Research On Indoor Location Technology Based On MEMS-IMU

With the development of technology and smart city,sensors and computing units continue to be miniaturized.The extensive use of mobile devices not only increases the demand of location information,but also promotes the development of location-based services industry.Global navigation Satellite System(GNSS)has high positioning accuracy in outdoor,but it can’t be positioned normally in indoor due to the poor coverage of satellite signals.Human beings often live indoors,and indoor positioning technology can be applied to medical care,functional management,emergency rescue and other fields,which can provide a lot of convenience for life and improve the quality of human life.Because of the indoor environment is complex,the indoor positioning technology based on wireless signals is easily affected by the environment.In addition,basic equipment needs to be installed.The technology is complex and costly.With the development of micro electro-mechanical system(MEMS),the inertial measurement unit(IMU)based on MEMS has the advantages of low cost,small size,and wide use,so it’s the best choice for the realization,the application and promotion of indoor positioning technology.Aiming at the key problems in indoor positioning technology based on MEMS-IMU,such as positioning error accumulation,heading drift and height channel divergence of positioning system,the study from pedestrian gait characteristics,attitude update filtering algorithm and height channel of pedestrian navigation system.The main research contents are as follows:(1)Aiming at the problems of error accumulation and heading drift in 2D inertial navigation positioning,an indoor positioning method based on low-cost IMU is studied.Firstly,based on the characteristics of pedestrian gait,an adaptive zero-speed detection algorithm is designed to solve the problem of low detection accuracy of zero-speed detection algorithm with fixed threshold in complex gait.Then,the accuracy of attitude calculation is improved by changing the working mode of PI controller in Mahony filtering algorithm.Finally,pedestrian steering is judged by the z-axis angular velocity data of gyroscope,and pedestrian heading is corrected.(2)In order to solve the problem of height channel divergence and heading drift in 3D inertial navigation,a 3D indoor positioning method based on EKF multi-sensor information fusion is studied.Firstly,a filtering algorithm based on the fusion of Mahony filter and EKF is proposed.The algorithm abandons the commonly used heuristic drift elimination(HDE)algorithm and corrects the heading by acceleration.The difference between the acceleration obtained by Mahony filter and the acceleration calculated by strapdown inertial navigation algorithm is used as the error of pedestrian acceleration,and the heading is corrected by the acceleration information.Then,based on the zero-speed detection algorithm,the errors of pedestrian velocity and angular velocity are obtained to reduce the error accumulation.Moreover,add the barometer,and we correct the barometer output and IMU height output.The difference between the two height outputs can be used to obtain the pedestrian height error and improve the stability of the system height channel.Finally,the information of each sensor is fused based on EKF to update the position and posture of pedestrians.Experiments on indoor linear walking,rectangular paths with different speeds,up and down stairs show that the positioning system in this thesis can effectively reduce error accumulation,decrease heading drift and improve positioning accuracy.