Research On Indoor Pedestrian Navigation Algorithm Based On Inertial Multi-Sensor Fusion

Indoor pedestrian navigation can provide pedestrians with accurate location navigation services in a complex and enclosed indoor environment,so the research value of indoor navigation is increasingly prominent.This article will focuses on the current problems of pedestrian navigation,and conducts research on indoor pedestrian navigation from the direction of inertial multi-sensor fusion.The research content of this article mainly includes the following parts:This article expounds the research on the related theory and technology of pedestrian navigation.Among then,the basic technical architecture of Pedestrian Dead Reckoning(PDR)is explained,the basic theoretical formula of Strapdown Inertial Navigation System(SINS)is discussed,the multi-sensor fusion algorithm is introduced,and the Kalman Filter algorithm is emphasized.Aiming at the poor performance of Global Navigation Satellite System in indoor navigation applications and high cost of wireless sensor network positioning,a multi-sensor fusion indoor pedestrian navigation algorithm that only relies on the information of inertial sensors such as gyroscope and accelerometer is studied.First,the SINS algorithm and the zero-speed gait detection algorithm are used to process the acquired gyroscope and accelerometer data to obtain navigation information such as pedestrian attitude,speed,and position,as well as pedestrian zero-speed Gait information.Then,based on the zero-velocity fusion strategy,the Extended Kalman Filter algorithm is used to fuse the zero-velocity information and the pedestrian navigation information of the SINS solution method,so that the overall navigation and positioning accuracy of indoor pedestrian navigation is improved.Finally,the performance of the indoor pedestrian navigation system under this method is verified by experiments.Aiming at the problem of low accuracy when using Kalman Filter for indoor pedestrian navigation data fusion,a multi-sensor fusion indoor pedestrian navigation algorithm based on Rank Kalman Filter is studied.By introducing the Rank Kalman Filter technology into the zero-velocity update fusion strategy under the PDR architecture,the errors caused by the nonlinear and non-Gaussian noise problems of the indoor pedestrian navigation system are better suppressed under the action of the rank sampling mechanism.The proposed algorithm performs fusion processing on the multi-sensor data measured during indoor pedestrian movement,which can realize more accurate indoor pedestrian navigation.And the test results display that the algorithm has a certain improvement compared with the zero-velocity fusion PDR algorithm using the Extended Kalman Filter method,which reduces the indoor pedestrian navigation positioning error by 18.91%.