在动态定位数据处理中,动态定位的精度和可靠性除受观测偶然误差和系统误差的影响外,还受时间相关的观测噪声的影响.在观测噪声不满足时间不相关的假设情况下,卡尔曼滤波将达不到最优滤波效果,并且其误差协方差阵也是错误的.分析了测量噪声时间相关对卡尔曼滤波结果的影响。给出了观测噪声时间相关时的卡尔曼滤波递推公式.实例计算结果表明,采用该算法能够有效地消除测量噪声相关性对滤波结果的影响.
The accuracy and reliability of the kinematic positioning are affected by not only the random noises and systematic wrong, but also the observation noises related to time. When the observation noise is temporal correlation, Kalman Filter will not be able to achieve optimization, and its covariance matrix will be wrong. The research analysed the influence of observational noises about temporal correlation, and gives recursive formula of Kalman Filtering. An example shows that the given algorithm is able to eliminate the influence of correlated observational noises effectively upon Kalman filter result.