GPS动态相对定位的自适应Kalman滤波算法

GPS动态数据处理中广泛应用卡尔曼滤波，经典Kalman滤波认为预报误差是白噪声，服从零均值的正态分布，并利用动态噪声协方差矩阵来控制它对当前信息的影响，但实际测量定位中难以保证观测对象的规则运动，因而容易出现模型误差。针对GPS动态定位的这一问题，探讨了在实际应用中存在模型误差时的卡尔曼滤波，介绍了一种自适应Kalman滤波算法，该法顾及了载体机动加速及接收机发生周跳时的影响，减少了滤波发散的机会。

Adaptively Kalman Filtering for Relative Dynamic GPS Survey

Kalman filtering have extension application in GPS kinetic data procession. Classic Kalman filtering deems that predict bias is white noises which obey zero mean value normal school, and controls current information effect by dynamic noise variance matrices. In fact, it is impossible keep normal moving and so it is very easy come forth error. This pa- per discusses on the Kalman filtering that have errors in apply and introduces a new adaptive- ly Kalman filtering approach, considering body rand accelerate and circle slips effect and the chance of divergence deduced.