基于环境认知的自适应波形优化设计研究

Cognitive radar is a new concept of radar system which draws widespread concern in recent years. It can overcome the shortcomings that traditional radar has, such as traditional radar transmits single waveform, lacks flexibility and hardly adapts to complex environment. Research of cognitive radar involves optimal control, radar signal processing and information theory and many problems need to solve. This project mainly researches adaptive waveform optimization and design technology, which includes cogniton of environment information based on multi-time-varying-parameter estimation and multi-sensor fusion, modeling method research of adaptive dynamic waveform library based on real-time updating environment information and optimization strategy research of waveform selection based on multi-source environment information in complex environment. Manpower considering transmitted waveform, environment, target and actual constraints, we can set up multi-constraint time-varying multi-parameter waveform optimization model based on mutual information, and then waveform library can be adaptively dynamic updated. For complex environment, we can set up nonlinear multi-stage waveform selection and optimization model in multi-source environment information based on stochastic dynamic programming, so that best match between transmitted waveform and environment information can be realized. The results of this project can supply theoretical basis of waveform library setup and transmitted waveform selection in cognitive radar, and can also supply effective tools for areas such as military, aerospace and communications applications.

认知雷达是近年来受到广泛关注的新概念雷达系统，它能够克服传统雷达发射波形单一，缺乏灵活性，很难适应复杂多变环境等缺点。认知雷达的研究涉及到最优控制、雷达信号处理和信息论等诸多学科，很多问题亟待解决。本项目主要研究基于环境认知的自适应波形优化设计技术，包括基于多时变参数估计和多传感器融合的环境信息的认知、基于实时环境信息的自适应动态波形库的建模方法研究、复杂环境下基于多源环境信息的波形选择优化策略研究。统筹考虑发射波形、环境和目标信息以及实际限定条件，基于互信息建立多约束时变多参数波形优化模型，进而自适应动态更新波形库。面向复杂环境，基于随机动态规划建立多源环境信息下非线性多阶段波形选择优化模型，从而实现发射波形与环境信息的最佳匹配。本项目的研究成果可以为认知雷达系统波形库的建立和发射波形的选择提供理论依据，为军事、航天和通信等领域的应用提供有效工具。

智能化是现代雷达发展的大方向，探测距离更远，跟踪精度更高，识别目标更准，抗干扰能力更强，反应速度更快。认知雷达是近年来受到广泛关注的新体制雷达系统，属于雷达领域和控制领域交叉研究的前沿课题。认知雷达能够克服传统雷达发射波形单一，缺乏灵活性，很难适应复杂多变环境等缺点，在不同的情况下根据环境信息选择发射适应的波形。本项目的主题围绕自适应波形优化设计展开，主要内容一是深入研究了各种杂波的性质及生成方法，作为环境认知的基础，为自适应波形库的建立提供依据；二是考虑实际环境因素限定条件，建立了波形设计的互信息模型，提出了相应的波形优化设计方法；三是基于随机动态规划理论建立波形自适应选择模型，提出了相应的波形选择算法。通过仿真结果可以看到，所提波形优化算法和波形选择算法都具有良好的性能，提高了雷达系统的综合能力。本项目的研究成果可以为认知雷达系统波形库的建立和发射波形的选择提供理论依据，为军事、航天和通信等领域的应用提供有效工具，有助于我国在未来雷达技术的发展中把握主动，占领先机。