抗差容积卡尔曼滤波及其在UWB室内定位中的应用  

Robust cubature Kalman filter and its application in UWB indoor positioning

作　　者：蔡赣飞 徐爱功[1] 洪州 隋心[1] CAI Ganfei;XU Aigong;HONG Zhou;SUI Xin(School of Geomatics,Liaoning Technical University,Fuxin,Liaoning 123000,China;Tieling Surveying and Mapping Management Office,Tieling Municipal Planning Bureau,Tieling,Liaoning 112000,China)

机构地区：[1]辽宁工程技术大学测绘与地理科学学院,辽宁阜新123000 [2]铁岭市规划局铁岭市测绘管理办公室,辽宁铁岭112000

出　　处：《测绘科学》2018年第12期123-129,共7页Science of Surveying and Mapping

基　　金：辽宁省高等学校创新团队项目(LT2015013);国家重点研发计划项目(2016YFC0803102)

摘　　要：针对超宽带(UWB)观测值异常引起的量测误差及系统噪声先验统计信息未知而导致状态估计误差增大的问题,该文提出了一种带噪声时变估计器的抗差容积卡尔曼滤波(CKF)算法。该算法在滤波过程中,利用预报残差因子构建抗差等价协方差矩阵,控制观测异常值对滤波参数解的影响,同时利用sage_husa算法对系统噪声的统计特性进行实时估计和修正,提高滤波精度和稳定性。实验结果表明,所提算法不仅能有效地消除量测误差对滤波解的影响,而且能在系统噪声先验信息未知的情况下更进一步提高UWB室内定位的精度和可靠性。To solve the problem that the measurement error caused by ultra wide band(UWB)observations anomaly and the state estimation error increase caused by which system noise priori statistical information is unknown,a robust cubature Kalman filter(CKF)with time-varying noise estimator was proposed.In the filtering process,a robust residual equivalent covariance matrix was constructed by using the forecast residual factor to control the effect of observation anomalies on the filter parameter solution.At the same time,the statistical characteristics of the system noise were estimated and corrected in real time by using the sage_husa algorithm to improve the filtering accuracy and stability.Experimental results showed that the proposed algorithm not only could effectively eliminate the influence of measurement error on the filtering solution,but also could improve the accuracy and reliability of UWB indoor positioning even if the system noise prior information is unknown.

关 键 词：量测异常 系统噪声 UWB室内定位 抗差CKF sage_husa算法 

分 类 号：P225[天文地球—大地测量学与测量工程]