GPS定位中2种自适应滤波的组合算法  

Combination Algorithm of Two Kinds of Adaptive Filter in GPS Positioning

作　　者：田宠[1] 王兴亮[1] 卢艳娥[1] 

机构地区：[1]空军工程大学电讯工程学院,陕西西安710077

出　　处：《现代防御技术》2012年第3期72-77,共6页Modern Defence Technology

基　　金：国家自然科学基金(61071014);学院科研创新基金(DYCX1002)

摘　　要：分析了卡尔曼滤波发散的原因,并给出了指数加权的衰减自适应记忆滤波和噪声加权自适应滤波2种算法用来抑制发散,相应地分析了衰减因子的选取以及噪声模型的在线估计,最后提出这2种定位方式组合的定位算法。仿真结果表明,衰减自适应记忆滤波和噪声加权自适应滤波明显提高了卡尔曼滤波的定位精度并且抑制了发散,组合算法在提高定位精度、抑制发散的同时,使这2种定位方式的优点形成了互补,增加了算法的稳定性。The reasons of the instability of Kalman filtering are analyzed, which indicates that the exponentially-weighted are the two ways to pre ding vent adaptive memory filtering and noise exponentially-weighted a filtering divergence, and the fading factor selection as well as daptive filtering the noise model online estimation is analyzed correspondingly. Finally, the position algorithm combining with two positioning methods is proposed. The simulation result indicates that fading adaptive memory filtering and noise exponentially-weighted adaptive filtering obviously increase the position accuracy of Kahnan filtering, and restrain the filtering divergence. In addition, the combinatorial algorithm makes the advantages of the two position methods complement each other, which increases the stability of the algorithm.

关 键 词：卡尔曼滤波 发散 衰减因子 定位精度 

分 类 号：P228.4[天文地球—大地测量学与测量工程] TN713[天文地球—测绘科学与技术]