雷达回波信号通常是同时从同一个天线、同一频段进入接收系统的多个不包含空间信息的非平稳信号。为了实现多目标信息的有效利用，多分量非平稳信号分解理论研究得到高度重视，成为现代雷达精细化处理的前提。本项目针对时频分解中面临分辨率差、分解分量存在模糊频率、时频域交叉的信号分量难以分离等难题，分析多分量非平稳信号在高阶空间中的表现，基于非平稳信号分量在瞬时调频率函数上的差别，提出时间-调频率域信号分解方法以精确地分解时频域交叉的分量；基于非平稳信号分量在瞬时频率函数上的差别，提出基于WVD的无模糊信号分解方法以精确、无模糊地分解时频域距离较近的分量。综合利用时间-调频率分解方法和时频分解方法，形成一套系统的多分量非平稳信号分解理论与方法，自适应地得到各信号分量的时域或频域波形，为复杂场景下目标回波检测与识别奠定基础。发表高水平论文20篇以上；申请国家发明专利8项以上；培养博士、硕士研究生10人以上。

In the illuminating area, there are more than one targets or reflection points. Radar echoes are composed of multiple non-stationary components, which are located at the same frequency band, without spacial information, and received by the same antenna simultaneously. To analyze every target, multi-component non-stationary signal decomposition method attracts plenty of attention, and will be the premise of the meticulous processing of the radar signal. This project aims at solving the poor resolution, the aliasing frequency of the decomposed component, and the decomposition of the components with crossing instantaneous frequencies in time-frequency decomposition. To show the difference between the instantaneous frequencies and instantaneous frequency rates, the time-high order parameter representation is considered, including time-frequency representation and time-frequency rate representation. 1）For the components with crossing instantaneous frequencies, we propose a time-frequency rate decomposition method. 2) For the components with close instantaneous frequencies, we propose a time-frequency decomposition method, without aliasing frequency. 3) By combining the two signal decomposition methods in the time-high order parameter representation, a signal decomposition system is obtained and provides the every component in sense of time or frequency adaptively. The decomposition system will be a base for the signal target detection and recognition in complex scene. It is expected to publish over 20 high quality papers, apply for over 8 patents, and train over 10 Ph.D. or Master students.