计算优化理论（膜算法）和多目标优化理论的框架下，研究雷达辐射源信号多特征信息融合与特征优化选择，雷达辐射源信号的异常脉冲检测、分析和雷达辐射源信号脉内和脉间参数联合分选模型设计3个相互关联内容的协同研究，解决未知雷达信号分选中先验知识的获取以及膜计算优化算法在工作方式复杂多变的雷达信号分选过程中的有效性和复杂度2个关键问题。根据ELINT、ESM和RWR的特点，实现Map/Reduce计算模型下基于膜算法的复杂体制雷达辐射源信号智能化分选的项目目标，探索拓展膜计算的应用领域，有助于丰富雷达对抗信号处理研究工作中的相关理论。

中文关键词： 雷达辐射源信号；膜算法；多目标优化；智能分选；特征选择

英文摘要： Aiming at intercepted unknown radar pulse signals in complex electromagnetism environment， membrane computing as a new branch of natural computing will be applied to this project, which conducts the research on clustering-based deinterleaving for advanced radar emitter signals based on membrane computing optimization theory. Starting from requirement for realization of automation and intelligence of electronic warfare surveillance system, taking analysis, selection and mining of radar signal features closely related to radar signal sorting as the objects, the project conducts the research under the framework of membrane computing optimization(membrane algorithm) and multi-objective optimization. By three collaborative researches of mutual associated topics on research on information fusion of multi-feature and feature optimization selection for radar emitter signals, outlier detection and analysis of radar pulse signals and the joint model of radar deinterleaving with inter and intra feature parameters, the two key problems, that is How to obtain priori-knowledge in unknown radar signal sorting and the validity and computation complexity of membrane algorithm during radar sorting process under complex and ever-changing environment, will be solved. According to ELINT, RWR and ESM characteristics, it is possible that the project can realize goal of constructing model of intelligent deinterleaving based membrane algorithm under Map/Reduce model and expand application areas of membrane computing, which contributes to enrich related theory in the research work of signal processing for radar countermeasure.

英文关键词： radar emitter signals;membrane algorithm;multi-objective optimization;intelligent deinterleaving;feature selection