论文部分内容阅读
提出一种结合无迹卡尔曼滤波(UKF)和小波阈值自适应滤波的混合粒子滤波方法。该算法在粒子产生过程中采用UKF方法以克服粒子发散,而在数据预处理和粒子产生时采用自适应阈值小波变换算法抑制观测数据噪声和数据处理过程误差。比较了EKF-PF、UKF-PF和WM-UKF-PF不同算法的性能。结果表明,经过改进的WM-UKF-PF混合粒子滤波算法能够有效地降低均方根误差,提高信噪比和相邻恒心日观测值的相关性,从而改善解算结果的统计特性。所采取的混合滤波算法的计算复杂性提高了,但能够有效减轻GPS多路径效应影响,对高精度定位测量和非平稳形变特征提取具有重要意义。
A hybrid particle filter method based on unscented Kalman filter (UKF) and wavelet threshold adaptive filtering is proposed. This algorithm uses the UKF method to overcome the particle divergence in the particle generation process, and uses the adaptive threshold wavelet transform algorithm to suppress the observed data noise and data processing error during data preprocessing and particle generation. The performance of different algorithms of EKF-PF, UKF-PF and WM-UKF-PF are compared. The results show that the improved WM-UKF-PF hybrid particle filter algorithm can effectively reduce the root-mean-square error and improve the correlation between the signal-to-noise ratio and the adjacent observational data, so as to improve the statistical properties of the solution. The computational complexity of the hybrid filtering algorithm is improved, but it can effectively reduce the impact of GPS multipath effect and is of great significance for high-precision positioning and non-stationary deformation feature extraction.