针对传统的无味卡尔曼相位展开算法效率和精度之间存在矛盾的问题,提出一种快速局部频率估计算法与量化质量图跟踪策略相结合的无味卡尔曼滤波(UKF)高效算法。快速局部频率估计算法首先利用二维傅里叶变换粗略计算出局部窗口内的复相位频谱的谱峰位置;然后,以谱峰为中心利用二维离散时间傅里叶变换快速算法计算出谱峰的精确位置,进而精确地获取局部频率值,得到UKF 相位展开系统模型所需的相位梯度估计值;最后,将UKF 算法与量化质量图跟踪策略相结合,利用量化质量图引导UKF 相位展开,进一步提高展开的效率。仿真实验验证了方法的有效性。

As for the contradiction between the efficiency and the precision of traditional unscented Kalman filtering phase unwrapping algorithm, an efficient phase unwrapping algorithm based on unscented Kalman filter (UKF) is proposed by combining a fast local frequency estimation approach and the strategy of quantizing paths-guided map. First, the spectral peak position of plural phase is calculated by the two-dimensional Fourier transform in the local window. Then, the accurate position of the spectral peak is calculated by the two-dimensional discrete-time Fourier transform, and then accurate values of local frequency to calculate the local frequency is gotten, according to the results, the phase gradient estimates for the UKF phase unwrapping system model are calculated. In the end, the UKF algorithm is combined with the strategy of quantizing paths-guided map to guide UKF phase unwrapping, further improving the efficiency of the unwrapping. The simulation results show the effectiveness of the proposed algorithm.

TN957