| 引用本文: | 王兴家， 王 彬， 刘岳巍， 晏学成， 丁 峰. 基于元知识转移的认知雷达波形设计[J]. 雷达科学与技术, 2024, 22(4): 443-453.[点击复制] |
| WANG Xingjia, WANG Bin, LIU Yuewei, YAN Xuecheng, DING Feng. Cognitive Radar Waveform Design Based on Meta⁃Knowledge Transfer[J]. Radar Science and Technology, 2024, 22(4): 443-453.[点击复制] |
|
| 摘要: |
| 认知雷达在实际应用中常常需要应对多样性的任务需求，传统的雷达系统往往是针对单一任务进行设计和优化。然而，现实场景中的雷达应用往往需要同时满足多个任务的需求。本文提出了一种多任务雷达波形设计方法，能够根据环境先验知识同时使用多个优化准则进行发射波形的设计。与此同时，引入了基于元知识迁移（Meta?Knowledge Transfer，MKT）的协方差矩阵自适应进化策略算法，通过使用更通用的MKT方法在有限的雷达资源下求解多个雷达任务。该方法通过转移雷达任务求解过程中产生的元知识，来提高每个雷达任务的求解效率。通过仿真实验验证了所提出的求解多任务雷达发射波形算法的可行性。相较于使用进化算法分别求解单一的雷达任务，避免了每个任务都需要从头开始学习优化策略，节省大量的计算资源和时间，加快了最优发射波形的求解速度。 |
| 关键词: 认知雷达 元知识转移 多任务雷达 波形设计 |
| DOI：DOI:10.3969/j.issn.1672-2337.2024.04.011 |
| 分类号:TN951；TN957.51 |
| 基金项目:石家庄铁道大学研究生创新资助项目（No.YC2023024） |
|
| Cognitive Radar Waveform Design Based on Meta⁃Knowledge Transfer |
|
WANG Xingjia, WANG Bin, LIU Yuewei, YAN Xuecheng, DING Feng
|
|
1. Hebei Provincial Collaborative Innovation Center of Transportation Power Grid Intelligent Integration Technology and Equipment,School of Electrical and Electronic Engineering, Shijiazhuang Tiedao University, Shijiazhuang 050043, China;2. Hebei Far East Communication System Engineering Co Ltd, Shijiazhuang 050299, China;3. Beijing Hangtian Xingke High Technology Co Ltd, Beijing 100043, China
|
| Abstract: |
| Cognitive radar often needs to deal with diverse task requirements in practical applications. Traditional radar systems are often designed and optimized for a single task. However, radar applications in real?world scenarios often need to meet the needs of multiple tasks at the same time. In this paper, a multi?task radar waveform design method is proposed, which can use multiple optimization criteria to design the transmitted waveform according to the prior knowledge of the environment. At the same time, a covariance matrix adaptive evolutionary strategy algorithm based on meta?knowledge transfer （MKT） is introduced to solve multiple radar tasks under limited radar resources by using a more general MKT method. This method improves the efficiency of each radar task by transferring the meta?knowledge generated during the radar task solving process. The simulation experiments verify the feasibility of the proposed algorithm for solving the transmitted waveform of multi?task radar. Compared with using evolutionary algorithms to solve a single radar task separately, the algorithm avoids the need to learn the optimization strategy from scratch for each task, saves a lot of computing resources and time, and speeds up the solution of the optimal transmission waveform. |
| Key words: cognitive radar meta⁃knowledge transfer multi⁃task radar waveform design |
