基于自适应动态规划的波形优化方法及在雷达通信中的应用

项目名称： 基于自适应动态规划的波形优化方法及在雷达通信中的应用

项目编号： No.60874108

项目类型： 面上项目

立项/批准年度： 2009

项目学科： 武器工业

项目作者： 汪晋宽

作者单位： 东北大学

项目金额： 30万元

中文摘要： 随着人类对宇宙空间的不断探索、开发和利用，电磁信号环境日益复杂，对雷达性能的要求越来越高，传统波形选择算法比较单一，无法满足实际的需要。针对这一问题，本项目设计了基于自适应动态规划的波形优化器，是认知雷达关键技术研究中的一个重要课题，包括环境参数的辨识、目标函数的构造、波形选择及波形优化等问题，并获得了一系列的研究成果。首先研究了复杂电磁环境下信号特征参数的提取方法；在此基础上，提出了波形选择问题的动态规划模型。该模型能够反映出雷达发射不同波形和检测不同目标时的代价差异，为动态规划方法在波形选择问题中的应用提供了理论基础；基于此模型，提出了基于动态规划的最优波形选择算法，从而能够得到最优波形序列；当状态空间和波形空间较大时，给出了基于Q学习的波形选择算法，提高了计算效率，避免计算状态观测转移概率；为了实现智能的认知雷达，需要同时考虑检测任务和估计任务，提出了基于互信息的最优波形设计算法，给出了认知雷达发射波形的设计准则。本项目的研究成果可以为未来的雷达系统和通信系统提供理论依据和技术储备，为军事、航天和通信等领域的应用提供有效工具。

中文关键词： 环境认知；自适应动态规划；波形优化

英文摘要： With the constant aerospace exploration, development and utilization, the complex electromagnetic signal environment results in increasingly higher demands towards radar performance. And traditional single waveform selection algorithms cannot meet those practical demands. Aiming at this problem, this project designs an adaptive-dynamic-programming-based waveform optimizer. As an important topic of cognitive radar key technologies, this project researches on environmental parameters identification, objective function formation waveform selection as well as waveform optimization, and fruitful achievements are acquired. The study firstly researched the extract method of signal characteristics in complex electromagnetic environment, proposing the dynamic planning model of waveform selection problems, which reflects cost difference from transmitting different waveforms and detecting different objectives and provides a theoretical basis to adaptive dynamic programming's application in waveform selection. Based on this model, the project then proposes an adaptive-dynamic-programming-based waveform selection algorithm to acquire the optimizing waveform sequence. A Q-Learning-based waveform selection algorithm is further developed, the calculation efficiency is increased, and the state observation transition probability calculation is avoided, considering the detection and estimation to achieve the smart cognitive radar, this project proposes the mutual-information-based optimal waveform design algorithm, and presents the cognitive radar transmission waveform design criterion. The research achievements of this project shall provide theoretical basis and technical reserve to future radar system and communication system, and shall be effective tools for military, aerospace, communication and other fields.

英文关键词： Environmental cognition; Adaptive Dynamic Programming; Waveform Optimization