基于非线性滤波和平滑相迭代 GPS 定位估计算法.pdfVIP

基于非线性滤波和平滑相迭代的GPS 定位估计算法 曹意，茅旭初1 （上海交通大学仪器科学与工程系，上海 200240 ） 摘要：提出了一种卡尔曼滤波与平滑相迭代的 GPS 定位估计算法，每一时刻点先用平淡卡 尔曼滤波进行位置估计，然后利用滤波结果对前一时刻点的位置估计进行平滑修正，接着再 用修正值进行滤波得到当前时刻点更新的估计值，通过若干次迭代，得到最终估计值。后向 平滑算法分别采用固定区间平滑和固定滞后平滑，前者适用于非线性系统模型，后者在平滑 步长增大时，其计算量较小。实验结果表明，迭代算法能够有效提高定位精度，不仅可以应 用在后验数据处理中，而且也可以应用于GPS 实时定位估计。 关键词：全球定位系统；最优平滑；迭代算法；平淡卡尔曼滤波 中图分类号：TN976.1 文献标识码：A A Nonlinear Iterative Filtering-Smoothing Algorithm for GPS Positioning Cao Yi, Mao Xu-chu (Dept. of Instrument Science Engineering, Shanghai Jiao Tong Univ., Shanghai 200240, China) Abstract: A new GPS positioning algorithm is proposed, which iteratively applies UKF and smoothing algorithm. UKF is employed to estimate user ’s position at each epoch, the estimation is used to smooth the estimation of previous epoch, then the updated estimation can be taken as the initial value for kalman filter in turn, the filtering algorithm is applied again. This process proceeds with several iterations ending up with final estimation. Two smoothing algorithms are addressed, Fixed-Interval smoother and Fixed-Lag smoother, the former can be applied to nonlinear system, and the latter effectively reduces the computational burden when smoothing step increases. GPS experiment result reveals that the iterative algorithm amounts to a method contributing to accuracy improvement, it can be used not only in the post-processing, but also in real-time GPS positioning estimation. Key words: global positioning system (GPS); optimal smoother; iterative algorithm; unscented Kalman filter (UKF) [1] 在GPS 定位中，应用扩展卡尔曼滤波 （Extended Kalman Filter, EKF）和平淡卡尔曼 [2] 滤波 (Unscented Kalman Filter, UKF)对接收机的位置和速度进行估计被证明