针对逆合成孔径激光雷达对机动目标成像时存在方位多普勒时变的问题, 提出了一种基于方位时频域keystone变换的机动目标逆合成孔径激光雷达方位成像快速算法.利用多分量线性调频子回波信号的调频斜率与起始频率的比值为常量这一特点, 在方位时频域采用keystone变换将多分量线性调频信号同时转换为多分量单频信号, 利用快速傅里叶变换实现方位聚焦.采用基于分数阶傅里叶变换和最小熵的线性调频参量估计方法, 实现了对调频斜率与起始频率比值的精确、快速估计.结果表明, 与现有的基于Radon-Wigner变换的距离-瞬时多普勒成像算法相比, 所提出的算法成像效率大大提高, 且能够保留更多的目标细节信息, 适合于逆合成孔径激光雷达的实时成像.

The Doppler frequency time-variation exist in the received echo when imaging maneuvering targets by Inverse Synthetic Aperture Laser radar (ISAL). Based on a keystone transform in the azimuth time-frequency domain, a fast azimuth imaging algorithm of ISAL for maneuvering targets is proposed. Since the ratio of the chirp rate to the initial frequency is constant for all multicomponent linear frequency modulation subechoes in each range cell, all these multicomponent linear frequency modulation subechoes can be simultaneously transformed into multi-component single frequency subechoes by using the keystone transform in the azimuth time-frequency domain, and hence fast Fourier transform can be used for azimuth focusing. In addition, a precise and efficient method based on the fractional Fourier transform and the minimum entropy is proposed to estimate the ratio of the chirp rate to the initial frequency. Simulation results prove that, comparing with the range-instantaneous Doppler algorithm based on the Radon-Wigner transform, the proposed algorithm is more efficient and can preserve more information details of the target.