This paper considers the problems of high-speed maneuvering target detection and motion parameters estimation with complex motions, involving range migration (RM) and Doppler frequency migration (DFM) induced by target’s radial velocity, acceleration and jerk within the coherent integration interval. To address these problems, we propose a novel method based on the third-order keystone transform (TOKT) and first-order discrete polynomial-phase transform (FDPT) in this paper. Firstly, the TOKT is carried out to correct the third-order RM. Secondly, the matched filtering function is constructed to remove the quadratic DFM caused by radial jerk. After that, the FDPT is applied to reduce the target phase order. Then, the TOKT is employed again to eliminate the residual first-order RM related to target’s radial acceleration, and a well-focused peak is acquired after azimuth fast Fourier transform (FFT) as well as range inverse FFT. Finally, based on the position of peak, the motion parameters of maneuvering targets can be estimated efficiently. Compared with the existing algorithms, the proposed method achieves a good balance between the computational complexity and detection ability. Numerical results verify the effectiveness of the proposed method.

中文翻译：

---

一种高速机动目标检测与运动参数估计的新方法

本文考虑了具有复杂运动的高速机动目标检测和运动参数估计问题，涉及相干积分区间内目标径向速度、加速度和加加速度引起的距离偏移(RM)和多普勒频率偏移(DFM)。为了解决这些问题，我们在本文中提出了一种基于三阶梯形变换（TOKT）和一阶离散多项式相位变换（FDPT）的新方法。首先进行TOKT校正三阶RM。其次，构造匹配滤波函数以去除由径向跳动引起的二次DFM。之后，应用 FDPT 来降低目标相序。然后，再次使用TOKT消除与目标径向加速度相关的残余一阶RM，并且在方位角快速傅里叶变换 (FFT) 以及距离逆 FFT 之后获得了一个聚焦良好的峰值。最后，基于峰值位置，可以有效地估计机动目标的运动参数。与现有算法相比，该方法在计算复杂度和检测能力之间取得了良好的平衡。数值结果验证了所提出方法的有效性。