基于Kalman滤波的动力学模型误差估计算法  

An algorithm based on Kalman filtering for estimating kinematic model errors

作　　者：高为广[1,2] 张晓东[3] 

机构地区：[1]信息工程大学测绘学院,郑州450052 [2]北京跟踪与通信技术研究所,北京100094 [3]西安测绘研究所,西安710054

出　　处：《测绘科学》2011年第2期56-58,共3页Science of Surveying and Mapping

基　　金：国家自然科学基金(40604003);"863"项目(2007AA12Z331)

摘　　要：本文分析介绍了模型误差对滤波解和预报残差影响的表达式。随后,针对GPS/INS松组合导航系统观测信息无冗余的情况,给出了基于Kalman滤波的动力学模型误差估计算法。最后利用一个车载实测数据证明了算法的有效性。Firstly, the influences of the dynamical model errors on navigation results and the predicted residuals were analyzed and introduced in the paper. Then, a new algorithm based on Kahnan filtering for estimating kinematic model errors was put forward. Finally, a practical example was given. The results proved that the new algorithm was valid and reliable when applied in loosely-coupled GPS/INS integrated navigation system.

关 键 词：GPS/INS松组合导航 模型误差 KALMAN滤波 

分 类 号：TN96[电子电信—信号与信息处理]