一种基于改进自适应卡尔曼滤波的GNSS/INS组合导航算法  

A GNSS/INS Integrated Navigation Algorithm Based on Modified Adaptive Kalman Filter

作　　者：吴晓倩 卢秀山[2] 王胜利[2] 王鸣鹤 柴大帅 WU Xiao-qian;LU Xiu-shan;WANG Sheng-li;WANG Ming-he;CHAI Da-shuai(College of Geomatics,Shandong University of Science and Technology,Qingdao 266000,China;College of Ocean Science and Engineering,Shandong University of Science and Technology,Qingdao 266000,China;School of Environment Science and Spatial Informatics,China University of Mining and Technology,Xuzhou 221000,China)

机构地区：[1]山东科技大学测绘科学与工程学院,青岛266000 [2]山东科技大学海洋科学与工程学院,青岛266000 [3]中国矿业大学环境与测绘学院,徐州221000

出　　处：《科学技术与工程》2020年第3期913-917,共5页Science Technology and Engineering

基　　金：国家重点研发计划(2016YFB0501705)。

摘　　要：在全球卫星导航系统/惯性导航系统(global navigation satellite system/inertial navigation system,GNSS/INS)组合系统中,状态模型误差和异常扰动的影响严重降低了标准卡尔曼滤波的性能,而基于预测残差自适应的卡尔曼滤波随计算次数的增加滤波效果降低,且使用统一的自适应因子调节不可靠。针对上述问题,提出一种改进算法,利用预测残差建立的统计量调节位置向量和速度向量,避免了其他参数对滤波的平衡作用;通过预测残差的概率密度建立马氏距离进行假设检验,在模型正常时使用标准卡尔曼滤波,模型异常时使用改进滤波算法;采用实测车载数据对标准卡尔曼滤波、单因子自适应滤波和本文的滤波方法进行评估,实验结果表明:改进的自适应卡尔曼滤波的滤波算法效果良好,证明了所提算法的有效性。The performance of Kalman filtering is severely degraded by state model errors and the effects of abnormal deviations in the loosely coupled GNSS/INS(global navigation satellite system/inertial navigation system)integration system.However,the adaptive Kalman filter basing on the predicted residuals reduced filtering effect with increasing number of calculations.And it is unreliable using uniform adaptive factors.In response to these problems,a modified adaptive Kalman filter algorithm is proposed.It uses the statistic established by the prediction residuals to adjust the position vector and velocity vector,which avoids the balance of other parameters on the filtering.And it establishes the Mahalanobis distance hypothesis test by the probability density of prediction residuals,uses standard Kalman filtering when the model is normal,and uses modified adaptive Kalman filtering when the model is abnormal.Kalman filtering,single-factor adaptive filtering and the filtering methods are evaluated using measured vehicle data.The experimental results show that the modified adaptive Kalman filter has good filtering effect and proves the effectiveness of the proposed algorithm.

关 键 词：组合导航 卡尔曼滤波 马氏距离 自适应因子 

分 类 号：P228[天文地球—大地测量学与测量工程]