FMCW雷达周期非线性相位估计与矫正

为获取高质量距离像,解线频调接收体制的调频连续波雷达通常需要进行非线性矫正。传统基于多项式模型的方法虽然能够矫正系统中绝大部分的非线性失真,但难以抑制由周期非线性引起的成对回波,距离压缩质量仍有较大提升空间。本文将剩余的周期非线性误差建模为多分量正弦调频信号,提出一种匹配初值与非线性最小二乘相结合的参数估计方法。所提算法仅采用傅里叶变换以及一维相位搜索即可实现正弦调频参数的快速估计,然后利用非线性最小二乘进一步提高估计精度,最后利用匹配傅里叶变换进行空变非线性矫正。仿真以及实测数据均表明,本文所提算法能够有效消除周期相位误差,抑制成对回波,显著改善距离压缩质量。

Nonlinearity phase estimation and correction of FMCW radar

In order to obtain a high-quality range image, frequency modulated continuous wave (FMCW) radar usually requires modulation nonlinearity correction under the mechanism of dechirp reception. Traditional methods based on the polynomial phase model can correct most nonlinear distortion in the system, but it is hard to eliminate the paired echoes invited by period nonlinearity, which reveals great improvement still exists in range compression. The residual nonlinearity error is modeled as a multi-component sinusoidal FM signal, and after that, a parameter estimation method combining initial value matching and nonlinear least squares optimization is proposed. The method can achieve high efficiency in the estimation of sinusoidal frequency modulation parameters via Fourier transform and one-dimensional phase searching. Then, the nonlinear least squares is used to further improve the estimation accuracy. Finally the range-dependent nonlinearity is corrected by match Fourier transform. Simulation and radar echo processing results demonstrate that the proposed method can effectively eliminate periodic phase nonlinearity, suppress paired echoes, and significantly improve range compression quality.