文献类型：期刊文献

中文题名：基于改进MCKF的UWB/IMU室内组合定位算法

英文题名：UWB/IMU indoor combination positioning algorithm based on improved MCKF

作者：廖天睿[1,2];吴向东[2];赵林惠[1];金晓明[1];贾之阳[2];戴亚平[2]

第一作者：廖天睿

机构：[1]北京联合大学机器人学院,北京100101;[2]北京理工大学自动化学院,北京100081

第一机构：北京联合大学机器人学院

年份：2022

卷号：41

期号：5

起止页码：127-130

中文期刊名：传感器与微系统

外文期刊名：Transducer and Microsystem Technologies

收录：CSTPCD;;北大核心:【北大核心2020】；CSCD:【CSCD_E2021_2022】；

基金：北京市智能机械创新设计服务工程技术研究中心开放课题项目(KF2019303);北京市教育委员会科技计划一般项目(KM202011417005);国家重点研发计划资助项目(201820655006)。

语种：中文

中文关键词：室内定位;最大相关熵卡尔曼滤波;组合定位算法;超宽带定位;惯性测量单元定位

外文关键词：indoor positioning;maximum correntropy Kalman filtering(MCKF);combination positioning algorithm;ultra wide band(UWB)positioning;inertial measurement unit(IMU)positioning

摘要：针对最大相关熵卡尔曼滤波(MCKF)难以处理室内组合定位系统中非高斯噪声的问题,提出了一种基于改进最大相关熵卡尔曼滤波(I-MCKF)融合方法的UWB/IMU室内组合定位算法。首先,利用新息对观测噪声协方差矩阵进行动态修正,以解决MCKF在被多种噪声干扰时难以准确估计观测噪声协方差矩阵的问题。然后,将修正后的观测噪声协方差矩阵引入MCKF中,以改善MCKF处理非高斯噪声的性能。最后,构建了基于I-MCKF的UWB/IMU室内组合定位算法,以实现室内高精度实时定位。室内动态定位实验的结果表明:基于I-MCKF的组合定位算法可有效抑制室内复杂环境与误差累积对定位结果的影响,相较于MCKF组合定位算法具有更高的定位精度。
Aiming at the problem that the maximum correntropy Kalman filtering(MCKF)is difficult to deal with the non-Gaussian noise in the indoor fusion positioning system,a UWB/IMU indoor fusion positioning algorithm based on the improved-MCKF(I-MCKF)fusion method is proposed.Firstly,to solve the problem that it is difficult for MCKF to accurately estimate the observation noise covariance matrix when disturbed by a variety of noise,the observation noise covariance matrix is dynamically modified by innovation.Then,to improve the performance of MCKF in dealing with Non-Gaussian noise,the modified observation noise covariance matrix is introduced into MCKF.Finally,a UWB/IMU indoor fusion positioning algorithm based on I-MCKF is constructed to realize indoor high-precision real-time positioning.The results of indoor dynamic positioning experiments show that the fusion positioning algorithm based on I-MCKF can effectively suppress the influence of indoor complex environment and error accumulation on the positioning results,and has higher positioning precision than MCKF combined positioning algorithm.