该文针对传统的基于二次时频分析和原子追踪匹配方法处理多分量LFM信号时存在的时频干扰和等振幅交叉分解等问题,提出了一种基于多尺度线性调频基信号稀疏分解的多分量LFM信号检测方法,该方法采用多尺度的线调频基函数对多分量LFM信号进行投影分解,通过从不同的时间支撑区内投影系数最大的基函数中寻找出使分解信号能量最大的基元函数组合,逐次获得信号包含的能量最大的LFM信号分量,从基元函数连接形成的频率曲线即可获得LFM信号分量瞬时频率的估计,再对分量瞬时频率求起始时间点的频率值和曲线斜率便可得到该LFM分量的中心频率及调频斜率,仿真试验表明该文方法能精确地提取等振幅多分量LFM信号的瞬时频率,并具有很强的抗噪声干扰能力。
A multicomponent LFM signals detection method based on multi-scale chirplet sparse signal decomposition is proposed to overcome the problems of cross-interference and cross-decomposition existed in the multicomponent LFM signals processing based on traditional quadratic time-frequency analysis and the atomic matching pursuit method.This method projects the multicomponent LFM signals onto the multi-scale chirplet base functions.Through the different support regions of time in which the projection coefficient is the largest,this method can separate the LFM signal component with the largest energy from the multicomponent LFM signals.The instantaneous frequency of the LFM signal component is estimated by connecting the chirplet base functions used in the current separation.Though seeking for the initial point and the slope of the instantaneous frequency curve,the center frequency and FM slope of the LFM signal component have been estimated.Simulation experiments show that the proposed method can accuratelly extract the instantaneous frequency of the multicomponent LFM signals with a strong anti-noise interference ability.